A report on the design and implementation of the
EU Kids Online survey of 9-16 year olds and
their parents in 25 countries
Technical Report and User Guide:
The 2010 EU Kids Online Survey
Sonia Livingstone, Leslie Haddon, Anke Görzig
and Kjartan Ólafsson, with members of the EU
Kids Online network
ISSN 2045-256X
www.eukidsonline.net
2
Technical Report and User Guide: The 2010 EU Kids Online Survey. This technical report describes the design and
implementation of the EU Kids Online survey of 9-16 year old internet using children and their parents in 25 countries
European countries. It has been produced on behalf of the project Coordinator by Sonia Livingstone, Leslie Haddon, Anke
Görzig and Kjartan Ólafsson, with members of the EU Kids Online network (Annex 2), as advised by the International
Advisory Panel (Annex 1). It builds on the technical survey report delivered by the fieldwork agency Ipsos MORI as part of
their contract with the London School of Economics and Political Science (LSE).
Cite this report as: Livingstone, S., Haddon, L., Görzig, A. and Ólafsson, K. (2011). Technical Report and User Guide: The
2010 EU Kids Online Survey. LSE, London: EU Kids Online.
Note that the dataset is archived in the UK Data Archive and available for public (but not commercial use). See
http://www.data-archive.ac.uk/
Previous reports and publications from EU Kids Online include:
Final recommendations for policy, methodology and research (O’Neill, B., Livingstone, S. and McLaughlin, S.,
2011)
Disadvantaged children and online risk (Livingstone, S., Görzig, A., and Ólafsson, K., 2011)
EU Kids Online Final Report (Livingstone, S., Haddon, L., Görzig, A., and Ólafsson, K., 2011)
Risks and safety on the internet: The perspective of European children. Full findings (Livingstone, S., Haddon, L.,
Görzig, A., and Ólafsson, K., 2011)
Risky communication online (Livingstone, S., and Ólafsson, K., 2011)
Digital literacy and safety skills (Sonck, N., Livingstone, S., Kuiper, E., and de Haan, J., 2011)
Social networking, age and privacy (Livingstone, S., Ólafsson, K., and Staksrud, E., 2011)
Patterns of risk and safety online. In-depth analyses from the EU Kids Online survey of 9-16 year olds and their
parents in 25 countries (Hasebrink, U., Görzig, A., Haddon, L., Kalmus, V. and Livingstone, S., 2011)
Cross-national comparison of risks and safety on the internet: Initial analysis from the EU Kids Online survey
of European children (Lobe, B., Livingstone, S., Ólafsson, K. and Vodeb, H., 2011)
Who bullies and who is bullied online? A study of 9-16 year old internet users in 25 European countries (Görzig,
A., 2011)
Comparing children’s online opportunities and risks across Europe: Cross-national comparisons for EU Kids
Online (2nd edn) (Hasebrink, U., Livingstone, S., Haddon, L. and Ólafsson, K., 2009)
What do we know about children’s use of online technologies? A report on data availability and research gaps
in Europe (2nd edn) (Staksrud, E., Livingstone, S., Haddon, L. and Ólafsson, K., 2009)
Best practice research guide: How to research children and online technologies in comparative perspective
(Lobe, B., Livingstone, S., Ólafsson, K. and Simões, J.A., 2008)
EU Kids Online II: Enhancing Knowledge Regarding European Children’s Use, Risk and Safety Online
This project has been funded by the EC Safer Internet Programme from 2009-11 (contract SIP-KEP-321803). Its aim is to
enhance knowledge of European children’s and parents’ experiences and practices regarding risky and safer use of the
internet and new online technologies in order to inform the promotion among national and international stakeholders of a
safer online environment for children.
Adopting an approach that is child-centred, comparative, critical and contextual, EU Kids Online II has designed and
conducted a major quantitative survey of 9-16 year olds experiences of online use, risk and safety in 25 European countries.
The findings will be systematically compared to the perceptions and practices of their parents, and they will be disseminated
through a series of reports and presentations during 2010-12.
For more information, and to receive project updates, visit www.eukidsonline.net
 
3
CONTENTS
Contents  3 
1. Introduction  5 
1.1.  THE EU KIDS ONLINE PROJECT................................... 5 
1.2.  THE RESEARCH CONTEXT........................................... 5 
1.3.  THE AIM OF EU KIDS ONLINE II ................................. 6 
1.4.  THE SURVEY AT A GLANCE ......................................... 7 
1.5.  FIELDWORK AGENCY ................................................ 8 
1.6.  MAIN LIMITATIONS.................................................. 9 
1.7.  ACCURACY OF THE FINDINGS ..................................... 9 
2. Survey development and piloting  11 
2.1.  QUESTIONNAIRE DEVELOPMENT............................... 11 
2.2.  COGNITIVE TESTING............................................... 11 
2.3.  SURVEY PILOT....................................................... 12 
2.4.  THE INTERVIEWS ................................................... 13 
2.5.  TRANSLATION....................................................... 14 
3. Sampling  15 
3.1.  SELECTION OF SAMPLE POINTS AND ADDRESSES........... 15 
3.2.  RANDOM WALK METHOD........................................ 18 
3.3.  OTHER METHODS USED .......................................... 18 
3.4.  RESPONDENT SELECTION ........................................ 18 
3.5.  CONTACT SHEETS AND THE SCREENING PROCESSES....... 18 
4. Fieldwork  21 
4.1.  FIELDWORK OVERVIEW........................................... 21 
4.2.  USE OF INCENTIVES ............................................... 21 
4.3.  INTERVIEWERS...................................................... 22 
4.4.  SURVEY MODE AND INTERVIEW LENGTH..................... 23 
4.5.  SUPPORT FOR RESPONDENTS ................................... 24 
4.6.  CONTEXT EFFECTS AND CHILD COMPREHENSION .......... 24 
4.7.  ETHICS AND CHILD PROTECTION ............................... 24 
4.8.  FIELDWORK OUTCOMES AND RESPONSE RATES ............ 26 
5. Data entry and quality control  29 
5.1.  DATA ENTRY AND PROCESSING................................. 29 
5.2.  QUALITY CONTROL ................................................ 29 
5.3.  DATA EDITING ...................................................... 30 
6. Data weighting and design effects  33 
6.1.  THE USE OF WEIGHTS............................................. 33 
6.2.  APPROACHES TO WEIGHTING................................... 33 
6.3.  STAGES OF WEIGHTING .......................................... 34 
6.4.  SAMPLING TOLERANCES ......................................... 36 
6.5.  DESIGN EFFECTS ................................................... 36 
6.6.  ANALYSING DATA ON THE COUNTRY LEVEL ................. 39 
7. The data set  41 
7.1.  THE DATA SET ...................................................... 41 
7.2.  SES MEASUREMENTS............................................. 42 
7.3.  EDUCATION ......................................................... 42 
7.4.  ROUTING AND HANDLING OF MISSING VALUES ............ 45 
7.5.  TREATMENT OF MISSING VALUES BY EU KIDS ONLINE .. 46 
List of figures  49 
List of tables  49 
Annex 1: EU Kids Online  51 
Annex 2: The network  52 
Annex 3: Ethics review  54 
Annex 4: Translation of difficult words  77 
Technical Report and User Guide: The 2010 EU Kids Online Survey
4
 
5
1. INTRODUCTION
1.1. The EU Kids Online project
The EU Kids Online project was organised as a direct
follow-up from the previous EU Kids Online I project which
was carried out in the years 2006 to 2009
1
. That project
examined research carried out in 21 European countries
into how people, especially children and young people,
use new media. In this three-year collaboration,
researchers across a diverse range of countries worked
together, through meetings, networking and dissemination
activities, to identify, compare and evaluate the available
evidence.
Key questions included:
 What research exists, is ongoing or, crucially, is still
needed?
 What risks exist, for which technologies, and in
relation to which (sub)populations?
 How do social, cultural and regulatory influences
affect the incidence and experience of, and the
responses to, different risks?
 Further, in accounting for current and ongoing
research, and anticipating future research, what
factors shape the research capability of European
research institutions and networks?
The aim was to identify comparable research findings
across member states on the basis of which
recommendations for child safety, media literacy and
awareness could be formulated. The project members
invited communications from the wider community,
practitioners and researchers in order to achieve this goal.
1.2. The research context
The rapidity with which children and young people are
gaining access to online, convergent, mobile and
networked media is unprecedented in the history of
technological innovation. Parents, teachers and children
are acquiring, learning how to use and finding a purpose
1
See Livingstone, S., & Haddon, L. (2009) EU Kids Online: Final
Report. LSE, London: EU Kids Online.
http://eprints.lse.ac.uk/24372/
for the internet within their daily lives. Stakeholders –
governments, schools, industry, child welfare
organisations and families – seek to maximise online
opportunities while minimising the risk of harm associated
with internet use.
Diverse and ambitious efforts are underway in many
countries to promote digital technologies in schools, egovernance
initiatives, digital participation and digital
literacy. As many families are discovering, the benefits are
considerable. New opportunities for learning, participation,
creativity and communication are being explored by
children, parents, schools, and public and private sector
organisations.
The previous EU Kids Online research identified a
complex array of online opportunities and risks associated
with children’s internet use.
2
Interestingly, the risks of
concern to children often are not those that lead to adult
anxiety.
3
Also, it appears that the more children go online
to gain the benefits, the more they may encounter risks,
accidentally or deliberately.
4
Risks may arise when children are sophisticated,
confident or experimental internet users, as observed in
‘high use, high risk’ countries or when, as in ‘new use,
new risk’ countries, children gain internet access in
advance of an infrastructure of awareness-raising,
parental understanding, regulation and safety protection.
So, although the popular fear that the internet endangers
all children has not been supported by evidence, there are
grounds for concern and intervention.
Further, despite the popular rhetoric of ‘digital natives’,
many children still lack resources to use the internet
sufficiently to explore its opportunities or to develop vital
2
See Livingstone, S., & Haddon, L. (2009) EU Kids Online: Final
Report. LSE, London: EU Kids Online.
http://eprints.lse.ac.uk/24372/ See also Livingstone, S., &
Haddon, L. (2009a). Kids online: Opportunities and risks for
children. Bristol: The Policy Press.
3
Optem (2007) Safer Internet for Children: Qualitative Study in
29 European Countries. Luxembourg: EC.
4
Livingstone, S. & Helsper, E. (2010) Balancing opportunities
and risks in teenagers’ use of the internet. New Media & Society,
12(2): 309-329.
Technical Report and User Guide: The 2010 EU Kids Online Survey
6
digital literacy skills.5
Thus it is important to encourage
and facilitate children’s confident and flexible internet use.
A difficult balancing act faces stakeholders: promoting
online opportunities without careful attention to safety may
also promote online risk, but measures to reduce risk may
have the unintended consequence of reducing
opportunities.
6
1.3. The aim of EU Kids Online II
A major conclusion in the EU Kids Online I project was
that a robust, comparable and up to date portrait of online
risks encountered by European children was lacking. The
available evidence base regarding users and their needs
clearly had many serious gaps; the methods used in the
existing research were often non-comparable across
projects or countries; also the available research in this
field dates quickly, given the pace of both technological
and social change. To rectify this lack would clearly
require a substantial investment, both in terms of funding
– given the scale, sensitivity and quality of the evidence
required, and in terms of collaborative effort among
experts in each country – given the task of interpreting
and exploiting the evidence produced.
The project aims were framed in accordance with Action
3.2 (Strengthening the knowledge base) of the 2008 Safer
Internet plus programme, namely To enhance the
knowledge base regarding children’s and parents’
experiences and practices regarding risky and safer use
of the internet and new online technologies in Europe, in
order to inform the promotion of a safer online
environment for children.
Enhancing the knowledge base is here understood as (i)
producing new, relevant, robust and comparable findings
regarding the incidence of online risk among European
children; (ii) pinpointing which children are particularly at
risk and why, by examining vulnerability factors (at both
individual and country levels); and (iii) examining the
operation and effectiveness of parental regulation and
awareness strategies, and children’s own coping
responses to risk, including their media literacy.
5
Helsper, E., & Eynon, R. (2010) Digital natives: where is the
evidence? British Educational Research Journal, 36(3), 502-520.
6
Livingstone, S. (2009) Children and the Internet: Great
Expectations, Challenging Realities. Cambridge: Polity.
Building on existing knowledge and experience, this aim
was operationalized in the EU Kids Online project as
specific objectives:
 To design a thorough and robust survey instrument
appropriate for identifying the nature of children’s
online access, use, risk, coping and safety
awareness.
 To design a thorough and robust survey instrument
appropriate for identifying the nature of parental
experiences, practices and concerns regarding their
children’s internet use.
 To administer the survey in a reliable and ethicallysensitive
manner to national samples of internet
users aged 9-16, and their parents, in member states.
 To analyse the results systematically so as to identify
both core findings and more complex patterns among
findings on a national and comparative basis.
 To disseminate the findings in a timely manner to a
wide range of relevant stakeholders nationally, across
Europe, and internationally.
 To identify and disseminate key recommendations
relevant to the development of safety awareness
initiatives in Europe.
 To identify any remaining knowledge gaps and
methodological lessons learned, to inform future
projects regarding the promotion of safer use of the
internet and new online technologies.
 To benefit from, sustain the visibility of, and further
enhance the knowledge generated by, the EU Kids
Online network.
In brief the main aims of the EU Kids Online project was
thus to enhance knowledge of European children’s and
parents’ experiences and practices regarding risky and
safer use of the internet and new online technologies, and
thereby to inform the promotion of a safer online
environment for children.
It has generated a substantial body of new data –
rigorously collected and cross-nationally-comparable – on
European children’s access, use, opportunities, risks and
safety practices regarding the internet and online
technologies. Significantly, findings come from interviews
conducted directly with children from 25 countries across
Europe (Figure 1).
 
7
Figure 1: Countries surveyed by EU Kids Online
Throughout this report and in various network outputs the
countries participating in the survey are referred to by a
two letter country code (see Table 1). These are the same
as used by Eurostat and almost the same as the ISO
3166-1 two letter code. The only difference between the
ISO two letter code and the two letter codes used by
Eurostat is that the United Kingdom is referred to as UK
(rather than GB) and Greece is referred to as EL (rather
than GR).
Table 1: Countries and two letter country codes
Country
Country
code
Country
Country
code
Austria AT Ireland IE
Belgium BE Italy IT
Bulgaria BG Lithuania LT
Cyprus CY Netherlands NL
Czech Republic CZ Norway NO
Germany DE Poland PL
Denmark DK Portugal PT
Estonia EE Romania RO
Greece EL Sweden SE
Spain ES Slovenia SI
Finland FI Turkey TR
France FR United Kingdom UK
Hungary HU
1.4. The survey at a glance
A total of 25,142 children who use the internet were
interviewed, as was one of their parents, during
Spring/Summer 2010, across 25 European countries.
Full details of the project’s methods are provided in the
accompanying Annexes (which are online at
www.eukidsonline.net).
Key features include:
 Two rounds of cognitive testing, in addition to piloting,
to check thoroughly children’s understandings of and
reactions to the questions.
 Random stratified survey sampling of some 1000
children (9-16 years old) per country who use the
internet.
 Survey administration at home, face-to-face, with a
self-completion section for sensitive questions.
 A detailed survey that questions children themselves,
to gain a direct account of their online experiences.
 Equivalent questions asked of each type of risk to
compare across risks.
 Matched questions to compare online with offline
risks, to put online risks in proportion.
 Matched comparison questions to the parent most
involved in the child’s internet use.
 Measures of mediating factors – psychological
vulnerability, social support and safety practices.
 Follow up questions to pursue how children respond
to or cope with online risk.
 The inclusion of the experiences of young children
aged 9-10, who are often excluded from surveys.
The design is comparative in several ways, comparing:
 Children’s experiences of the internet across
locations and devices.
 Similarities and differences by children’s age, gender
and SES.
 A range of risks experienced by children online.
 Children’s perception of the subjective harm
associated with these risks.
 Children’s roles as ‘victim’ and ‘perpetrator’ of risks.
 Accounts of risks and safety practices reported by
children and their parents.
 Data across countries for analysis of national
similarities and differences.
The population interviewed in the EU Kids Online survey
is children aged 9-16 years old who use the internet at all.
Technical Report and User Guide: The 2010 EU Kids Online Survey
8
Note that, in countries where nearly all children use the
internet, internet-using children are almost the same as
the population of children aged 9-16 years in those
countries. But in countries where some children still do not
have access, or for whatever reason do not use the
internet, internet-using-children (the population sampled
for this project) is not the same as all children.
In section 6.2 there is an estimate of the proportion of
internet-using children out of all children in each country.
It is particularly important to keep this in mind when
interpreting cross-country differences.
Additionally, to pinpoint the support children can call on at
home, the EU Kids Online survey interviewed the parent
‘most involved in the child’s internet use’, while also
recording the existence of other adults in the household.
The term ‘parent’ is used to refer to the parent or carer
most involved in the child’s internet use. This was more
often mothers/female carers (some three in four) than
fathers (in a quarter of cases).
1.5. Fieldwork agency
Following a public procurement procedure conducted in
accordance with EC guidelines, Ipsos MORI was
commissioned to work with EU Kids Online (coordinated
by LSE – the London School of Economics and Political
Science) to provide support with questionnaire design and
testing, and to conduct the fieldwork and produce the data
sets. Ipsos MORI, in turn, contracted with fieldwork
agencies in each country (see Table 2), in order to ensure
a standard approach across Europe.
In each of 24 European countries, around 1,000 children
aged 9-16 who use the internet were interviewed, as was
one of their parents. (In the 25
th
country, Cyprus, it proved
problematic to achieve this sample size and so 800
children were interviewed.) Households were selected
using random sampling methods and interviews were
carried out face-to-face in homes using CAPI (Computer
Administered Personal Interviewing) or PAPI (Paper
Administered Personal Interviewing).
The LSE Research Ethics Committee approved the
methodology and appropriate protocols were put in place
to ensure that the rights and wellbeing of children and
families were protected during the research process. At
the end of the interview, children and families were
provided with a leaflet providing tips on internet safety and
details of relevant help lines.
The EU Kids Online network worked closely with
Ipsos MORI at both national and pan-European levels
to ensure the quality of the research.
 The EU Kids Online network is entirely responsible
for the survey questionnaire design, the sampling
decisions, and all data analysis.
 The network worked with Ipsos MORI on finalising
and implementing the survey questionnaire, cognitive
and pilot testing, translation, fieldwork procedures
and implementation, and data editing.
Table 2: List of fieldwork agencies
Country
AT SPECTRA
BE IPSOS BELGIUM
BG MARKET TEST
CY CYPRONETWORK
CZ IPSOS TAMBOR CZ
DE IPSOS GmbH
DK DMA/RESEARCH A/S
EE TURU UURINGUTE A.S
EL OPINION S.A.
ES IPSOS SPAIN
FI TALOUSTOUKIMUS OY
FR
ALTERNATIVE INTERNATIONAL RESEARCH
(OBJECTIF MARKETING)
HU IPSOS SZONDA
IE IPSOS MORI
IT IPSOS ITALY
LT RAIT
NL IBT
NO IPSOS NORWAY
PL IPSOS POLAND
PT IPSOS PORTUGAL
RO MERCURY RESEARCH
SE IMRI
SI IPSOS PULS SLOVENIA
TR IPSOS KMG
UK ROSSLYN RESEARCH
 
9
1.6. Main limitations
Every effort was made in designing and administering the
survey to provide the best account possible of children’s
internet use in Europe. Also the data set containing the
responses has been thoroughly checked for consistency.
Inevitably, however, the project has limitations, and these
should be borne in mind when using the data set and
interpreting the results.
 Limits on sampling – despite repeated return visits to
sampled households and every effort made to
encourage participation, it must be acknowledged
that the recruitment process may not have reached
the most vulnerable or marginalised children.
 Questionnaire limits – the questionnaire was
designed to take, on average, 30 minutes for children
to complete (and 10 minutes for parents), although in
practice, it took rather longer than this (just under one
hour for the child and parent interviews combined). It
is difficult to hold children’s attention for longer than
this, and so difficult decisions had to be taken about
which questions to include or exclude.
 In over half the countries, the self-completion section
of the questionnaire was completed by pen and paper
– this limited the degree of routing (i.e. the degree to
which questions could follow up on children’s
answers). Last, for ethical reasons (as confirmed by
cognitive testing and pilot interviews), intimate,
embarrassing or certain explicit questions could not
be asked.
 Survey context – every effort was made to encourage
honest answers, to promise anonymity and privacy
(including reassuring children that their parents would
not see their answers). However, any survey takes
place within some social context. Here, the fact that it
was conducted in homes with parents in the vicinity
may have influenced the answers of some children,
meaning they gave more ‘socially desirable’ answers.
As detailed in the online technical report, in two thirds
of cases, interviewers reported that parents were
wholly uninvolved in the child’s interview; in a fifth of
cases they were ‘not very much’ involved, and in one
in seven cases they were more involved.
1.7. Accuracy of the findings
To judge the accuracy of numbers in studies like the one
carried out in the EU Kids Online project it is first
necessary to distinguish between two types of error:
random error and systematic error (or bias). All numbers
calculated from the EU Kids Online data set are to some
extent affected by these and are thus essentially
estimates of some true (but unknown) values.
Systematic error (or bias) occurs when the estimates
provided in the study are systematically higher or lower
than the true value. This can for example be the result of
sampling procedures or measurements (e.g. question
wording). The EU Kids Online survey was carefully
designed to avoid such error. The cognitive testing of the
survey instruments is an example of efforts taken to
minimise systematic bias.
Random error is the result of the fact that not all children
in all of the 25 countries have been interviewed. The
results obtained from the samples of approximately one
thousand children in each country will invariably depart
slightly from the findings that would have been obtained
had it been possible to interview all children in these
countries. In most cases this difference is small and gets
smaller the more children there are in the sample. At the
same time however, the smaller the group that is being
analysed, the greater the random error. Another property
of the random error is that very small (or very large)
percentages (such as when a small number of children
have experienced a particular risk) are more accurate
than percentages that are closer to 50%.
The figure below shows how the random error behaves
for three typical kinds of groups in the EU Kids Online
study. The lowest line shows approximately how the
margin of error varies for estimates based on the whole
data set (all children in all countries). The middle line
shows how the margin of error varies for estimates based
on data from all children in a single country. The top line
shows how the margin of error varies for analysis based
on small groups (for example just children that have
experienced a certain kind of risk and been bothered). In
general it is not advisable to conduct analysis of children
who have experienced a risk and been bothered within a
single country by using simple cross tabulation as the
base number will become very low.
Technical Report and User Guide: The 2010 EU Kids Online Survey
10
Figure 2: Estimated margin of error for findings based
on the EU Kids Online data set
0
1
2
3
4
5
6
7
8
9
10
5 15 25 35 45 55 65 75 85 95
Percentage (finding in data)
Marginoferror(+/-percentagepoints)
Whole group Single country
Small group
To illustrate how this works it is possible to look at the
number of children who have seen sexual images on any
websites which is estimated at 14% (as estimated by
using the weighted data set). This estimate is based on
answers from over 23 thousand respondents and thus has
a very small margin of error (only around ± 0.4 percentage
points). In Turkey approximately the same number of
children (13%) say that they have seen sexual images on
any websites but as this estimate is based on answers
from about one thousand respondents in Turkey the
margin of error becomes larger (around ± 2.4 percentage
points). The margin of error is then lower for Germany
(5% ± 1.6 percentage points) but higher for Estonia (30%
± 3.4 percentage points) where the same number of
respondents has participated in the survey in each
country but where the lower figure (5%) has a lower
margin of error than the higher figure (30%).
These examples show that that when working with the
overall findings from all children in all countries or for all
children within each country the random error is in most
cases very small. For analysis of some parts of the data
set, however, the groups that are being examined can get
quite small. For the findings that are presented in the
report due care has been taken not to exceed the
analytical possibilities of the data but readers of the report
should also take care not to over generalise from any
findings based on small subsets of the data. This applies
for example about those children that have experienced
particular risk factors (such as the 14% who have seen
sexual images on any websites) and then go on and
answer questions about that experience.
 
11
2. SURVEY DEVELOPMENT
AND PILOTING
The questionnaires used in the survey were developed by
EU Kids Online network in collaboration with the fieldwork
agency Ipsos MORI. They were then tested and refined
through a two-phase process of cognitive interviewing and
pilot testing.
 Phase one cognitive testing involved 20 cognitive
interviews (14 with children and six with parents) in
England using an English language questionnaire.
Several refinements were then made to the
questionnaires.
 The amended master questionnaires were then
translated and cognitively tested via a total of 113
interviews across the remaining 24 countries (at least
4 in each country), to ensure testing in all main
languages. Again, amendments to the questionnaires
were made for the final versions.
 Prior to main-stage fieldwork, a pilot survey was
conducted to test all aspects of the survey including
sampling, recruitment and the interview process. A
total of 102 pilot interviews (43 with children aged 9
and 10 years and 59 with children aged 11 to 16
years) were carried out across five countries:
Germany, Slovenia, Ireland, Portugal and the UK.
2.1. Questionnaire development
In terms of the scope and topics the questionnaire was
based on previous work carried out in the EU Kids Online
network
7
. This involved amongst other things a
comprehensive review of existing research on children’s
internet use in Europe both in terms of findings and the
questionnaires used.
An initial draft of the questionnaire was made by the LSE,
as project coordinator, in close conjunction with the EU
Kids Online network in the autumn of 2009. This
development stage took the research design from a
7
See Livingstone, S., & Haddon, L. (2009) EU Kids Online: Final
Report. LSE, London: EU Kids Online.
http://eprints.lse.ac.uk/24372/
scoping of the theoretical framework and pressing
research and policy issues, through to a draft
questionnaire to children and to parents that
encompassed the key issues to be addressed, and
seeking to optimise question formats and response
options so as to be readily comprehensible by children.
Following this early development work, the fieldwork
agency (Ipsos) was involved in numerous revisions of the
draft questionnaires, making recommendations with
regards to ensuring question wordings conformed to best
practice for generating accurate and meaningful answers
from respondents, and in particular making
recommendations for the approach to child question
elements.
2.2. Cognitive testing
Cognitive testing is a diagnostic technique that explores
the processes employed by people when they answer
survey questions, such as comprehension, recognition,
recall and decision-making/response (e.g. how do they
respond to being asked potentially sensitive questions
and/or how suitable are the pre-code lists for capturing all
types of valid response)
8
.
By exploring in a qualitative way the processes by which
people interpret and respond to questions, we can identify
potential sources of measurement error and ideally
address them via appropriate revisions to the
questionnaire to ensure it measures what we want it to
measure as accurately as possible. This can be
particularly helpful for surveys among children, given the
difference in cognitive ability between adult researchers
who are designing the questionnaire, and the child
informants completing them. In the context of international
surveys, cognitive testing can help to ensure that the
8
R. Groves, F. Fowler Jr, M. Couper, J. Lepkowski, E. Singer
and R. Tourangeau, Survey Methodology, (2004), p. 202.
Technical Report and User Guide: The 2010 EU Kids Online Survey
12
wording of questions and response options generate and
capture the same meaning across all countries.
Findings can also be useful when interpreting findings in
the sense that they provide extensive qualitative data on
the types of aspects respondents are thinking about when
they give particular answers to particular questions.
Two rounds of cognitive testing were conducted for this
study by Ipsos and local fieldwork agencies. The first
stage involved 20 cognitive interviews (14 with children
and six with parents) in England. Four of the children were
aged 9-10, four aged 11- 12, five were aged 13-14 and
one was aged 15-16. There were eight girls and six boys.
In terms of social economic status, three parents were
from social groups ABC1 (households where the chief
income earner is in a professional, managerial or clerical
position) and three were from social groups C2DE
(households where the chief income earner is a skilled
manual worker, semi-skilled or unskilled or not working).
This stage of testing tested all key aspects of the main
questionnaire, including respondent comprehension, the
layout of the self completion module, and the acceptability
and suitability of approaches for sensitive subject matter.
A significant amount of refinement was implemented
following this wave. Many changes were made in order to
increase clarity and comprehension and ensure consistent
and unambiguous interpretation. For example, further
clarification was given regarding specific timeframes to
think about when asking children about frequency of
internet based activities; more specific definitions and
supporting examples were given to describe generic
internet terms and concepts, such as social networking.
Some changes were also made to increase ease of
completion of the self-completion elements, such as
reducing complexity of routing, and making instructions for
navigation more prominent through the use of colour for
younger children. The questionnaire was then translated
into all languages relevant to the 25 country study
The second stage involved cognitive interviews (113 in
total) in the remaining 24 countries, to ensure testing
across different languages and cultural contexts. Four or
more interviews were conducted with children in each
country, and a small number of parent interviews were
also conducted. Whilst a range of age groups were
included, 9-10 year olds were over-sampled to ensure that
the questionnaire was sufficiently tested among the age
group likely to have most difficulties with completing it.
This stage of testing was designed to assess the
suitability and efficacy of questioning approaches used
and comparability of meaning generated from the
translated questionnaires across countries, languages
and cultures. It also tested the effectiveness of the
questionnaire following amendments made after stage
one testing.
The testing identified a range of country specific
translation issues, which were then addressed. It also
highlighted differing issues in different countries relating to
the sensitivity of some questions, and concerns about the
length and complexity for younger age groups. As a
result, the length of the questionnaire and level of filtering
was reduced for all children, and some further sensitive
items cut out for 9-10 year olds, especially detailed
questions relating to online content of a sexual or violent
nature.
A particular challenge emerged for generating comparable
meanings across countries for questions measuring
negative emotional impact of risk exposure on children. A
challenge lay in identifying a wording that generated
meaning of the same level of harm in each country. The
wording finalised for use in the survey focused on whether
the children were ‘bothered’ by an experience, together
with related words like ‘upset’, ‘worried’ or ‘uncomfortable.’
However, users of the data set should note that there
remain some differences in interpretation across
countries.
2.3. Survey pilot
Before the main fieldwork, a dress rehearsal pilot survey
was conducted to test key aspects of implementation, in
as close to “live conditions” as possible. A total of 102
pilot interviews were carried out across five countries:
Germany, Slovenia, Ireland, Portugal and the UK (43 with
children aged 9-10 and 59 with children aged 11-16).
The pilot study checked the efficacy of random walk
sampling procedures, contact and screening procedures,
fieldwork materials, and all protocols for how to
communicate about the survey, gain informed respondent
consent and respondent co-operation. It also tested the
length and effectiveness of the survey tools themselves in
“live” conditions.
As a result of the pilot, some final minor modifications
were made to the questionnaire, mainly to reduce length.
 
13
Refinements were also made to the screening contact
sheets to make them more user-friendly for interviewers,
taking into account the large quantity of addresses that
needed to be screened to identify eligible households.
The pilot also identified challenges relating to respondent
engagement in communicating the survey and parental
concern about the sensitivity of the subject matter. The
guidance already provided to interviewers on how to
handle this during fieldwork was therefore expanded on
for the main stage, taking into account learning from the
pilot.
2.4. The interviews
The questionnaires for the children consisted of three
main components which were administered in a
sequence. The children were interviewed face to face to
obtain responses to questions in most sections of the
questionnaire, and then were given the most sensitive
questions in a questionnaire form for them to complete on
their own. For each child, one parent/carer was
administered a questionnaire with a selection of questions
that matched to the questions in the child survey. The
sections in these three questionnaires are outlined below
Items with matched child-parent questions are marked
with an asterisk. An additional screening questionnaire
was used to obtain socio-demographic information about
the household and its internet use.
1. Interviewer administered (face-to-face) the child
questionnaire, covering:
- Patterns of child’s internet usage *
- Activities online
- Digital skills
- Perceptions of parent’s/carer’s, teachers’
and friends’ mediation of online risks *.
2. Child questionnaire for self-completion (simple
version for 9–10 year olds, more complex version for
11–16 year olds ), covering:
- Psychological factors
- Risky offline activities
- Experience of online risks *
- Coping with online risks
- Sources of education, advice and support.
3. Interviewer administered parent questionnaire,
covering:
- Additional and repeated household demographics
and internet access
- Parental patterns of internet usage *
- Perceptions of the child’s internet usage and
exposure to online risks *
- Parental mediation of the child’s online risks *
- Sources of parental education, advice and support.
The “contact sheets” used by interviewers to introduce the
survey, screen for eligible households, and gain informed
respondent consent to the study was also designed to
collect a small amount of demographic information about
screened households where possible (i.e. before
respondent refusal, for example).
The survey was carried out face to face in home, rather
than by telephone, for example, due to the sensitivity of the
subject matter and the need to gain rapport with families to
engage them in the survey work. Questionnaires were
administered either using Computer Assisted Personal
Interviewing (CAPI) or on paper (PAPI), depending on
local practice in each country (see Table 7). Furthermore,
whilst the first two survey tools were administered by
interviewers face to face with the respondent, a self
completion mode was used among children to help ensure
confidentiality of responses to sensitive questions, and to
minimise the potential of social desirability bias – e.g.
under-reporting of exposure to online risks – that might be
caused by the presence of the interviewer or other
household members.
Children were carefully briefed by interviewers about how to
complete the self-completion questionnaire, and were also
provided with clear written instructions about how to do so.
All children were given an envelope in which to place their
completed forms, to help reassure them about the
confidentiality of their responses. Two versions of the self
completion tool were developed, one for 9-10 year olds and
one for 11-16 year olds.
The version for 9-10 year olds excluded some questions
relating to sex and violence related to online risks that were
thought to be less appropriate for this age group. To keep
the length to an acceptable minimum for this age group,
some of the follow-up questions relating to the detail of
specific risks experienced were also omitted and asked
only of 11-16 year olds. This version was also divided into
five separate documents so that the interviewer could
provide more guidance at each step of the way about how
each one should be completed. For this age group, text that
gave instructions about routing through the questionnaire
was also shown in red font to help ensure that it was not
missed.
Technical Report and User Guide: The 2010 EU Kids Online Survey
14
2.5. Translation
A master questionnaire was finalised in English. National
versions were then produced in appropriate languages (see
Table 3). After the master questionnaire was finalised and
approved the translation process progressed as follows:
(i) The master questionnaire was sent to the national
agencies using a specific format designed for
multilingual questionnaires. It was easy to understand
as the source language and the target language
could be simultaneously viewed.
(ii) In the national agencies, two researchers that had at
least two years of experience of opinion surveys
independently translated the questionnaire into their
mother tongue. After this, they met to compile the two
translations into one which was then sent to the Ipsos
coordination centre.
(iii) The core team in the coordination centre verified that
everything had been translated, after which the
questionnaires were sent to back-translation. A native
English speaker with a sufficient level of the source
language then translated it back to English.
(iv) The back-translated documents were returned to the
coordination centre where the team checked them
against the original English master. Each country was
given feedback based on this exercise and all
necessary adjustments were made to the final
questionnaire by the national agencies.
(v) The national agencies sent the final national
questionnaires to the coordination centre.
Academic representatives in every country in the EU Kids
Online network also reviewed translations to double check
that the meaning of key terms was as intended. In
particular, a list of concepts for which there were
challenges ensuring translation generated identical
meaning across countries was drawn up (“upset” is one
example) and network members input to ensure the most
comparable terminologies were used (see Annex 4).
Network members also helped to provide nationally
relevant examples to support communication of key
concepts, such as social networking.
Table 3: Languages provided in the EU Kids Online
survey in each of the participating countries
Country Language(s)
AT Austria German
BE Belgium Dutch, French
BG Bulgaria Bulgarian
CY Cyprus Greek
CZ Czech Republic Czech
DE Germany German
DK Denmark Danish
EE Estonia Estonian, Russian
EL Greece Greek
ES Spain
Spanish (Castilian),
Catalan
FI Finland Finnish
FR France French
HU Hungary Hungarian
IE Ireland English
IT Italy Italian
LT Lithuania Lithuanian, Russian
NL Netherlands Dutch
NO Norway Norwegian
PL Poland Polish
PT Portugal Portuguese
RO Romania Romanian
SE Sweden Swedish
SI Slovenia Slovene
TR Turkey Turkish, Kurdish
UK United Kingdom English
 
15
3. SAMPLING
A representative sample of ~1,000 internet using children
aged 9–16 and one of their parents or carers, from each
of the 25 European countries, was selected. The overall
sample size was 25,142. A three-stage (sampling points,
addresses, and individuals), random probability clustered
sample was achieved. Details of the sampling process are
outlined below. The sampling for the project followed a
robust approach, for example, reflecting processes and
standards common for many large scale Europe-wide
surveys conducted by and on behalf of the European
Commission.
3.1. Selection of sample points
and addresses
An official and complete register of geographical units was
used as the sampling frame for each country. However, in
some countries, certain areas were excluded from the
sampling frame for reasons of practicality, reflecting
standard approaches to fieldwork in the country
concerned. These regions included Mount Athos in
Greece, The Wadden Eilanden in the Netherlands,
Madeira and Azores Islands in Portugal, Ceuta and Melilla
in Spain and The Channel Islands, Isle of Man, and the
area north of the Caledonian Canal in the UK. In all
countries where small geographical areas have been
excluded, population coverage is still extremely high (e.g.
over 95%) meaning negligible impact on survey
estimates. The approach taken reflects standard
approaches to survey work in each country in this regard.
Prior to selection of sampling points, the list of
geographical units was stratified (ordered) by:
(i) Region (NUTS
9
2, 3 or 4, or other nationally
appropriate system of regional classification)
(ii) Population density or degree of urbanisation, where
data was available.
9
Nomenclature of Territorial Units for Statistics, see see
Eurostat, 2010 Eurostat (2010) ‘Introduction’, in NUTS –
Nomenclature of territorial units for statistics,
http://epp.eurostat.ec.europa.eu/portal/page/portal/nuts_nomencl
ature/introduction
Table 4 below outlines the method of stratification (region
and degree of urbanisation) used in each country for both.
In all countries sampling points were then selected with
Probability Proportionate to Size (PPS). This means that
the chance of selection is equivalent to the number of
children living there. For example, if the total population of
children aged 9-16 is 2 million, the probability of selecting
an area with 50,000 children is 0.025 and the probability of
selecting an area with 10,000 children is 0.005. The
number of sampling points varied by country, according to
local circumstances (see Table 4).
All addresses were selected using random probability
sampling approaches, but the precise approach varied by
country reflecting different circumstances on the ground,
the nature of sample frames available, and cultural
differences with regards to whether initial contact was
thought to be most appropriate by telephone or face to
face, bearing in mind the sensitive subject matter. In most
cases “random walk” sampling and face to face recruitment
was used. In a small number of countries, households were
selected from national population registers (either
households in general, or households with children) and
pre-selected addresses were visited in person, or contacted
by telephone in the first instance.
Table 5 below shows the number of sampling points
selected in each country, along with the address-selection
method used. More detailed information about the different
methods then follows.
It should be noted that the relatively low number of
sampling points in Norway does not indicate a lower quality
of the sample in Norway. Typically, a larger number of
sampling points is preferred since they reduce the risk of
homogenous responses within clusters which has the
potential to reduce a survey’s effective sample size, (the
extent to which there are systematic differences in findings
between survey clusters). However, the lower number of
sample points in Norway has not caused a problem in this
regard: despite the relatively small number of sampling
points, the effective sample size for Norway is estimated at
729 which is in line with other countries (see Table 12).
This means that the smaller number of sample points used
Technical Report and User Guide: The 2010 EU Kids Online Survey
16
in Norway did not have a larger negative impact on the
reliability of Norway’s findings.
Table 4: Method of stratification by region and urbanisation
Type of Primary Sampling Unit
Indicator for stratification by
region
Indicator for stratification by degree of
urbanisation
AT Locality (village/town/city) NUTS 2
Total number of inhabitants living in locality and
number of children aged 9-16 living in locality
BE Locality (village/town/city) NUTS 2
Total number of inhabitants living in locality and
number of children aged 9-16 living in locality
BG Locality (village/town/city) NUTS 2
Total number of inhabitants living in locality and
number of children aged 9-16 living in locality
CY Municipalities
by district (Nicosia, Limassol,
Larnaca, Pafos, Famagusta)
Municipalities defined as Urban/Rural by the
Department of town Planning and Housing in Cyprus.
CZ
Municipalities and postal districts for the
cities with over 50 thousands inhabitants.
NUTS 3
Total number of inhabitants in municipalities or postal
districts
DE Postal district ADM sampling points
ADM (Arbeitskreis deutscher Marktforscher) sample
points have urban/rural indicators
DK Postal district NUTS 2 Number of children aged 9-16 living in locality
EE Locality (village/town/city) NUTS 3 Number of children aged 9-16 living in locality
EL Administrative district NUTS 2 Total number of inhabitants living in locality
ES Administrative area NUTS 2 Number of children aged 10-15 living in locality
FI Postal district NUTS 2 Total number of inhabitants living in locality
FR Locality (village/town/city) UDA 5 (regions) Population Density
HU
Locality (village/ town/ city/districts of the
capital)
NUTS 2
Total number of inhabitants living in locality and
number of children aged 9-16 living in locality
IE Electoral district NUTS 2 Total number of inhabitants living in locality
IT Locality (village/town/city) NUTS 2 Total number of inhabitants living in locality
LT Locality (village/town/city) Counties Population density
NL
Locality (village/town/city) and postal for
larger cities
NUTS 1 Number of addresses per km2
NO Locality (village/town/city) NUTS 2 Total number of inhabitants living in locality
PL Administrative areas - Gminas NUTS 2
Population density and number of children aged 9-16
living in locality
PT Locality (village/town/city) NUTS 2 Total population resident in the locality
RO Locality (village/town/city) NUTS 2 Total number of inhabitants living in locality
SE Administrative area NUTS 2 Number of children aged 10-15 living in locality
SI
Administrative areas defined by
Slovenian statistical office
NUTS 3
City size (number of inhabitants) and percentage of
agricultural population
TR Administrative district NUTS 1 Total number of inhabitants living in locality
UK NUTS 4 NUTS 1 Number of children aged 9-16 living in locality
 
17
Table 5: Sampling information
Methodology Type of national register used
Sampling
points
AT Random Walk 125
BE Random Walk 102
BG Random Walk 290
CY Random Walk 84
CZ
Pre-selected households - telephone
recruitment
Registered directory of fixed line telephones. Held by Nexos. 140
DE Random Walk 212
DK
Pre-selected households of children
aged 0-17, telephone recruitment
Sample was purchased from ”Forbrugerliv” a company owned by
Jyllands-Posten Holding AS (the largest media-provider of
Denmark)
148
EE Random Walk 137
EL Random Walk 125
ES Random Walk 140
FI Random Walk 100
FR Random Walk 120
HU
Pre-selected households with children
aged 9-16
Addresses were selected from the Citizens’ Personal Data and
Address Register, held by The Central Office for Administrative
and Electronic Public Services (Hungary).
163
IE Random Walk 170
IT Random Walk 103
LT Random Walk 101
NL
Pre-selected households - telephone
recruitment
Addresses were selected from the Nationale Telefoongids,
published by KPN Telecom.
125
NO
Pre-selected households - telephone
recruitment
Addresses were purchased from “Norstat” using the
“EasyConnect” database – the largest database of private
households and telephone numbers in Norway
16
PL
Pre-selected households of
children aged 9-16
PESEL - Universal Electronic System for Registration of the
Population . Addresses were selected by the Ministry of Internal
Affairs and Administration
218
PT Random Walk 128
RO Random Walk 135
SE
Pre-selected households with children
aged 9-16 - telephone recruitment
Addresses were selected from a random sample of households
with children aged 9-16. The sample was provided by PAR
(Postens Adressregister, the postal office address register, which
itself is drawn from SPAR, the Swedish Population register.
40
SI
10% Random Walk – 90% national
register of households with 9-16s
Central Population Register 350
TR Random Walk 115
UK Random Walk 179
Technical Report and User Guide: The 2010 EU Kids Online Survey
18
3.2. Random walk method
In each of the selected sampling points, one address was
drawn at random from the register of households or from
the listing of streets in the geographical area of the
sampling point. This ‘seed’ address is the first in the
sample and acts as the start point for the random walk.
The remaining addresses in the sample point were
selected using a strict pre-defined random-walk procedure
which makes the selection independent of the
interviewer’s decision. Specifically, the interviewer
selected a batch of five addresses before counting five on
their route and then selecting another batch of five. The
procedure is as follows:
 Standing at the seed address, the interviewer faced
the street and turns left. He/she identifies the next
four immediately neighbouring addresses as the next
in the sample – a batch of five addresses has been
selected together.
 The interviewer then continued along the route
counting houses/flats/apartments, leaving five
addresses before identifying the next five
neighbouring addresses as the next in the sample.
 When turning at the end of the street, the interviewer
did not stop counting housing units/addresses.
Every effort was made to screen each sampled address
and achieve an interview at eligible households, with the
following fieldwork requirements followed:
 At least 4 attempts to make contact at each address.
 Contact attempted at different times of day (including
evenings), and at weekends as well as weekdays.
 No substitution of selected addresses – this means
that if an address is unproductive or appears
unsuitable from the outside, the interviewer still had
to make contact there; they could not choose a
neighbour to try instead.
3.3. Other methods used
In-home recruitment from national registers: In
Hungary, Poland and Slovenia10
a sample of households
with children aged 9-16 were drawn from population
10
In Slovenia, the survey began with a random walk
methodology, but the approach was switched to this method
early in on in fieldwork due to difficulties identifying eligible
households using random walk methods.
sample frames as the sample to be issued. In each of these
countries, all selected addresses were sent a copy of the
letter in advance. This served to notify them of the survey
and inform them to expect an interviewer to call. The
selected household was then visited by an interviewer and
screened on the doorstep following exactly the same
contact procedure as in countries using Random Walk (and
discussed further below).
Telephone recruitment from national registers: Sweden
used a register which identified households (in the selected
sampling points) with children aged 9-16 and Denmark
used a register that identified households with 0-17s. Czech
Republic, Norway and the Netherlands used national
registers of households in general. In each case,
households (in the selected sampling points) were
randomly selected from the register for contact and
screening. In all four of these countries, the pre-selected
households were initially contacted and screened by
telephone with an interviewer then visiting responding
households to conduct the interviews in person after
appointments had been made. In the Netherlands, in cases
where an appointment with a respondent was broken and
could not be rescheduled, the interviewer had the option of
sampling/screening new households using strict random
walk methods.
3.4. Respondent selection
Each selected household was screened to identified eligible
households (with a child aged 9-16 who uses the internet).
An interview with one child and one parent/carer was
required. Where there was more than one eligible child
present, one child per household was selected using the
last birthday method.
The parent/carer interview was conducted with the
parent/carer who knew the most about the child and their
internet use. In around three-quarters of households, the
mother was interviewed, around one-fifth, the father, and in
around one-in-twenty households another household
member (step parent, grandparent, or other) was
interviewed.
3.5. Contact sheets and the
screening processes
Two types of contact sheet were provided to interviewers,
guiding them through the screening process, and on which
 
19
key screening and sample outcome information was
recorded.
First, a summary contact sheet was used to check if the
property was residential/occupied and if so, to identify if a
child aged 9-16 was present in the household.
Secondly, at households with a 9-16 year old present, fuller
screening processes were carried out using a more detailed
follow-up contact sheet:
 completing eligibility screening (identifying children
using the internet)
 identifying and selecting the appropriate child and
parent/carer respondent
 securing co-operation and informed consent from
parents and children (see section below: 6. Ethics
and child protection)
 capturing some profile information about all
households with children that could be used for
profiling and weighting purposes: age, gender and
internet use of all children in the household, and
education and employment status of the chief income
earner in the household.
In order to support communication of the survey
requirements and gain respondent co-operation, a letter
from the LSE was shown to the respondents, emphasising
the importance and value of the study. A copy of the
English version of the letter is provided in annex 3. In
countries using face to face recruitment from pre-selected
addresses, the letter was posted in advance.
Technical Report and User Guide: The 2010 EU Kids Online Survey
20
 
21
4. FIELDWORK
4.1. Fieldwork overview
Fieldwork started in April 2010 and was completed by
October 2010 (week 26); however, more than half of the
countries completed by early July (week 11). Fieldwork
was shortest in Romania and Hungary (6 weeks) and
longest in Norway (23 weeks).
4.2. Use of incentives
The decision whether or not to use incentives was taken
at the local agency level. Using their experience of
conducting in home surveys with parents and children
within their market, agencies considered whether they
thought the offer of incentives would increase response
rates enough to offer value for money. In some cases,
incentives were introduced part way through fieldwork to
help improve response rates. Incentives were offered in
the following countries:
 Austria: A 5 EUR Amazon voucher given to the child
upon completion
 Belgium: A 5 EUR voucher for the child, conditional
on taking part.
 Bulgaria: Stationary for the child (coloured pencils,
ruler, pocket books worth approximately €1.5).
 Czech Republic: Incentives given to the parent: a
gift bought by the interviewer – most often some kind
of premium coffee, chocolate or tea costing on
average 4 EUR. The children were given a flash disk
costing 8 EUR. Both conditional on participation
 Denmark: Each responding household received an
incentive of 100 DKR. Normally the child was offered
the incentive. Each respondent could choose
between a gift-card and donating the amount to a
Child Welfare Organisation. 42% of respondents
chose charity donation.
 Finland: A small chocolate or candy bar was
provided to the child as a gift after the interview was
completed (worth approximately €2).
 Netherlands: The original incentive was a lottery with
prizes as follows (or cash equivalent); 5x weekend in
a bungalow park (worth approximately €400 each); 5x
game consoles (worth approximately €250 each); 10x
Nintendo DS (worth approximately €200 each); 1x
weekend EuroDisney (family max. 4 persons €450
per person); To boost response rates part way
through fieldwork, a conditional incentive of 10 EUR
was given
 Norway: Every family received 300 NOK
 Poland: Chocolate was given to one of the parents
conditional on participation (worth approximately €3)
 Romania: A key holder or a pocket calculator for the
child on completion (worth approximately €3)
 Spain: An incentive of 6 EUR (gift card) was given to
parents as a gift for the children. The incentive was
provided upon completing the interview
 Sweden: A gift voucher of SEK 100 (ca €10), signed
for by the parent but aimed at the child; this incentive
was later increased to two cinema tickets (value ca
€18).
 Turkey: A notebook and a pen were given to the
child upon completion (worth approximately 2TL)
 UK: £10 per household upon completion of the
survey
Incentives were higher in those countries where fieldwork
took longer; alternatively, when fieldwork seemed to be
progressing slowly, the level of incentives was raised. It
appeared that incentives were lower when there were
more sampling points. This finding might be a
methodological artefact due to both sample points and
incentives being related to the number of interviewers. An
unexpected finding was that interviews took longer when
incentives were higher. In addition, in those countries
where addresses were pre-selected, (higher) incentives
were more likely because interview times were longer. To
explain these findings, three regression analyses were
conducted11
with response rates, incentives, and fieldwork
length as the dependent variables and all other sampling
and fieldwork variables as well as country size area and
number of children as predictors. None of the predictors
reached statistical significance suggesting that the
11
Source: Görzig, A. (in press) Methodological framework: the
EU Kids Online project. In Livingstone, S., Haddon, L., and
Görzig, A. (Eds.) Children, Risk and Safety on the Internet: Kids
online in comparative perspective. Bristol: The Policy Press.
Technical Report and User Guide: The 2010 EU Kids Online Survey
22
relations become meaningless when other variables are
held constant.
Table 6: Fieldwork dates, incentives and number of
interviewers
Start End Incentives
used
Number of
interviews
AT 24.04 25.07 Yes 45
BE 06.05 14.07 Yes 44
BG 06.05 24.06 Yes 133
CY 17.05 20.09 No 39
CZ 21.05 02.07 Yes 146
DE 20.05 07.07 No 400
DK 30.04 14.06 Yes 160
EE 10.05 14.07 No 70
EL 10.05 02.07 No 52
ES 10.05 15.07 Yes 60
FI 28.04 02.07 Yes 54
FR 06.05 03.07 No 83
HU 10.05 15.06 Yes 123
IE 05.05 24.07 No 103
IT 28.04 03.07 No 56
LT 23.04 06.07 No 52
NL 03.05 05.08 Yes 100
NO 21.05 19.10 Yes 90
PL 06.05 26.07 Yes 149
PT 29.04 30.07 No 47
RO 16.05 25.06 Yes 67
SE 27.05 20.09 Yes 64
SI 03.05 27.08 No 200
TR 03.05 17.06 Yes 27
UK 01.05 21.06 Yes 105
4.3. Interviewers
All countries recruited interviewers based on their
experience, not just in research, but more specifically with
face-to-face surveys and random walk procedures where
appropriate, and experience of research with children.
Agencies acknowledged the complexity and sensitive
nature of the questionnaires and allocated the individuals
they thought would achieve the best results. As detailed in
Table 4, the number of interviewers working on the project
ranged from 27 in Turkey, to 400 in Germany.
All interviewers received intensive project-specific training
and briefings and written guidance materials, covering all
aspects of survey implementation, including guidance on
how to conduct sensitive interviews with children.
All project managers and interviewers were supplied with
detailed and uniform instructions supplied by the Ipsos
coordination centre. These Training Booklets and
Interviewer Packs covered the following topics:
 Overall briefing on EU Kids Online Survey:
 Detailed description of the sampling procedures and
random walk methodology where applicable
 Full questionnaire review, clarifying terminology and
data collection
 Review of ESOMAR ethical rules and other ethical
issues and protocols associated with this project,
including relating to child protection, and informed
respondent consent
 Briefings on key techniques and protocols for
interviewing children and parents
 Fieldwork management rules
 Specific techniques to convert refusals and maximise
the response rate
 A reminder of how the quality of their work will be
supervised and managed, including back-checking
procedures.
Interactive telephone briefings with the project managers
from each country were led by the Ipsos Coordination
centre during early April 2010. Further to discussing the
information detailed in the Training Booklets above,
briefings also gave guidance on data processing and how
project managers should deliver local interviewer
briefings. Finally, country specific interviewer briefings
were then conducted locally. These half-day or one-day
sessions are organised centrally or at regional level and
often included role plays where interviewers worked in
pairs to practice delivering the questionnaire.
 
23
4.4. Survey mode and interview
length
Questionnaires were administered either using Computer
Assisted Personal Interviewing (CAPI) or on paper (PAPI).
As mentioned earlier, some sections were intervieweradministered,
whilst sensitive questions among children
were administered via a self interviewing in a selfcompletion
questionnaire.
The interview length was measured per household,
encompassing the length of time it took to complete the
parent, child face-to-face and child self-completion
questionnaires. The average across all countries was 55.8
minutes.
Table 7 gives an overview of the survey mode for each
country, and summarises the range in interview duration
across the countries and provides a comparison between
households where a child aged 9-10 was interviewed and
those where a child aged 11-16 was interviewed. The
interview duration covers the period of time taken to
complete the questionnaire tools, not the full time spent in
the household.
Table 7: Survey mode and interview length
Average interview time for child and
parent combined
Survey
mode
All
With 9-10
year olds
With 11-16
year olds
AT PAPI 59.4 61.8 58.6
BE PAPI 53.3 51.9 53.8
BG PAPI 56.2 56.2 56.2
CY PAPI 42.4 40.6 42.7
CZ PAPI 58.0 59.5 57.5
DE CAPI 49.0 47.7 49.4
DK CAPI 63.8 62.1 64.4
EE CAPI 68.1 69.9 67.6
EL PAPI 52.9 54.3 52.2
ES CAPI 56.3 51.7 57.7
FI CAPI 54.6 50.8 55.8
FR PAPI 47.3 58.5 56.7
HU PAPI 63.6 64.5 63.4
IE CAPI 53.5 52.1 53.9
IT CAPI 53.3 53.5 53.2
LT PAPI 56.9 56.8 57.0
NL PAPI 65.6 66.8 65.2
NO CAPI 66.4 67.4 66.1
PL PAPI 57.8 60.6 57.0
PT PAPI 49.8 51.0 49.3
RO PAPI 53.5 52.1 53.9
SE CAPI 61.2 59.7 61.8
SI CAPI 48.4 45.2 49.3
TR CAPI 55.3 54.9 55.5
UK PAPI 48.6 48.8 48.5
All 55.8 55.9 56.3
Technical Report and User Guide: The 2010 EU Kids Online Survey
24
4.5. Support for respondents
It was important to ensure that where possible, children
and parents were not excluded from the research due to
language or communication difficulties. In cases where
child or parent did not speak the main language(s) of the
country well enough to complete the survey, another
household member was asked to provide support. If a
child had communication difficulties, where appropriate,
the parent or interviewer provided support. However, for
the self-completion element of the study, interviewers
were instructed to ensure that support was kept to a
minimum, to avoid biasing the findings. Types of support
received by respondents were recorded by interviewers,
and this information is included in the data set (see
section 5.4 below).
4.6. Context effects and child
comprehension
As part of the survey’s quality procedures, interviewers
were asked to record details relating to the child’s
comprehension of survey questions and who was present
in the room during the child’s interview. The detail below
comments on the overall average and maximum and
minimum findings across all countries; further detail, by
country, can be found in the data set (QC343-QC348). It
should be noted that the figures outlined below are based
on all unweighted data.
Interviewers were asked to observe how well they thought
the child understood the questions asked during the
interview. Overall, more than nine in ten children were
thought to have understood the interview questions very
or fairly well (93%), rising to as much as 98% in Greece
and Italy. Comprehension was less proficient in Belgium
and Turkey where 13% of children were thought to
understand questions not very well/not at all well.
In total, one in ten children had some form of help
(language or communication) from a family member in
order to answer the survey questions (10%). Overall, two
per cent of adults and three percent of children required
language help to take part in the survey; five per cent of
children required some form of communication help.
Showing the importance of the self-completion sections of
the questionnaire, more than three in five child interviews
were conducted with the parent respondent present in the
room (63%); a further three per cent had another adult
present other than the parent respondent. The proportion
of households where the parent respondent was present
ranged from 29% in the Czech Republic to 80% in Spain
and Romania, and 83% in Turkey.
As well as noting adult presence during the child survey,
interviewers were also asked to observe the extent to
which the parent respondent tried to involve themselves in
the child interview (for example, if they were concerned
about the sensitivity of some of the subject matter). In the
vast majority of cases this was not an issue: overall, twothirds
of parents made no attempt to be involved (66%),
with a further fifth having made little attempt (21% not very
much); equating to 87% of parents overall. In contrast,
four per cent of parents attempted to be involved a great
deal with a further one in ten a fair amount (10%). Parents
in Spain were the most fervent, with around three in ten
attempting to be involved a great deal/a fair amount
(29%). Interviewers were fully briefed on how to manage
these types of situation, for example, explaining the
importance of confidentiality, reassuring that the child
could skip any question they did not like, and allowing the
parent to see a blank copy of the questionnaire before the
child interview took place.
4.7. Ethics and child protection
Children’s exposure to risks on the internet is a particularly
sensitive topic; it was therefore paramount that fieldwork
was conducted in an appropriately ethical manner. The
project received ethical clearance from LSE’s Research
Ethics Committee and all aspects of methodology and
approaches to survey implementation were developed with
child and respondent wellbeing in mind (See: Research
Ethics review questionnaire in Annex 3). Key points are
described below.
An essential requirement was to gain informed consent from
both the parent and the child. Several, several mechanisms
were put in place to ensure that parents and children had all
the information necessary to make an informed judgement
about taking part in the survey.
 Each house was presented with written information
about the study, as well as interviewers explaining
this carefully to parents and children verbally. The
letter contained both LSE and Ipsos branding and
was translated into the relevant local languages and
was available online on the EU Kids Online website.
The key points covered including the funding and
purposes of the project, the nature of the interview,
the value of the project to policy makers seeking to
 
25
improve internet safety for children, and contact
details for the national fieldwork organisation
(contracted by Ipsos), the national EU Kids Online
network representative, and the project director
(Sonia Livingstone for EU Kids Online at LSE).
Where a parent wished for more time to consider
taking part, the information letter was left with the
household for several days before the interviewer
returned at a later date.
 A signature was required from parents confirming
consent to their own interview and consent to us
approaching the child to invite their participation in
the child interview in all countries except from
Germany, where local laws prohibited written
signatures being obtained and where instead
interviewers were asked to sign to confirm that the
parent had given their permission for the interview to
take place. Child consent was also recorded by the
interviewer signing in writing that this had been given
verbally by the child.
 Particular attention was taken to ensure that the text
and words spoken in the letter and consent form were
age appropriate. Across all languages, separate
versions of the text were tailored for parents and
children of different ages (A copy of the information
letter, safety tips leaflet and consent form can be
found in Annex 3).
 Anonymity and confidentiality of responses were
guaranteed to both parents and children, with the
exception that if the child reported that they are being
harmed in some way, this would limit the promise of
confidentiality and action would be taken (see below).
All fieldwork was conducted in line with stipulated ESOMAR
ethical guidelines for conducting research with children and
young people, as well as those specified by the LSE
Research Ethics Committee.
Interviewers were selected on their experience of working
with children and further training and briefing was provided
as outlined above in section 4.3. Relevant security checks
were carried out on interviewers where appropriate
according to country specific legal requirements.
Confidentiality and anonymity was guaranteed to survey
questions but at the same time interviewers were instructed
to ensure that parents remained in the vicinity within the
household whilst the children interview was being conducted
(with the door open, for example).
Whilst in the field, all children were advised of the fact that it
was their right to stop the interview at any point and that
they could choose not to answer a question if they felt
uncomfortable doing so.
In designing the questionnaire, several measures were also
put in place to make the child as comfortable as possible.
 The most sensitive questions relating to risky
behaviour were asked in a self completion format
where children were assured that neither the
interviewer nor the parent would be able to see their
answers, since (for CAPI) the screen was turned so
only they could see it or (for PAPI) a pen-and-paper
questionnaire was provided for their answers along
with a sealed envelope for the child to use.
 Discretion was used to consider whether questions
were suitable for the youngest participants, the most
sensitive and more mature themed questions were
only asked to those aged 11 years and above.
 A Prefer not to say option was also included in those
questions where a child might feel uncomfortable
about disclosing their behaviour.
 The routing and introduction to questions ensured
that the interview does not introduce the child for the
first time to ideas or material that may be ethically
problematic. For example, children were immediately
routed out of sections about risky behaviour if it
became apparent that they had not experienced the
risk, and introductory wording was used where
appropriate to forewarn of the nature of the
subsequent questions.
All respondents, parents and children, were provided with an
information leaflet at the end of the survey visit, containing
tips and advice about online risk and safety. The leaflet was
tailored for each country and included the contact details of
local help lines (or other appropriate provision for children
identified through the conduct of the survey as in some way
‘at risk’), whereby the respondent can access private,
confidential help and advice. These leaflets were developed
for the project by the national Insafe nodes of the EC’s Safer
Internet Programme, with input also from Child Helpline
International (see www.childhelplineinternational.org).
Given the topics considered in this project, it was
important to establish an agreed approach to intervention
prior to fieldwork, as to what would happen if it became
apparent that a child was at risk of harm. This approach
was agreed between Ipsos and the LSE and cleared by
the LSE Research Ethics Committee.
To ensure guarantees of confidentiality and anonymity,
intervention from fieldworkers was only considered on the
basis of relatively serious harm being identified, i.e. on the
Technical Report and User Guide: The 2010 EU Kids Online Survey
26
broad principle that the risk identified was “something any
reasonable person could not ignore”. The notes below
outline the agreed approach of dealing with identified risk,
although it is important to note that a different approach
was considered depending on whether or not the risk was
identified within the survey questions.
 The questionnaire design and methodology meant
that risk of current harm would not identifiable from
the study at the time of the interview. First, survey
questions ask about exposure to risks in the past and
do not directly identify current issues; secondly
questions on risk were asked within self completion
modules and as such interviewers were not aware of
the child’s responses. We therefore took a universal
approach to responding to possible risk for all
children.
 Interviewers explained to all children that if they have
they have experienced harm, they should tell a
trusted adult;
 As mentioned above, the interviewer left a leaflet with
helpline numbers and ‘top-tips’ for online safety.
 In addition, fieldwork agencies abided by any local
laws regarding actions required to protect children.
A protocol was in place for actions to be taken if a
participant made a disclosure to the interviewer outside
their response to a survey question and/or the interviewer
witnessed something in the household suggesting that a
child was at risk.
 If the interviewer became aware of risk of harm to a
child that no reasonable person could ignore, or that
required action within national laws, they were to
follow specific agreed protocols as below.
 Given that disclosure of harm in this scenario is
outside the main interview questions, this approach
does not conflict with guarantees of respondent
confidentiality with regards to survey responses.
 The interviewer was instructed to report the “incident”
to the project manager/field supervisor for action to
be taken by the Institute, according to national law.
Where institutes are not competent to make a
decision of this kind, a legal person was to be
consulted before action is decided upon.
 In such cases, the interviewer was also instructed to
tell the child that they are concerned and talk to them
about the action that they will be taking.
 As mentioned above, the interviewer was also briefed
to encourage the child to talk to a trusted adult (if they
have not already done so) and provide them with the
leaflet of top tips/help line support services.
Importantly, and reassuringly, there were no such incidents
reported during fieldwork.
Finally, confidentiality and anonymity was guaranteed during
the data processing stage of the project by removing key
identifiers from the data set.
4.8. Fieldwork outcomes and
response rates
The interviewers needed to complete the following steps
to achieve an interview:
 Make contact at the selected address (up to four
attempts)
 Obtain consent for the screening questionnaire and
establish whether at least one child aged 9–16 years
old lived at the address and was using the internet
 Obtain consent for the child and parent/carer
interviews.
Contact, cooperation and response rates were calculated
in accordance with standard definitions12
. It was estimated
that in 53% of interviewers’ attempts to contact an eligible
address (i.e., a residential address with at least one child
age 9–16 that uses the internet), this was successful
(contact rate). Contact rates ranged from 31% in Germany
to 89% in Romania. In 79% of the estimated eligible
cases, when contact was made, the interviews were
completed (cooperation rate), with a rate of 36% in the
Netherlands to 100% in Poland
13
and Greece. The
estimated overall response rate was 42% of all potentially
eligible cases (regardless of successful contact).
Response rates ranged from 17% in the Netherlands to
83% in Romania (see Table 8).
12
American Association for Public Opinion Research (AAPOR)
(2008) Standard definitions: Final dispositions of case codes and
outcome rates for surveys (5th edn)
13
In Poland households were preselected using the ‘Universal
Electronic System for Registration of the Population’, which
perhaps explains the high cooperation rate.
 
27
Figure 3: Field work steps and respondent mortality
N = 25.142, 42% of all
estimated eligible
cases that were
contacted
Respondents complete
interview
Estimated overall
response rate
Child 9-16 that used the
internet in household
Interview refused or not
completed
Residents screened
No child 9-16 that used
the internet in household
Contact made at address
Residents refused
screening
N =384.856 residential
properties visited
(N = 60.232 are
estimated to be eligible)
No contact made at
address
Reason for drop out: Non-contact Refusal Ineligible Refusal
Source: Görzig, A. (in press) Methodological framework: the EU Kids Online project. In Livingstone, S., Haddon, L., and
Görzig, A. (Eds.) Children, Risk and Safety on the Internet: Kids online in comparative perspective. Bristol: The Policy
Press.
Differences in response rates may be related to
differences in sampling methodology and unexplained or
unmeasured cultural differences. In Sweden, for example,
respondents were pre-selected and recruited via the
phone, possibly explaining the high contact rate in that
country (80%). However, the same methodology was
used in Norway, and this had one of the lowest contact
rates (34%). The low cooperation rate in Cyprus might be
due to a lack of respondent incentives, but, on the other
hand, the average incentive of €10.50 per respondent,
among the highest in the sample, resulted in a low rate of
cooperation in the Netherlands; a finding that is in line
with past research
1415
. Note that incentives were offered
in 13 countries, in the course of all or part of the fieldwork.
The monetary value of these incentives ranged from an
average of €1 (Turkey) to €38 (Norway) per household,
with a range of €3 to €12 in those countries within the two
centre quartiles (middle 50%). Methodological issues
mostly explain cross-country differences in response
14
De Heer, W. (1999) ‘International response trends, results of
an international survey’, Journal of Official Statistics, vol 15, no 2,
pp 129–42
15
De Leeuw, E., and de Heer, W. (2002) ‘Trends in household
survey non-response: a longitudinal and international
comparison’, in R.M. Groves, D.A. Dillman, J.L. Eltinge and
R.J.A. Little (eds) Survey nonresponse, New York: Wiley, pp 41–
54
rates, but not in all cases, which suggests unmeasured
cultural differences played a role.
Table 8: Contact, cooperation and response rates, by
country
Contact rate
(%)
Cooperation
rate
(%)
Response
rate
(%)
AT 54 91 49
BE 54 98 53
BG 85 75 64
CY 36 69 25
CZ 38 70 27
DE 31 100 30
DK 66 48 32
EE 88 89 78
EL 74 100 74
ES 64 85 54
FI 79 86 68
FR 45 90 41
HU 62 100 61
IE 39 65 25
IT 53 77 40
LT 79 100 79
NL 48 36 17
NO 34 61 21
PL 38 100 38
PT 78 97 76
RO 89 93 83
SE 80 89 70
SI 33 88 29
TR 62 99 61
UK 71 92 66
ALL 53 79 42
Technical Report and User Guide: The 2010 EU Kids Online Survey
28
 
29
5. DATA ENTRY AND
QUALITY CONTROL
5.1. Data entry and processing
As noted above in section 4.4 some countries
administered surveys using CAPI, others used PAPI.
CAPI captures respondents’ answers electronically during
fieldwork, so no data entry is required. For countries using
PAPI, the data from paper questionnaires were either
scanned or the data were entered by local data
processing teams. Industry standard quality control and
back check procedures were carried out to ensure a high
quality of data.
Although all local agencies processed their own data, a
uniform collection of data across all countries was
ensured through the use of a single data map provided
centrally by the core survey team. Raw data sets were
uploaded by agencies to - a centralised online data
processing platform – with each case containing contact
sheet, screening, parent and child questionnaire data for
one household.
To ensure that data were processed correctly, local
agency data sets had to pass a series of basic quality
checks before being accepted by the online platform.
Such checks included considering if responses were valid
and whether ID variables were consistent. A range of
further quality, consistency and edits checks were
considered centrally by the core project team using Initial
data – more detail about the edits applied to the data set
is provided below.
At all times, and in line with data protection legislation and
professional industry standards (ESOMAR), data were
held securely and kept confidential. Furthermore, only
anonymised data were uploaded via the online platform
for anonymised central analysis.
5.2. Quality control
Strict quality measures were implemented at every stage of
the data collection and production process. This tight
monitoring allowed for the early detection of any potential
problems which could be addressed in a timely way, thus
maintaining quality of data throughout.
Checks for all returned materials included:
 Check of returned Summary Contact Sheets: to
ensure that the pre-defined random-walk procedure
was strictly applied and that a summary outcome was
coded for the addresses contacted.
 Check of returned Follow-Up Contact Sheets: to
ensure that the birthday method for random-child
selection was correctly used, to ensure that the
parent and child consent was obtained for all
interviews and that the interviewers had completed
the child and head of household profile information
for all households with a child aged 9-16.
 Check of returned interview packs: to ensure that the
correct survey forms were used and none were
missing.
In a small number of cases in the final data set, a nonselected
child had been interviewed. However, the profile
by age and gender was reviewed and addressed in the
overall approach to non-response weighting.
In total 15% of interviews for each interviewer received a
quality back check, focused on either the contact sheet or
the interview itself: for around five per cent, local
supervisors checked contact sheet processes were
implemented correctly on the ground during fieldwork. For
around 10%, telephone call backs to respondents checked
the following:
 Respondent’s memory of the interview (gender of
interviewer, day, time and duration, mode of
interviewing, use of show cards, topics of the survey)
 Answers to some key questions (mainly screener
questions about the parent and child).
Checks on early completed questionnaires check:
 If filtering and routing was working correctly and was
being respected.
 If questions had been missed out due to interviewer
error.
Technical Report and User Guide: The 2010 EU Kids Online Survey
30
 The general quality of the data.
Each agency completed a quality-check monitoring form
(provided by central project team) early on in fieldwork
confirming that the appropriate checks had been
completed, and any issues rectified.
At the data entry stage, for a proportion of cases in each
country, data entry was back-checked to verify that data
entry was set up according to the data map provided and to
check if responses were captured exactly in the way they
were recorded by interviewers and respondents. Checks
required by local agencies included:
 Ensuring filtering has been set up correctly
 No questions missed
 No responses miss-keyed
 If there were blanks or don’t knows in the
demographic section, the fieldwork department was
encouraged to contact the interviewer or interviewee
in order to complete the missing information.
 If Contact sheet ID numbers were missing, they had
to be identified and entered for 100% of cases
 If there were multiple blanks or don’t knows across
the entire questionnaire and/or sections of the
questionnaire are not filled in or filters/routings are
not respected properly, the questionnaire was not
retained for subsequent processing. A data count
was run checking for instances where more than 30%
of responses to the parent and the child
questionnaire were not valid, and this enabled the
survey team order to consider whether such
instances should be treated as incompletes and
potentially removed from the data set. There were no
cases where both the parent and child interview had
over 30% invalid responses and needed to be
removed.
5.3. Data editing
A wide range of automatic routing and edit checks (i.e.
checks to disallow out of range responses) are built into
CAPI to ensure accuracy of completion.
However, for paper-based surveys this is not possible, and
as for all PAPI studies it was necessary to carry out edit
checks on the data to identify and address errors on a small
proportion of cases for some questions. Inconsistencies are
particularly likely to occur with any self-completion
questionnaire due to the lack of interviewer administration.
Therefore particular attention was paid to the child self
completion questionnaires.
The first step was to investigate any inconsistencies found
with fieldwork agencies to identify possible courses and
solutions – for example, checking for any data entry errors
that could be corrected, or raising issues with interviewers
to establish why issues might have occurred. Where
inconsistencies still remained, data editing was considered,
and applied where logical to support data quality and
consistency. Importantly, edits were also applied in ways
that supported consistency with edit checks and routing
implemented in CAPI. The level of editing required was low
reflecting that children had a good level of understanding of
the questionnaire. The edits applied were as follows.
Routing: A check was carried out to identify instances
where questions with filtered bases routed from responses
to previous questions had been answered by the
respondents whose previous responses indicated eligibility
to proceed. Based on a review of the responses to those
follow-up questions, edits were applied to route
respondents out of later questions where earlier responses
indicated that the questions were not relevant to them. For
example, a review of follow-up responses identified that in
many cases respondents had coded response options such
as “don’t know” or “not very much,” or “not applicable”. This
approach also provided consistency between PAPI and the
routing built into CAPI.
Inappropriate multi-coding: There were some instances
where multiple codes were selected at single code
questions. In these cases it is not possible to know which is
the “correct” answer, so items were coded as “no answer”
for cases where this applied. There were also some
instances of multi-code questions, where a respondent had
chosen one or more answer options – and also a “don’t
know” or “prefer not to say” option. In these cases, based
on a review of the data it seemed appropriate to edit out the
“don’t know/prefer not to say” response, because the main
response codes coded seemed likely to be valid.
Addressing inconsistent responses: A range of
consistency checks were carried out to check responses
that were illogical based on responses to other questions,
or general reasonableness. The table below details the
checks carried out, and any edits which were applied to
address these.
 
31
Table 9: Details of non-routing based edits
Questions Approach and edits applied
Child age:
Checking contact sheet: SCR.3b/4b Age of selected child against
the child age question in the parent questionnaire: Q.201 What is
the age of your child?
The age of the interviewed child in the contact sheet was edited,
where necessary, to ensure it referenced the child who had
completed the questionnaires.
If there was more than one possible match (among the children
recorded in the contact sheet data) then the child that uses the
internet (SCR3D) was identified as the selected child. If both/all
(or neither/none) used the internet then one child was selected at
random. In order to avoid confusion, the contact sheet selected
child age variable was not included in the main survey data set
(just in the contact sheet data set). This ensured that all data
users will use the same variable for analysis on child age (as
recorded during the main interview). All selected children were
then coded as internet users at SCR3D for consistency (as per
the profile of survey participants desired).
Child gender:
Checking contact sheet: SCR.3c/4c Gender of selected child
against the child gender question in the parent questionnaire:
Q.201b Gender1 of child?
As above.
Number of children living in house:
Checking contact sheet: SCR.2 Number of children aged 9-16
living in the household against parent questionnaire variable:
Q202 number of children aged 0-17 living in the household.
If more children were reported at SCR2 than Q202, Q202 was
edited to be equal to the response at SCR2. If there was no
valid response at Q202 and SCR2, answers were backcoded
from SCR3. If there was no data recorded at
SCR2,SCR3 and Q202 responses were edited to refer to 1
child.
Child use of communication media on the internet:
Q324a-f asked children which of a range of activities they had
done in the last year. This was checked against answers at
Q308a-f which asked how often they had done the same activities
in the past month.
a) email usage
b) visited a social networking profile
c) Visited a chat room
d) used instant messaging
e) Played games with other people on the internet
f) Spent time in a virtual world
If a child had coded “no” (not done in the past year) at Q324 for
activities they had reported doing in the past month at Q308,
the response at Q324 was edited to show that they had
participated in it
Technical Report and User Guide: The 2010 EU Kids Online Survey
32
 
33
6. DATA WEIGHTING AND
DESIGN EFFECTS
6.1. The use of weights
The data set has three kinds of weights. The weights are
applied to the data to improve the representativeness of
the achieved sample. There are three forms of weighting
applied to the data set:
 country-specific design weights which adjust for
unequal probabilities of selection; for example, these
correct for the fact that children in households with
two eligible children only had half the chance of
selection as one-child households;
 country-specific non-response weights which correct
for bias caused by varying response rates across
different types of respondent within each country.
These weights correct for differences between the
achieved profile of respondents and the population
profile on key demographic variables – age, gender,
region and education of the chief income earner in
the household;
 a European level weight which adjusts for country
level contribution to the overall results. This weight
corrects for the fact that the same number of
interviews were conducted per country, despite the
fact that the population of (internet-using) 9-16 year
olds in each country is different. This weight adjusts
each country’s contribution to the European-level
results in proportion to the actual population size of
internet-using children.
There are five variables in the SPSS file: Weight,
Weightb, Weightc, Weightd, and Weighte. The EU Kids
Online network generally follows a consistent approach to
weighting: for descriptive statistics weights are applied to
make them representative of the population, for statistical
significance testing weights are not applied to avoid
biased standard errors.
The first main weighting variable (‘Weight' in the
SPSS file) is generally used for all European level
analysis as it incorporates individual respondent weights
as well as the country-level adjustment. As a function of
the survey design – i.e. equal numbers of interviews in all
European countries, irrespective of their population size –
the final European adjustment weights are large for some
countries. For example, respondents in Turkey have large
up-weights because the country has a large population.
This variable was used for overall results, and analysis at
the European level by age, gender, and socioeconomic
status.
The second main weighting variable (‘Weightb’ in the
SPSS file) is generally used for country by country
analysis, and for analysis looking at any single
country. This weight incorporates the individual withincountry
weights which combine any non-response and
design weights that were calculated.
Note that the SPSS file contains three additional
weighting variables (‘Weightc’, ‘Weightd’ and ‘Weighte’).
These should not be used for data analysis. These are
intermediate weights that cover the first stages of the
weighting calculations. ‘Weighte’ includes the Design
Weight, ‘Weightd’ the Non-response 1 weight, and
‘Weightc’ the Non-response 1 and design weights. These
variables are included for users to judge the impact of the
final stage of non-response weighting has had (in some
cases very little).
As a rule of thumb, for descriptive statistics the
variable ‘Weight’ is used for analysis on the whole
data set but ‘Weightb’ is used when analysing data
within each country or comparing two or more
countries with one another.
6.2. Approaches to weighting
Non-response weights were calculated separately for
each country. Most survey designs would require only one
stage of non-response weighting: the achieved sample
would be weighted back to the profile of either the issued
sample or the survey population. However, with this
survey the achieved sample is purposively different from
the issued sample, since the entire issued sample has
been screened to identify a sub-set of households in the
Technical Report and User Guide: The 2010 EU Kids Online Survey
34
population (i.e. those containing at least one child aged 9-
16 who used the internet).
The sample issued in each country was representative of
the country’s population, while the population we
interviewed was children who use the internet. As such,
the issued sample includes households which were
ineligible for the survey: i.e. households which did not
include children, and households which included children
who were not internet users. Given the specific nature of
the population the survey represents – i.e. children aged
9-16 who use the internet – there is no accurate
population data available to use for weighting.
Instead, the non-response weights are based on data
collected during the screening process on contact sheets
and combined with general population data relating to
households with children at national level. This has been
done based on data from two stages of the sampling and
recruitment process. First our screened sample
(consisting of all children in screened households
including both internet users and non-users) was
weighted according to the known population data for all
children aged 9-16 (users + non-users) by age, gender
and region.
Once the first stage of weights had been applied, the noninternet
users were excluded to provide a sample of
internet users that is representative of the population of
internet-using children in terms of age, gender and region.
It is this that was used to weight the interviewed sample
back to. By weighting the screened sample first, we can
be confident that the starting point (the screened sample
of children) is representative of the population in terms of
these variables and therefore when the sample of users is
extracted, we can be confident that the profile used to
weight the interviewed sample is also representative.
6.3. Stages of weighting
The three types of weighting, with non-response weighting
being split into two, meant that weights are calculated and
applied in four stages (see below). However they are
combined to give a single weight for analysis:
Together, the design weights and the two stages of nonresponse
weighting, produce an individual weight for each
respondent. This weight should be applied whenever any
analysis is conducted for a single country (for example,
looking at results and sub-group differences for Denmark).
This weight is labelled ‘Weightb’ in the SPSS file.
The final European adjustment is calculated at the country
level, which means that every respondent in the same
country will be given the same final adjustment factor.
This factor is combined with the individual weight to give a
single weight which should be applied when analysis of
the whole data set is conducted (for example, looking at
results and sub-group differences for Europe). This weight
is labelled ‘weight’ in the SPSS file.
Further information about the construction of the individual
weight is provided below.
Non-response weights 1 – applied to the sample of all
screened children (i.e. this will include not only those who
completed an interview, but those who were eligible but
were not interviewed and those who were ineligible nonusers
of the internet). For each country, population
distributions of the population of children aged 9-16 by
age, gender and region were identified by local agencies.
These are used as targets for rim weighting for each
country. Rim weighting is a process whereby the
population figures are fed into a piece of software which
iteratively runs through different possibilities until it comes
to the best fit weights for the data.
With this approach, rather than interlocking all weighting
variables, each is treated on a marginal basis. For
example breaking the sample down into cells by age
within sex within region is usually impractical due to
limitations on the sample size. All that rim weighting
requires is the distribution for each of these variables. The
computer then calculates the ‘best’ fit for the data across
all the variables included in the weighting. The
advantages to this approach are that the weighting can
include a greater number of variables, and it is not
necessary to have targets for all the interlocked cells. As
such, rim-weighting is the preferred option in most
situations.
Profiles for the rim weights were created for each country
based on age x gender, and region. The regions used
were those corresponding to the region variables in the
final data set, and are typically the regions used at the
sampling stage.
Design weights – applied to the sample of all eligible
children (all children aged 9-16 who use the internet).
Design weights adjust for unequal probabilities of
selection during sampling: at eligible addresses one child
per household was selected for interview from all those
who were eligible. This introduces unequal probabilities of
 
35
selection whereby a child from a household with a number
of eligible children has a lower chance of selection than a
child from a household with only one eligible child. We
applied design weights to correct for these unequal
probabilities of selection.
The weights are calculated as the inverse of the selection
probability, for example where there are 3 eligible children
(aged 9-16 who use the internet) the weight would be
1/(1/3) = 3.
Non-response weights 2 – applied to the final sample of
all interviewed children. The weighted profile (i.e. with
NR1 x DW already applied to the data) of all eligible
children – distributions of children by age, gender, region
and education of the chief income earner in the household
– are used as targets for rim weighting for each country.
The regions used for weighting are the same as those
used for Non-response weights 1, as described above.
These three stages are then combined to produce one
single weight for each respondent. Weights are capped (a
maximum of 6 times the average weight is set) to avoid
any extreme weights which could cause peculiarities in
the data as well as large design effects. The weights were
then rescaled (divided by the average weight for each
country); a purely aesthetic process which means the
weighted base reflects the number of respondents
interviewed. Since these individual weights are calculated
separately for each respondent based on household
make-up and demographic profile of the country, the
range and average weight varies from one country to the
next.
European weights – applied to the full aggregate dataset
(all countries) as the last stage of the weighting process,
in order to adjust the contribution each country makes to
the data at the European level
This is a final weight for European level analysis which
adjusts for country level contribution to the overall results
relative to population size. Respondents in countries with
a large population of child internet users are given a
greater weight than those in countries with a smaller
population which means that the larger countries
contribute more to the total figures than smaller ones.
As there is no available data on the population of children
aged 9-16 who use the internet by country to use for this
stage these figures have been estimated using a
combination of data from a range of sources. For most
countries data from the Eurobarometer and Eurostat has
been used.
Figures for internet penetration are estimated from a
combination of data from the Eurobarometer (% children
using the internet in 2008) and Eurostat (change in
internet penetration, as measured among 16-24s 2008-
2009).
Table 10: Estimated number of children aged 9-16
who use the internet, by country
Children in
population
9-16 years
(N)
Estimated
children
online
(%)
European
internet-using
children per
country
(%)
AT 739,722 86% 1.49%
BE 974,461 78% 1.78%
BG 554,032 91% 1.2%
CY 82,059 68% 0.13%
CZ 809,443 90% 1.71%
DE 6,419,300 86% 12.95%
DK 558,236 97% 1.27%
EE 105,460 96% 0.24%
EL 862,481 59% 1.19%
ES 3,401,338 80% 6.38%
FI 501,387 98% 1.15%
FR 6,005,850 87% 12.26%
HU 854,406 93% 1.86%
IE 458,260 93% 1. 00%
IT 4,516,646 55% 5.83%
LT 320,821 96% 0.72%
NL 1,582,903 96% 3.57%
NO 503,160 98% 1.16%
PL 3,490,271 97% 7.94%
PT 871,444 78% 1.59%
RO 1,821,471 78% 3.33%
SE 861,183 98% 1.98%
SI 154,063 95% 0.34%
TR 10,297,791 65% 15.70%
UK 5,861,598 98% 13.20%
Technical Report and User Guide: The 2010 EU Kids Online Survey
36
Internet penetration for 2010 was estimated by taking the
actual penetration in 2008 and extrapolating the rate of
growth in internet use measured by Eurostat across 2009-
2010. As 2009 data was unavailable for the UK and
Belgium, estimates for UK and Belgium are based on
2008 data, scaled up by the average population change
across the countries where 2009 data are available.
Eurostat gives figures for the changing proportion of 16-24
year olds who have used the internet in the past year, and
those who have ever used the internet. The change in
internet penetration was estimated at being between
these two figures. Where data on the change in internet
penetration among 16-24 year olds was unavailable, the
average rate of change of 2 percentage points was
assumed. Generally figures were rounded up rather than
down, since the change in internet use among 9-16s was
assumed to be higher than among 16-24 year olds. Note
that figures for Norway were unavailable and so were
estimated based on the data for Sweden. Figures for
Turkey were estimated from two local sources: the
Ministry of Social and Family Research, whose data
showed 67.2% children age 13-18 use the Internet, and
results from the ‘ICT Usage in Households, 2004-2010’
from the Turkish Statistical Institute (2010) which showed
62.9% 16-24 had used the internet in the last 3 months.
An average of these two figures was taken and used as
the internet penetration rate for 9-16 year olds.
These figures were used to generate an estimate of the
total number of 9-16 year old internet-users in the
population of each country. These figures were then used
to calculate the proportion of internet users across the 25
countries covered by the survey that fall within each
country. For example, 4% of all internet users across the
countries covered by the survey are in Belgium, and
therefore results from Belgium are weighted down to
account for only 4% of the total 25,000 interviews. The EU
relative weights therefore adjust the data to be
representative of the internet-using 9-16 year old
population of the 25 countries covered by the survey.
6.4. Sampling tolerances
When interpreting the findings it is important to remember
that the results are based on a sample of children aged 9-
16 who use the internet, and not the entire population of
9-16 year olds in each country. Therefore, we cannot be
certain that the figures obtained are exactly those we
would have if the whole population of 9-16 year olds in
each participating jurisdiction had been interviewed (the
‘true’ values).
The “margin of error” is a common summary of sampling
error, which quantifies uncertainty about (or confidence in)
a survey result. Usually, one calculates a 95 percent
confidence interval of the format: survey estimate +/margin
of error.
The margin of error depends on the size of the sample:
the more interviews conducted (sample size), the smaller
the margin of error. It also depends on the study design:
any sample design that departs from a simple random
design, and any weighting applied to the data set normally
results in a “design effect” that reduces the effective
sample size (the size that is effective for statistical
reliability tests), and a higher margin of error.
6.5. Design effects
Design effects are ‘the ratio of the sampling variance for a
static computed using a [particular design] divided by the
sampling variance that would have been obtained from a
[Simple Random Sample] of exactly the same size’
16
. The
design effect statistic can be usefully applied to indicate
the loss of precision in survey results derived using a
particular methodology compared with the reliability of
results derived using a Simple Random Sampling method.
This loss of precision is often indicated by showing how
the margin of error for each survey statistic is widened as
a result of the survey design. Each statistic in a survey
has its own design effect.
Design effects apply to the methodology used for EU Kids
Online in a number of ways:
 Clustering of interviews: because a face-to-face
fieldwork methodology was used, interviews in each
country were clustered in geographical areas (rather
than being spread randomly across the country). This
clustering leads to a loss of precision, insofar as
variance in survey results differs between rather than
across clusters.
 Weighting: as described above, several stages of
weights were applied to adjust country-level
estimates. All weights applied are associated with a
design effect.
16
Groves, R. M. (2004) Survey Methodology. Hoboken, New
Jersey, Wiley.
 
37
In addition, at the European level: disproportionate
stratification of samples: rather than being sampled in
proportion to the population of children within each
country, 1,000 interviews were conducted per country.
This has the advantage of producing reliable estimates
per country. At the aggregate level, however, this design
requires corrective weighting (so that each country’s
results are weighted back to reflect that country’s relative
population size within the 25 participating countries).
These weights are also associated with a design effect.
For example, whilst ca. 1,000 interviews are being
conducted in both Ireland and Germany, in the European
data set as a whole, Ireland cases will be weighted down,
whilst Germany cases will be weighted up, reflecting the
smaller and larger sizes of the eligible population in each,
respectively. As would be the case for any study
generating European estimates, design effects arising
from this are large, due to the considerable variability in
population size between each country. The variables used
to create the design are shown in Table 11.
Table 11: Variables used to calculate design effects
Country Language(s)
QP215: Do you personally use the internet? Yes/No QC301a: Please tell me where you use the internet these days?
Your bedroom (or other private room) at home. Yes/No
QP220a: Which of the following things, if any, do you (or your partner/
other carer) sometimes do with your child? Talk to him/her about what
he/she does on the internet. Yes/No/Don’t know
QC303: How often do you use the internet? Every day or almost
every day/ Once or twice a week/ Once or twice a month/ less
than once a month/ Don’t know
QP220b: Which of the following things, if any, do you (or your partner/
other carer) sometimes do with your child? Sit with him/her while s/he
uses the internet (watching what s/he is doing but not really joining
in). Yes/No/Don’t know
QC110: In the PAST 12 MONTHS, have you seen or experienced
something on the internet that has bothered you in some way? For
example, made you feel uncomfortable, upset, or feel that you
shouldn’t have seen it? Yes/No/Prefer not to say/ Don’t know
QP224a: Do you (or your partner/carer) make use of any of the
following for the computer that your child uses MOST OFTEN at
home? Parental controls or other means of blocking or filtering some
types of website. Yes/No/Don’t know
QC106a: How true is this of you? I am easily distracted and find it
difficult to concentrate. Not true/ A bit true/ Very true
QP224b: Do you (or your partner/carer) make use of any of the
following for the computer that your child uses MOST OFTEN at
home? Parental controls or other means of keeping track of the
websites they visit. Yes/No/Don’t know
QC106b: How true is this of you? Other people my age often treat
me as if I wasn’t there. Not true/ A bit true/ Very true
QP224c: Do you (or your partner/carer) make use of any of the
following for the computer that your child uses MOST OFTEN at
home? A service of contract that limits the time your child spends on
the internet. Yes/No/Don’t know
QC106c: How true is this of you? If I am in trouble I can usually
think of something to do. Not true/ A bit true/ Very true
QP228: As far as you are aware, in the past year, has your child seen
or experienced something on the internet that has bothered them in
some way? For example, made them feel uncomfortable, upset, or
feel they shouldn’t have seen it? Yes/No/Prefer not to say/Don’t know
QC106d: How true is this of you? I take things that are not mine
from school, home or elsewhere. Not true/ A bit true/ Very true
QP235a: Please tell me whether or not your child has done [each of
the following] in the PAST YEAR, as far as your are aware: Gone to a
meeting with someone face to face (in person) that he or she first met
on the internet. Yes/No/Don’t know.
QC106e: How true is this of you? I get on better with adults that
with people my own age. Not true/ A bit true/ Very true
QP235b: Please tell me whether or not your child has done [each of
the following] in the PAST YEAR, as far as your are aware: Seen
images on the internet that are obviously sexual – for example,
showing people naked or people having sex. Yes/No/Don’t know.
QC106f: How true is this of you? I can generally work out how to
handle new situations. Not true/ A bit true/ Very true
QP235f: Please tell me whether or not your child has done [each of
the following] in the PAST YEAR, as far as your are aware: Sent
someone else sexual messages (e.g. words, pictures of videos) on
the internet. By this we mean images of people naked or having sex.
Yes/No/Don’t know.
QC106g: How true is this of you? I have many fears, and I am
easily scared. Not true/ A bit true/ Very true
Technical Report and User Guide: The 2010 EU Kids Online Survey
38
Since every estimate in a survey has a different design
effect, design effects were calculated in STATA on a
range of survey variables. These variables were selected
purposively to cover a range of different types of question,
and therefore to give an indication of the range of design
effects that may apply to different types of question.
These questions were also selected to cover some of the
key measures of interest from the survey (including of
internet use, parental monitoring and knowledge,
exposure to risks online and child self-sufficiency) and to
provide an indication of the psychological profile of
children from different sampling points.
Table 12 below shows the results by country, and for the
European sample as a whole, giving the unweighted
sample size for each country – i.e. the actual number of
interviews conducted – as well as the design effects
Table 12: Design effects and effective sample sizes by country
Actual sample size
Approximate design
effect
Approximate effective
sample size
Approximate effective
sample efficiency
Approximate design
factor
AT 1,000 1.79 591 59% 1.34
BE 1,006 1.68 644 64% 1.30
BG 1,088 1.56 711 65% 1.25
CY 806 1.79 591 73% 1.34
CZ 1,009 1.60 668 66% 1.27
DE 1,023 1.67 626 61% 1.29
DK 1,001 1.45 723 72% 1.20
EE 1,005 1.51 688 68% 1.23
EL 1,000 1.75 616 62% 1.32
ES 1,024 1.69 640 62% 1.30
FI 1,017 1.38 830 82% 1.17
FR 1,000 1.36 744 74% 1.17
HU 1,000 1.57 662 66% 1.25
IE 990 1.31 784 79% 1.14
IT 1,021 2.05 533 52% 1.43
LT 1,004 1.62 651 65% 1.27
NL 1,004 1.79 591 59% 1.34
NO 1,019 1.47 729 72% 1.21
PL 1,034 1.75 634 61% 1.32
PT 1,000 1.63 661 66% 1.27
RO 1,041 1.71 663 64% 1.31
SE 1,000 1.40 771 77% 1.18
SI 1,000 1.51 682 68% 1.23
TR 1,018 2.39 473 46% 1.55
UK 1,032 1.52 694 67% 1.23
 
39
The easiest way to interpret the design effect is with
reference to the effective sample size (calculated as:
actual sample/design effect). The effective sample size
shows the amount of confidence we have in the reliability
of our figures, after adjusting for the impact of the survey
design – for example, although 1,005 children in Estonia
were interviewed, we have as much confidence in the
results as we would have from a simple random sample of
688 children in Estonia.
The European level design effect in particular is inevitably
large with this type of design: equal numbers of interviews
were conducted in all countries, despite the very large
differences in population size, and then large weights
were applied to weight the contribution of each country
appropriately within the aggregate figures. The main
contributing factor to these large design effects is the
large European weights. While the total number of
interviews conducted was over 25,000 therefore, this
equates to an effective sample of 8,509 (i.e. the same
level of reliability applies to our achieved sample of
25,000 using a clustered and disproportionately stratified
design, as to a sample of 8,509 using a simple random
sample).
6.6. Analysing data on the
country level
When analysing the EU Kids Online data set on a country
level and wanting to maintain claims of
representativeness, it is necessary to take care not to
extent beyond the analytical possibilities of the data and
to pay attention to base numbers in the analysis. To take
an example, let us look at the UK data set, which has
1,032 responses. When making inferences about all
children who use the internet this is roughly the base
number that defines the standard error for point estimates
in the data. For percentages the standard error can be
obtained by the following formula:
(100 )
1
P P
SE
n



As can be seen the standard error will be bigger for
numbers close to 50% than for numbers close or 100 or
zero (due to the multiplication of the percentage times 100
minus the percentage). A confidence interval for the
percentage can then be calculated by multiplying the
standard error with the appropriate Z value (usually 1.96
for a 95% confidence interval). To estimate the accuracy
of percentages it is therefore only necessary to know the
percentage itself and the correct base on which that
percentage is calculated. Let us take an example.
In the UK data set there are 93 children who claim to have
seen sexual images on any websites. This is based on the
unweighted data set and to obtain the correct point
estimate it is necessary to apply weighting and deduct
individuals with missing values on this particular variable
but let us for the moment imagine that in the UK sample
93 out of 1,032 respondents have seen sexual images on
any websites or some 9% (let us also ignore the fact that
there is a clustering effect in the data set that reduces the
effective sample size a bit). To estimate the accuracy of
this finding we would calculate a 95% confidence interval
in the following way:
(100 ) 9(100 9)
1.96 1.96 1.75
1 1.032 1
P P
CI
n
 
    
 
Then we conclude that some 9% (±1.75) of UK children
have seen sexual images on any websites. If we want to
compare boys and girls then we must split the group by
gender and then the accuracy of the point estimate for the
boys will be based on the number of boys in the sample
and similarly the accuracy for the point estimate for the
girls will depend on the number of girls in the UK sample.
In the UK sample there are 510 boys and of those some
56 have seen sexual images on any websites or about
11%. As can be seen from the formula that we use to
calculate the standard error the accuracy of the 11%
figure for boys in the UK sample is affected by both the
11% number itself and the n which in this case is 510 (as
there are 510 boys in the sample). The confidence interval
for our estimate of how many UK boys have seen sexual
images on any websites thus becomes:
(100 ) 11(100 11)
1.96 1.96 2.72
1 510 1
boys
P P
CI
n
 
    
 
Note that by going from estimating how many UK children
have seen sexual images on any websites and to
estimating how many UK boys have seen such images
the confidence interval goes from 1.75 to 2.72 and the
difference is almost exclusively the result of going from
the group of all UK children and to the group of UK boys.
But note at the same time that it is not problematic here
Technical Report and User Guide: The 2010 EU Kids Online Survey
40
that only 56 UK boys have seen sexual images on any
websites.
To further demonstrate this, let us look at our estimate for
a very rare activity like sending sexual messages. In the
UK sample only some 22 children admit (or claim) to have
sent such messages or only some 2% of the UK sample
of 1.032 children. As before we can calculate a 95%
confidence interval for our estimate that 2% of UK children
have sent sexual messages:
(100 ) 2(100 2)
1.96 1.96 0.85
1 1.032 1
P P
CI
n
 
    
 
Note that as with sexual images for all children in the UK
sample this estimate is based on 1.032 children but the
confidence interval becomes smaller as there are fewer
children who have sent sexual messages than have seen
sexual images. As mentioned before, this is because
there is less uncertainty for numbers close to zero or
100% than numbers close to 50% and as 2% is a smaller
number than 9% the confidence interval is smaller.
If we wish to see how many UK boys have sent sexual
messages we will see that 12 out of 510 UK boys admit
(or claim) to have done so or roughly 2%. We can
calculate a confidence interval as follows and becomes
larger than the confidence interval because it is based
only on the 510 boys in UK but not the whole UK sample.
(100 ) 2(100 2)
1.96 1.96 1.22
1 510 1
P P
CI
n
 
    
 
So far we have looked at how the confidence intervals
change when moving from the overall data set of c.a.
1,000 respondents and down to the subset of boys only
(or girls only) where one could expect around 500
respondents. It is possible to break the data down even
further and look for example at two age groups by gender
(going down to roughly one fourth of the overall data) or
even further. However, as one goes into smaller sub
groups the standard errors for the point estimates will
grow increasingly large (splitting a group in half will result
in a standard error that is roughly 50% bigger than the
standard error for the overall group).
The small number of children who have experienced most
of the risks asked about in the EU Kids Online survey
becomes a limitation if there is desire to look at only those
who have experienced a certain risk. It might be possible
to look a the group of 93 UK children who have seen
sexual images on any websites and see for example how
many of them have been bothered or upset by it (the base
number for that analysis would be 93 minus perhaps
some small internal mortality if not all of those 93
respondents have responded to the question that is then
being analysed within the group of 93). However it is
questionable if it is possible to look at gender differences
within the group of 93 UK children who have seen sexual
images on any websites as that analysis would be based
on only 56 boys and 37 girls. Also it would be impossible
to look any further at the 22 UK children who claim to
have sent sexual messages.
For analysis of this kind where there is interest in looking
at specific aspects of online experience it is however
possible to use the whole data set with answers from all
countries in a multivariate analysis where country
differences are either controlled for or estimated along
with other independent variables.
 
41
7. THE DATA SET
7.1. The data set
One of the main objectives of the EU Kids Online project
was to make data available – to the EU Kids Online
network and the wider research community. Attention was
paid to ensuring that the variables in the data set were
consistently labelled and coded. The main types of
variables are screening, core and derived variables.
Screening variables contain selected socio-demographic
information about the household and its members; core
variables provide data on the survey questions; and
derived variables are created or computed from the
information derived from the other variables.
All variables were labelled according to a similar structure
containing a prefix, a root and a suffix. Core variables
were named according to which questionnaire and
question they referred to (see Table 13).
Table 13: Variable names of core variables
Core
variables
Prefix
Origin of
variable
Root
Question
number
Suffix
Response
option
Screener form SCR 1 b
Child
interview (f2f) QC 300 d
Child
interview (self
completion)
QC 100 a
Parent
interview
QP 200 c
Derived variables were named according to which
variables they were derived from, which concept they
incorporated and/or what calculation was used to derive
them (see Table 14).
Table 14: Variable names of derived variables
Derived
variables
Prefix
Origin of
variable
Root
Concept /
group
Suffix
Variable
type
Child
interview DC
Parent
interview DP
e.g.:
SES
age
webuse
e.g.:
MN: Mean
NM: Count
2: Number
of
categories
Reversed
items RC or RP
Original
question
number
Response
option
The exact naming, labelling and coding of variables
can be found in the data dictionary (downloadable as
an excel file from the UK Data Archive).
Education and occupation of the household’s main wage
earner were obtained from the screening questionnaire.
Country-specific codes were standardised to obtain
comparable variables across countries. Socio-economic
status indicators were derived based on a combination of
the occupation and education variables (see SES pack).
Socio-economic status is not evenly distributed across
countries; the proportion of respondents with a high socioeconomic
background ranges from 12% in Turkey to 82%
in Norway, for medium socio-economic background the
range is 16% in Norway to 67% in Italy, and low socioeconomic
background ranges from 2% in Norway to 54%
in Portugal and Turkey. Finding related to socio-economic
status could be an indicator of between country
differences and vice versa.
Psychological differences were measured on scales
derived or adapted from existing measures for self-
efficacy
17
, the Strength and Difficulty Questionnaire
(SDQ)
18
, sensation-seeking
19
, and internet addiction
20
.
17
Schwarzer, R. and Jerusalem, M. (1995) ‘Generalized selfefficacy
scale’, in J. Weinman, S. Wright, and M. Johnston (eds),
Technical Report and User Guide: The 2010 EU Kids Online Survey
42
The data set also contains paradata, metadata and
auxiliary data21
. Paradata give information on data
collection processes, in this case variables for interview
mode (CAPI, PAPI), screening outcome, interview
completion, property type, interviewer observations and
identifiers for each respondent, household, sample point
and country. This technical report accompanying the data
set provides information on questionnaire duration times
and incentives per country. Metadata are data on the
data, such as sample design and question coding, which
are contained in the data set variables on sample points,
in the questionnaires and in the interviewer briefing
documents which contain introductory texts, coding
instructions and definitions of complex terms; they are
also provided in this technical report which provides
information on actual numbers of interviewers per country.
In addition, socio-economic status and education packs
(downloadable as excel files from the UK data archive)
provide information on national coding and recoding
procedures concerning educational levels and
occupational status into cross-national variables. Auxiliary
data are data from external sources and include variables
for information such as regions, population density and
area size.
Measures in health psychology: A user’s portfolio. Causal and
control beliefs, Windsor: NFER-Nelson, pp 35–7.
18
Goodman, R. (1997) ‘The strengths and difficulties
questionnaire: a research note’, Journal of Child Psychology and
Psychiatry, vol 38, pp 581–86; Goodman, R.R., Ford, T.T.,
Simmons, H.H., Gatward, R.R. and Meltzer, H.H. (2003) ‘Using
the strengths and difficulties questionnaire (SDQ) to screen for
child psychiatric disorders in a community sample’, International
Review of Psychiatry, vol 15, nos 1–2, pp 166–72.
19
Stephenson, M.T., Hoyle, R.H., Palmgreen, P. and Slater, M.D.
(2003) ‘Brief measures of sensation seeking for screening and
large-scale surveys’, Drug and Alcohol Dependence, vol 72, no 3,
279–86.
20
Šmahel, D., Vondráčková, P.,Blinka, L. & Godoy-Etcheverry, S.
(2009). Comparing addictive Behavior on the Internet in the
Czech Republic, Chile and Sweden. In G. Cardosso, A. Cheong,
J. Cole (Eds.), World wide internet: Changing societies,
economies and cultures (pp. 544-582). Macao : University of
Macau.
21
Nicolaas, G. (2011) ‘Survey paradata: a review. ESRC National
Centre for Research Methods review paper, London: National
Centre for Research Methods.
For a list of key measurements used in the analysis of
the data see Annex 5.
7.2. SES measurements
Information relating to the chief income earner’s level of
education and occupation was collected during the
screening process. As outlined in Table 12, responses to
level of education and employment were then grouped
and cross-referenced with each other to calculate one of
three levels of SES: low, middle and high.
However, it should be noted that, as is often the case with
European research, a uniform approach was taken to the
calculation of SES across all 25 countries, and therefore
SES is not relative to the differences between the sociodemographic
make up of each country.
7.3. Education
Derived variables were also created to consider the level
of education within the household. Information on the
education of parents came from three questions.
 SCR6orig comes from the screening interview and
asks about the highest education level of the head of
household.
 QP209 is in the parent questionnaire and asks about
the highest education level completed by the parent
(or carer) that is being interviewed.
 QP210 is in the parent questionnaire and asks about
the highest level of education completed by the other
parent (or carer) if there is such a person.
One of the challenges for the project was to create a
central understanding of the different levels of education that
could be applied across all countries - whilst taking
into account the different education systems that exist
across Europe.
 
43
Table 12: Socio-Economic Status of the Chief Income Earner
Education of Main Wage Earner (SCR6orig)
Occupation of Main wage earner (SCR7)
Less than
primary Primary Secondary Tertiary
General management / Self employed professional Low Middle High High
Employed professional / Middle management / Business prop Low Low High High
Farmer / Fisherman Low Low High High
Employed desk position / Owner of shop, craftsmen Low Low Middle High
Employed position, not at a desk / Supervisor, skilled manual worker Low Low Middle High
Unskilled manual worker, servant Low Low Low Low
Non active (housework, student, unemployed) Low Low Low Middle
Non active retired Low Low Middle Middle
Therefore although respondents answered a question that
was specific to their country, and reflected the different
levels within their system of education, responses to
SCR6orig, QP209 and QP210 were all later mapped into
the derived variable DPEDUHH comprising a central
model of seven different levels of education in line with
the International Standard Classification of Education
(ISCED; UNESCO, 2006):
 Not completed primary education
 Primary or first stage of basic
 Lower secondary or second stage of basic
 Upper secondary
 Post secondary, non tertiary
 First stage of tertiary
 Second stage of tertiary
The mapping of individual education systems to these
seven central codes was undertaken in consultation with
the relevant academics from the EU Kids Online network;
however there remained several challenges. For example
several education systems have courses or levels that fall
in between or transcend across two of the seven
variables; or for cultural differences such as in Germany,
respondents underrepresented their tertiary education
because not all gained qualifications at the end of their
study.
A further difficulty in interpreting level of education is that
the level of education profile of the survey population is
unknown. Although Eurostat data
22
of adults 25-64 is used
to generate an indicative comparison below, the adults in
the EU Kids Online project take a different profile: namely
they are parents (not aged 25-64 per se), of children 9-16,
and whose children use the internet. Crucially, information
about respondents’ level of education has been collected
in different ways by Eurostat and EU Kids Online. It is
therefore not possible to use level of education as a
variable in weighting the data, and comparisons of the
population data and the survey profile should be
treated with caution.
The difficulty in translating and mapping different
education systems together and the inability to weight the
data to a known population profile for education help
explain why the level of education appears under- or
overrepresented in some countries. Table 15 shows the
education level as measured in the EU Kids Online data
for the head of household and by Eurostat in the adult
population aged 25-64 years
23
. In the EU Kids Online data
set the estimated percentage of households where the
head of household has completed tertiary education
ranges from 9 percent in Turkey to 82 percent in Norway.
22
Eurostat can be found here:
http://epp.eurostat.ec.europa.eu/portal/page/portal/education/dat
a/database#
23
The highest education level of the household (the variable
DPEDUHH) is calculated by taking the highest level of education
across SCR6, Q209 and Q210.
Technical Report and User Guide: The 2010 EU Kids Online Survey
44
The Eurostat figures for the individual adult population
range from 12 percent to 38 percent.
Table 15: Level of education as measured in the EU
Kids Online data and by Eurostat
Percent with a tertiary education
As measured in the
EU Kids Online data
As estimated by
Eurostat for the
adult population
SCR6 DPEDUHH 2009 2010
Difference
between
DPEDUHH
and
Eurostat
2010
AT 13 17 19 19 -2
BE 35 41 33 35 6
BG 25 33 23 23 10
CY 21 24 34 36 -12
CZ 20 24 16 17 7
DE 7 12 26 27 -15
DK 52 54 34 34 20
EE 23 31 36 35 -4
EL 20 24 23 24 0
ES 15 19 30 31 -12
FI 18 26 37 38 -12
FR 32 38 29 29 9
HU 16 20 20 20 0
IE 21 26 36 37 -11
IT 9 15 15 15 0
LT 25 32 31 33 -1
NL 51 61 33 32 29
NO 76 82 36 37 45
PL 19 22 21 23 -1
PT 9 10 15 15 -5
RO 15 19 13 14 5
SE 38 54 33 34 20
SI 28 33 23 24 9
TR 8 9 12 12 -3
UK 16 19 33 35 -16
As demonstrated in Figure 4 below, the correlation
between the percentages obtained in the EU Kids Online
data set and the Eurostat figures is perhaps lower than
expected.
Figure 4: Education as measured in the EU Kids
Online survey and as estimated by Eurostat
AT
BE
BG
CY
CZ
DE
DK
EE
EL
ES
FI
FR
HU IE
IT
LT
NL
NO
PL
PT
RO
SE
SI
TR
UK
0
10
20
30
40
50
60
70
80
90
100
0 10 20 30 40 50
%Withtertiaryeducation(EUKidsOnline)
% With tertiary education (Eurostat)
It could be hypothesised that the EU Kids Online figure
should be slightly higher than that of the Eurostat adult
population, especially considering that the EU Kids Online
figure accounts for the highest level of education across
the household as a whole rather than just individual
adults. Looking at Table 15, it therefore appears that the
level of education is overestimated in three countries:
Norway (+45), Netherlands (+29), Denmark (+20) and
Sweden (+20); in contrast, the level of education seems to
be underrepresented in the UK (-16), Ireland (-11),
Finland (-12), Germany (-15) and Cyprus (-12). However it
is worth noting that although comparisons between some
countries should be treated with caution, the level of
education variable provides a useful indication of the
variation in education between households within the
same country for which additional country specific
variables can be found in the data set (ATeduc to
UKeduc) and the education pack.
To allow analyses on the European level and country
comparisons which include the education variable the
EU Kids Online network took the following approach:
Following the suggestion of the International
Telecommunication Union (ITU, 2010) the derived
variable DPEDUHH4 was created for cross-country
analyses that include education. The variable contains a
 
45
four-way classification of education using ISCED97 as
follows:
1. Primary education or lower – no formal education, preprimary
(ISCED 0) or primary education (ISCED 1);
2. Lower secondary education (ISCED 2);
3. Upper secondary or post-secondary non-tertiary
(ISCED 3, 4); and
4. Tertiary (ISCED 5, 6).
7.4. Routing and handling of
missing values
The use of routing in the questionnaire calls for special
care in handling of missing values in the analysis of the
EU Kids Online data set. The following is the question on
bullying experienced in past 12 months:
Has someone acted in this kind of hurtful or nasty way to you in the PAST
12 MONTHS?
PLEASE TICK ONE BOX ONLY
Yes  Answer question on next page
No 
Don’t know  Go straight to section C
Prefer not to say 
This is the frequency table in SPSS showing that some
93% of the children (16.6+73.6) give a definite answer to
this question. The remaining 7% say that they don’t know
(coded ad -98), that they prefer not to say (coded as -97)
or simply do not answer the question (coded as -99).
QC112 Has someone acted in this kind of hurtful or nasty way to you in the past 12
months?
48 ,2 ,2 ,2
1126 4,8 4,8 5,0
416 1,8 1,8 6,8
4587 19,6 19,6 26,4
17243 73,6 73,6 100,0
23420 100,0 100,0
-99
-98 Don't know
-97 Prefer not to say
1 Yes
2 No
Total
Valid
Frequency Percent Valid Percent
Cumulative
Percent
Only those who answer with a definite yes continue to
answer the following question on how often bullying has
been experienced in the past 12 months.
How often has someone acted in this kind of way towards you in the PAST
12 MONTHS?
PLEASE TICK ONE BOX ONLY
Every day or almost every day 
Once or twice a week 
Once or twice a month 
Less often 
Don’t know 
Below is the frequency table in SPSS and here a new
missing value has been introduced (-96) for those who
were routed out of the section in the previous question. It
is important to note however that this value contains a
mixture of answers from the previous question and thus
can’t be seen as representing those who have not
experienced bullying in the past 12 months (although this
group is the vast majority of those ending up in the -96
category). As in the question on if the children had been
bullied at all they can also in the question on how often
choose to say that they don’t know (coded as -98 as
before) or skip the question (coded as -99).
QC113 How often has someone acted in this kind of way towards you in the past 12
months?
18 ,1 ,1 ,1
387 1,7 1,7 1,7
18833 80,4 80,4 82,1
345 1,5 1,5 83,6
659 2,8 2,8 86,4
870 3,7 3,7 90,1
2308 9,9 9,9 100,0
23420 100,0 100,0
-99
-98 Don't know
-96
1 Every day or almost
every day
2 Once or twice a week
3 Once or twice a month
4 Less often
Total
Valid
Frequency Percent Valid Percent
Cumulative
Percent
The table below shows how the answers from question
113 on how often bullying has been experienced map
onto the answers from question 112 on whether bullying
has been experienced at all in the past 12 months. This
shows how all the missing values from question 112 have
been put together into one missing value in question 113
(the -96 group). This shows also how some 9% of those
who said in question 112 that they had experienced
bullying in the past 12 months do not give a valid answer
in question 113 on how often this has happened.
QC112 Has someone acted in this kind of hurtful or nasty way to you in the past 12 months?
-99
Missing
-98
Don't
know
-97
Prefer
not to
say 1 Yes 2 No Total
-99 Missing 18 18
-98 Don't know 387 387
-96 Routed out 48 1126 416 17243 18833
Valid answer 4182 4182
Total 48 1126 416 4587 17243 23420
Internal mortality 9%
QC113 How often has
someone acted in this
kind of way towards
you in the past 12
months?
Technical Report and User Guide: The 2010 EU Kids Online Survey
46
The next question presents a new issue to think about.
Here the children are asked how bullying has happened in
the past 12 months and as this can happen in more than
one way they can tick as many boxes as they want.
At any time during the last 12 months, has this happened … ?
PLEASE TICK AS MANY BOXES AS NEEDED
In person face to face 
By mobile phone calls, texts or
image/video texts 
Some other way 
Don’t know 
The frequency table for question 114a shows that there
are two kinds of missing values. Those who do not tick
any of the response options are coded as -99 and those
who were routed out in question 112 have been coded as
-96. Those who ticked the box for ‘In person face to face’
are coded as ‘Yes’ and everyone else is coded as ‘No’.
QC114a At any time during the last 12 months, has this happened ...?: In
person face to face
45 ,2 ,2 ,2
18833 80,4 80,4 80,6
1518 6,5 6,5 87,1
3024 12,9 12,9 100,0
23420 100,0 100,0
-99
-96
0 No
1 Yes
Total
Valid
Frequency Percent Valid Percent
Cumulative
Percent
As before in comparing questions 112 and 113 the same
thing can happen here that children who have said in
question 112 that they have been bullied do not give a
valid answer in question 114. The table below shows how
the first option in question 114 maps onto question 112.
QC112 Has someone acted in this kind of hurtful or nasty way to you in the past 12 months?
-99
Missing
-98
Don't
know
-97
Prefer
not to
say 1 Yes 2 No Total
-99 Missing 45 18
-98 Don't know 0 387
-96 Routed out 48 1126 416 17243 18833
Valid answer 4542 4182
Total 48 1126 416 4587 17243 23420
Internal mortality 1%
QC114a At any time
during the last 12
months, has this
happened ...?: In
person face to face
The internal mortality between question 112 and question
114 is much lower than between questions 112 and 113.
The reason is that in the variable holding the information
from response option 114a the only missing values are
those who do not pick any response option in question
114 and those who tick the ‘Don’t know’ option are all
coded as ‘No’ in the other response options.
QC114a At any time during the last 12 months, has this happened ...?: In person face to face
* QC114d At any time during the last 12 months, has this happened ...?: Don't know
Crosstabulation
Count
45 0 0 0 45
0 18833 0 0 18833
0 0 1128 390 1518
0 0 3024 0 3024
45 18833 4152 390 23420
-99
-96
0 No
1 Yes
QC114a At any time
during the last 12
months, has this
happened ...?: In
person face to face
Total
-99 -96 0 No 1 Yes
QC114d At any time during the last 12 months,
has this happened ...?: Don't know
Total
It should be stressed that the issues related to
missing values in the EU Kids Online data are more
complex than in many other surveys. The preferred
setting of missing values depends however on the
nature of the analysis and is by no means default or
natural in the data set. However, as a
recommendation it is advised to follow the approach
taken by the EU Kids Online network.
7.5. Treatment of missing values
by EU Kids Online
The exact number of percentages reported will be
dependent on how missing values are treated and which
of them are included or excluded from the base.
The base determines which respondents were included
for reporting percentages of a particular variable. In other
words, the base is the actual number that makes up 100%
of the reported data. There are different ways in defining
the base dependent on how non-responses (-96 and -99),
‘don’t know’ (-98), and ‘prefer not to say’ (-97) responses
are treated. In the EU kids online II project we decided to
exclude all of the above from the base, that is,
we defined the base as: All respondents, who have
given a valid answer to a question, such as ‘yes’, ‘no’
or any response option that is not ‘don’t know’ or
‘prefer not to say’
The assumption underlying this decision was that the
likelihood of each missing respondent for one of the
response options equals the likelihood which with that
 
47
Have you acted in a way that might have felt hurtful or nasty to someone
else in the PAST 12 MONTHS?
PLEASE TICK ONE BOX ONLY
Yes  Answer questions on next page
No 
Don’t know  Go straight to section D
Prefer not to say 
In which of the following ways have you acted like this in the PAST 12
MONTHS? …
PLEASE TICK AS MANY BOXES AS NEEDED
In person face to face 
By mobile phone calls, texts or
image/video texts 
On the internet 
Other way(s) 
Don’t know 
-96
response option was chosen in the sample (missing at
random).
Two exceptions to this rule were made:
1)When the response ‘don’t know’ or ‘prefer not to say’
was considered meaningful to report – e.g., due to the
topic in question or because a large number of
respondents had chosen them – then these responses
were included in the base. This was, for example, the
case when we reported parental awareness of their
child experiencing a risk. Here we reported the
percentages of parents who had said ‘yes’, ‘no’ and
‘don’t know’ to the question whether they thought their
child had encountered a particular risk.
2)When not applicable was assigned because a
respondent was not routed to a question due to having
given a response at a previous question that made
him/her not receive that particular question AND the
respondent should still be included in the base for
percentage reporting. This was mainly the case when
percentages were reported for follow up questions of
online risks.
An example for the second case would be, when
reporting the percentage of all children who use the
internet who have bullied others online (i.e. said ‘yes’ to
QC125 and QC127c). Now those children who have not
said ‘yes’ to the question whether they have bullied others
at all (QC125) will not have been routed to the follow up
question which asked them in which mode they have
bullied others (QC127) and therefore been coded as ‘not
applicable’ (-96) for this question.
However, we can assume that those who have said ‘no’ to
whether they have bullied at all (QC125) would also have
said ‘no’ to whether they have bullied online (QC127c).
Hence, for those who have said ‘no’ to whether they have
bullied at all (QC125) the ‘not applicable’ will need to be
recoded into ‘no’ if the base for percentage reporting are
intended to be all children who use the internet.
A similar approach was taken for other follow up
questions when the intention was to include those that
were routed out into the base for percentage reporting.
A different procedure would be taken if we would like to
report the percentage of those who have bullied others
online (i.e. said ‘yes’ to QC127c) of all children who have
bullied in general. In this case the base for calculation
would be all those who said ‘yes’ to the question of
whether they have bullied in general (QC125) and have
been routed to the question whether they have bullied
online (QC127c). In this case the complete base has been
routed to the variable in question and no recoding of
missing values would be needed.
Have you acted in a way that might have felt hurtful or nasty to someone
else in the PAST 12 MONTHS?
PLEASE TICK ONE BOX ONLY
Yes  Answer questions on next page
No 
Don’t know  Go straight to section D
Prefer not to say 
In which of the following ways have you acted like this in the PAST 12
MONTHS? …
PLEASE TICK AS MANY BOXES AS NEEDED
In person face to face 
By mobile phone calls, texts or
image/video texts 
On the internet 
Other way(s) 
Don’t know 
-96
no
yes
no
no
-96
no
yes
no
no
no
yes
no
no
-98
-96
-96
-98
-96
-96
Have you acted in a way that might have felt hurtful or nasty to someone
else in the PAST 12 MONTHS?
PLEASE TICK ONE BOX ONLY
Yes  Answer questions on next page
No 
Don’t know  Go straight to section D
Prefer not to say 
In which of the following ways have you acted like this in the PAST 12
MONTHS? …
PLEASE TICK AS MANY BOXES AS NEEDED
In person face to face 
By mobile phone calls, texts or
image/video texts 
On the internet 
Other way(s) 
Don’t know 
-96 → no
no
yes
no
no
-96 → no
no
yes
no
no
no
yes
no
no
-98
-96
-96
-98
-96
-96
Technical Report and User Guide: The 2010 EU Kids Online Survey
48
 
49
LIST OF FIGURES
Figure 1: Countries surveyed by EU Kids Online............. 7 
Figure 2: Estimated margin of error for findings based on
the EU Kids Online data set........................................... 10 
Figure 3: Field work steps and respondent mortality ..... 27 
Figure 4: Education as measured in the EU Kids Online
survey and as estimated by Eurostat............................. 44
LIST OF TABLES
Table 1: Countries and two letter country codes.............. 7
Table 2: List of fieldwork agencies................................... 8
Table 3: Languages provided in the EU Kids Online
survey in each of the participating countries.................. 14
Table 4: Method of stratification by region and
urbanisation ................................................................... 16
Table 5: Sampling information ....................................... 17
Table 6: Fieldwork dates, incentives and number of
interviewers.................................................................... 22
Table 7: Survey mode and interview length................... 23
Table 8: Contact, cooperation and response rates, by
country........................................................................... 27
Table 9: Details of non-routing based edits.................... 31
Table 10: Estimated number of children aged 9-16 who
use the internet, by country............................................ 35
Table 11: Variables used to calculate design effects..... 37
Table 12: Design effects and effective sample sizes by
country........................................................................... 38
Table 13: Variable names of core variables................... 41
Table 14: Variable names of core variables................... 41
Table 13: Level of education as measured in the EU Kids
Online data and by Eurostat .......................................... 44
Risks and safety on the internet: The perspective of European children
50
 
51
ANNEX 1: EU KIDS ONLINE
Overview
EU Kids Online II: Enhancing Knowledge Regarding
European Children’s Use, Risk and Safety Online is
funded from 2009-2011 by the EC’s Safer Internet
Programme.
The project aims to enhance knowledge of European
children’s and parents’ experiences and practices
regarding risky and safer use of the internet and new
online technologies, in order to inform the promotion of a
safer online environment for children among national and
international stakeholders.
Adopting an approach that is child-centred, comparative,
critical and contextual, EU Kids Online has conducted a
major survey of children’s experiences (and their parents’
perceptions) of online risk in 25 European countries. The
findings will be disseminated through a series of reports
and presentations during 2010-12.
Objectives
 To design a robust survey instrument appropriate for
identifying the nature of children’s online access, use,
risk, coping and safety awareness.
 To design a robust survey instrument appropriate for
identifying parental experiences, practices and
concerns regarding their child’s internet use.
 To administer the survey in a reliable and ethicallysensitive
manner to national samples of internet
users aged 9-16 and their parents in Europe.
 To analyse the results systematically to identify core
findings and more complex patterns among findings
on a national and comparative basis.
 To disseminate the findings in a timely manner to a
wide range of relevant stakeholders nationally, across
Europe, and internationally.
 To identify and disseminate key recommendations
relevant to the development of safety awareness
initiatives in Europe.
 To identify remaining knowledge gaps and
methodological guidance to inform future projects on
the safer use of online technologies.
Work packages
WP1: Project Management and Evaluation: ensure
effective conduct and evaluation of work packages.
WP2: Project Design: design a robust survey instrument
and sampling frame for children and parents.
WP3: Data Collection: tender, select and work with the
subcontractor appointed to conduct the fieldwork.
WP4: Data Reporting: cross-tabulation, presentation and
report of core findings.
WP5: Statistical Analysis of Hypotheses: analysis and
hypothesis testing of relations among variables.
WP6: Cross-National Comparisons: interpretation of
similarities and differences across countries.
WP7: Recommendations: guide awareness and safety
initiatives and future projects in this field.
WP8: Dissemination of Project Results: dissemination to
diverse stakeholders and the wider public.
International Advisory Panel
 María José Cantarino, Corporate Responsibility
Manager, Telefonica, Spain.
 Dieter Carstensen, Save the Children Denmark,
European NGO Alliance on Child Safety Online.
 Professors David Finkelhor and Janis Wolak, Crimes
against Children Research Center, University of New
Hampshire, USA.
 Will Gardner, CEO of Childnet International, UK.
 Dr Ellen Helsper, Department of Media and
Communications, London School of Economics, UK.
 Amanda Lenhart, Pew Internet & American Life
Project.
 Prof Eileen Munro, Department of Social Policy,
London School of Economics, UK.
 Annie Mullins, Global Head of Content Standards,
Vodafone, UK.
 Kjartan Ólafsson, University of Akureyri, Iceland.
 Janice Richardson, project manager at European
Schoolnet, coordinator of Insafe, Brussels, Belgium.
 Agnieszka Wrzesień, Project Coordinator, Polish
Safer Internet Node, Nobody’s Children Foundation.
Technical Report and User Guide: The 2010 EU Kids Online Survey
52
ANNEX 2: THE NETWORK
Country National Contact Information Team Members
Austria (AT) Ingrid Paus-Hasebrink ingrid.paus-hasebrink@sbg.ac.at
Department of Audiovisual Communication, University of
Salzburg, Rudolfskai 42, A-5020 Salzburg, Austria
Ingrid Paus-Hasebrink
Andrea Dürager
Belgium (BE) Leen D'Haenens Leen.DHaenens@soc.kuleuven.be
Centrum voor Mediacultuur en
Communicatietechnologie (OE), OE Centr. Mediacult.&
Comm.technologie,
Parkstraat 45 – bus 3603, 3000 Leuven, Belgium
Leen d'Haenens
Verónica Donoso
Sofie Vandoninck
Joke Bauwens
Katia Segers
Bulgaria (BG) Jivka Marinova gert@mbox.contact.bg
Gender Education, Research and Technologies
foundation, P.O.B. 963, Sofia 1000, Bulgaria
Jivka Marinova
Diana Boteva
Cyprus (CY) Yiannis Laouris laouris@cnti.org.cy
Cyprus Neuroscience & Technology Institute
Science Unit of the Future Worlds Center
5 Promitheos, 1065 Lefkosia, Cyprus
Yiannis Laouris
Tatjana Taraszow
Elena Aristodemou
Aysu Arsoy
Czech Republic (CZ) David Šmahel smahel@fss.muni.cz
Faculty of Social Studies, Masaryk University
Joštova 10, 602 00 Brno, Czech Republic
David Šmahel
Štepán Konečný
Lukáš Blinka
Anna Ševčíková
Petra Vondráčková
Alena Černá
Denmark (DK) Gitte Stald stald@itu.dk
IT University of Copenhagen,
Ruud Langgaards Vej 7, 2300 Copenhagen, Denmark
Gitte Stald
Estonia (EE) Veronika Kalmus Veronika.Kalmus@ut.ee
Institute of Journalism and Communication, University of
Tartu, 18 Ülikooli St., 50090 Tartu, Estonia
Veronika Kalmus
Pille Pruulmann-
Vengerfeldt
Pille Runnel
Andra Siibak
Kadri Ugur
Lennart Komp
Finland (FI) Reijo Kupiainen reijo.kupiainen@uta.fi
Department of Journalism and Mass Communication,
University of Tampere, 33014 Finland
Reijo Kupiainen
Kaarina Nikunen
Annikka Suoninen
Riitta Kauppinen
France (FR) Dominique Pasquier Dominique.Pasquier@ehess.fr
Ecole Nationale Supérieure des Télécommunications
46 rue Barrault, 75013 Paris, France
Dominique Pasquier
Sylvie Octobre
Elodie Kredens
Pauline Reboul
Germany (DE)
(Management Group)
Uwe Hasebrink u.hasebrink@hans-bredow-institut.de
Hans Bredow Institute for Media Research
Warburgstr. 8-10, D - 20354 Hamburg, Germany
Uwe Hasebrink
Claudia Lampert
Greece (EL) Liza Tsaliki etsaliki@media.uoa.gr
Department of Mass Media and Communications
National and Kapodistrian University of Athens
5 Stadiou Street, Athens 105 62, Greece
Liza Tsaliki
Despina Chronaki
Eleni-Revekka Staiou
Kalpaki Kornilia
Konstantina
Michalopoulou
Hungary (HU) Bence Ságvári bence.sagvari@ithaka.hu
Information Society and Network Research Center –
ITHAKA, Perc u. 8, Budapest, 1036 Hungary
Anna Galácz
Bence Ságvári
Erik Gerhradt
Zsófia Rét
 
53
Ireland (IE)
(Management Group)
Brian O’Neill brian.oneill@dit.ie
College of Arts and Tourism, Dublin Institute of
Technology, Rathmines Road, Dublin 6, Ireland
Brian O’Neill
Nóirín Hayes
Simon Grehan
Sharon McLaughlin
Italy (IT) Giovanna Mascheroni giovanna.mascheroni@unicatt.it
OssCom, Università Cattolica del S. Cuore
Largo Gemelli, 1, 20123 Milano, Italy
Fausto Colombo
Piermarco Aroldi
Barbara Scifo
Giovanna Mascheroni
Maria Francesca Murru
Lithuania (LT) Alfredas Laurinavičius allaur@mruni.eu
Department of Psychology, Mykolas Romeris University,
Ateities st. 20, LT-08303 Vilnius, Lithuania
Alfredas Laurinavičius
Laura Ustinavičūtė
Rita Žukauskiene
Netherlands (NL) Jos de Haan j.de.haan@scp.nl
Netherlands Institute for Social Research | SCP
P.O. Box 16164, 2500 BD Den Haag, The Netherlands
Jos de Haan
Patti M. Valkenburg
Marion Duimel
Els Kuiper
Linda Adrichem
Jochen Peter
Maria Koutamanis
Nathalie Sonck
Norway (NO) Elisabeth Staksrud elisabeth.staksrud@media.uio.no
Dept. of Media and Communication, University of Oslo
Boks 1093 Blindern, 0317 Oslo, Norway
Elisabeth Staksrud
Ingunn Hagen
Jørgen Kirksæther
Poland (PL) Lucyna Kirwil lucyna.kirwil@swps.edu.pl
Department of Psychology
Warsaw School of Social Sciences and Humanities
ul. Chodakowska 19/31, 03-815 Warsaw, Poland
Lucyna Kirwil
Aldona Zdrodowska
Portugal (PT)
(Management Group)
Cristina Ponte cristina.ponte@fcsh.unl.pt
Departamento de Ciências da Comunicação
Faculdade de Ciências Sociais e Humanas,
Universidade Nova de Lisboa (UNL)
Av. de Berna, 26-C, 1069-061 Lisboa, Portugal
Cristina Ponte
José Alberto Simões
Daniel Cardoso
Ana Jorge
Romania (RO) Monica Barbovschi moni.barbovski@gmail.com
Babes-Bolyai University, Faculty of Sociology and Social
Work, 21 Decembrie 1989 st. no.128-130, Cluj-Napoca,
Romania
Monica Barbovschi
Maria Diaconescu
Eva Laszlo
George Roman
Valentina Marinescu
Anca Velicu
Slovenia (SL)
(Management Group)
Bojana Lobe bojana.lobe@fdv.uni-lj.si
Centre for Methodology and Informatics
Faculty of Social Sciences, University of Ljubljana
Kardeljeva pl. 5, Ljubljana, Slovenia
Bojana Lobe
Sandra Muha
Spain (ES) Maialen Garmendia maialen.garmendia@ehu.es
Depto. de Sociología, Universidad del País Vasco,
Apartado 644, 48.080 Bilbao, Spain
Carmelo Garitaonandia
Maialen Garmendia
Gemma Martínez
Fernández
Miguel Angel Casado
Sweden (SE) Cecilia von Feilitzen cecilia.von.feilitzen@sh.se
The International Clearinghouse on Children,
Youth and Media, Nordicom, Goteborg University,
Box 713, 405 30 Goteborg, Sweden
Cecilia von Feilitzen
Elza Dunkels
Olle Findahl
Turkey (TR) Kursat Cagiltay kursat@metu.edu.tr
Department of Computer Education and Instructional
Technology, Faculty of Education, Middle East
Technical University, 06531, Ankara, Turkey
Kursat Cagiltay
Engin Kursun
Duygu Nazire Kasikci
Christine Ogan
Turkan Karakus
United Kingdom (UK)
(Coordinator,
Management Group)
Leslie Haddon leshaddon@aol.com
Department of Media and Communications
London School of Economics and Political Science
Houghton Street, London WC2A 2AE, UK
Sonia Livingstone
Leslie Haddon
Anke Görzig
Daniel Kardefelt-Winther
Technical Report and User Guide: The 2010 EU Kids Online Survey
54
ANNEX 3: ETHICS REVIEW
Questionnaire submitted to the LSE Research Ethics committee
Researchers should consider the following questions when devising research proposals involving human participants,
personal, medical or otherwise sensitive data or methodologically controversial approaches. N.B. not all of these questions
will be relevant to every study. These questions provide pointers to direct researchers’ thinking about the ethical
dimensions of their research. It is expected that researchers will already have addressed the academic justification for the
project in their proposal; the guidance questions set out below aim to help researchers address specific ethical issues in
so far as they relate to participants or data.
In particular, consideration of risks to the research participants versus benefits need to be weighed up by researchers. It is
important to think through carefully the likely impact on participants or vulnerable groups of any data collection methods.
Certain groups are particularly vulnerable, or will be placed in a vulnerable position in relation to research, and may
succumb to pressure; for example children or people with learning disability, or students when they are participating in
research as students. Some participants will have diminished capacity to give consent and are therefore less able to
protect themselves and require specific consideration (see further guidance given on the RPDD web pages regarding
informed consent). The Research Ethics Committee (REC) recognizes that it is not only research with human participants
that raises relevant ethical concerns. Researchers may be assessing sensitive information, the publication or analysis of
which may have direct impact on agencies, communities or individuals. For example, collection and use of archive,
historical, legal, online or visual materials may raise ethical issues (e.g for families and friends of people deceased), and
research on provision of social or human services may impact user provision. Similarly, use of other people’s primary data
may need clearance or raise concerns about its interpretation. The Research Ethics Committee will assess whether the
relevant questions have been adequately addressed when it scrutinises proposals. Please ensure that each answer
provides the Committee with enough information to make an informed decision on the ethical dimensions of the proposal.
The LSE Research Ethics Policy and guidance will be reviewed annually and may be subject to further development.
I. Project Details
Project Title:
EU Kids Online II:
Enhancing knowledge regarding European children’s use, risk and safety online
II. Applicant Details
Name: Sonia Livingstone
Status (delete as applicable) Professor, Department of Media and Communications
Email address: s.livingstone@lse.ac.uk
Room number/contact address: S105/ 7710
 
55
III. Research Aims
Please provide brief details of the research aims and the scientific background of the research. A full copy of the proposal
should be attached to this document.
During 2008, the European Commission’s 2005-8 Safer Internet Plus Programme called for “knowledge enhancement
projects that aim to increase the knowledge relevant to the issue of safer online technologies”, specifically to strengthen
the knowledge base by conducting “a comparable quantitative study of children's use of online technologies, with a
mapping of parents’ views of their children's use of online technologies”. See
http://ec.europa.eu/information_society/activities/sip/index_en.htm
The London School of Economics and Political Science, as Coordinator of the multinational EU Kids Online network
(see www.eukidsonline.net), has been awarded a contract for this work from 1/7/2009 to 30/6/2011. The aim is to
enhance the knowledge base for children’s and parents’ experiences and practices in relation to risky and safer use of
the internet and new online technologies in Europe, in order to inform the promotion of a safer online environment for
children.
The objectives are as follows:
- To design a thorough and robust survey instrument appropriate for identifying the nature of children’s online
access, use, risk, coping and safety awareness.
- To design a thorough and robust survey instrument appropriate for identifying the nature of parental
experiences, practices and concerns regarding their children’s internet use.
- To administer the survey in a reliable and ethically-sensitive manner to national samples of internet users
aged 9-16, and their parents, in member states.
- To analyse the results systematically so as to identify both core findings and more complex patterns among
findings on a national and comparative basis.
- To disseminate the findings in a timely manner to a wide range of relevant stakeholders nationally, across
Europe, and internationally.
- To identify and disseminate key recommendations relevant to the development of safety awareness initiatives
in Europe.
- To identify any remaining knowledge gaps and methodological lessons learned, to inform future projects
regarding the promotion of safer use of the internet and new online technologies.
- To benefit from, sustain the visibility of, and further enhance the knowledge generated by, the EU Kids
Online network.
These objectives will be achieved through the design and conduct of a comparable quantitative survey of children’s use
of online technologies across member states, together with a survey of parents’ experiences, practices and concerns
regarding their children’s online risk and safety. The survey questionnaires will be conducted in home, face to face,
with one parent and then the selected child.
Pilot research and cognitive testing with children will inform the design of the survey questionnaire, as will the detailed
literature review conducted by the Safer Internet programme’s previous grant to the EU Kids Online network (2006-9).
The network comprises experienced social researchers in 25 countries - member states, EEA and candidate countries
that vary in geography (north/south, urban/rural), wealth, culture (language, religion), position in Europe (EU15, recent
entrants from Eastern Europe) and internet history and penetration. 1000 children will be interviewed in each country,
drawn using a random stratified sampling procedure (see the attached statement from Ipsos Mori on detailed sampling
procedures).
The countries included are: Austria, Belgium, Bulgaria, Cyprus, Czech Republic, Denmark, Estonia, France, Germany,
Greece, Hungary, Ireland, Italy, Lithuania, Netherlands, Norway, Poland, Portugal, Romania, Slovenia, Spain, Sweden,
Turkey, United Kingdom and, on a self-paying basis, Finland. In each country, the research teams are paired with the
national node for the EC’s Insafe network of awareness-raisers, educators and policy/government stakeholders, to
ensure the evidence is used to inform policy (see www.saferinternet.org). These nodes are also producing the safety
information to be left with each child during fieldwork.
Technical Report and User Guide: The 2010 EU Kids Online Survey
56
As is the norm for a multi-country study, ethics approval is sought by the Coordinator at the LSE for the whole study,
rather than seeking approval from each participating member of the consortium. This was specified in the Description
of Work which forms the technical annex to the contract between LSE and the EC; it also serves as an annex to the
Network Members’ Agreement, signed by each institutional (university or research institute) member of the network
and countersigned by LSE. Additionally, the contracted fieldwork company, Ipsos Mori, is bound by the ethical
requirements of its professional market research association, ESOMAR, see http://www.esomar.org/index.php/codes-
guidelines.html.
In all that follows, everything will take place in the national (official) language(s) of the country concerned. Thus there
will be careful translation into all languages of the interviewer protocols, the letter of project introduction, the parent
and child survey questionnaires, the information leaflet and the final posting of accessible findings on the project
website. The EU Kids Online national teams will check translations provided by Ipsos Mori. The survey questionnaires
will be both translated and back translated, according to international procedures and standards governing such survey
translation processes.
1. Informed consent.
1.1 Will potential participants be asked to give informed consent in writing and will they be asked to confirm that they have
received and read the information about the study? If not, why not?
The fieldwork will conducted by Ipsos MORI - a highly reputable market research (polling) organisation appointed
following a European tender process. A requirement for the award of the contract was that data collection will be
conducted in a timely, efficient, rigorous and ethically sensitive manner by interviewers trained to deal with children,
so as to ensure high quality results that will command widespread respect. Accordingly, informed consent and
confirmation of receipt of information about the study will be a requirement for participation.
The survey will be conducted face-to-face in the child’s home, as this permits optimal sampling of individual children,
the convenience of obtaining parental permission, a parent interview and a child interview, and best ensures a reliable
and valid interview with the child. Consent from both parents and children will a prerequisite of both the main
fieldwork and also the prior phases of cognitive and pilot testing.
The process of gaining consent
- Ipsos Mori fieldwork interviewers will present written information about the project to participating parents
(where ‘parent’ refers to a person legally responsible for the child, and so could be the step-parent, foster-
parent).
- This letter will explain the funding and purposes of the project, the nature of the interview, the value of the
project to policy makers seeking to improve internet safety for children, and contact details for the national
fieldwork organisation (contracted to Ipsos Mori), the national EU Kids Online network representative, and
the project director (Sonia Livingstone for EU Kids Online at LSE).
- Those parents who agree to participate in the survey will be asked to sign a written consent form stating the
purpose and nature of the project (see Annex 2), this giving informed consent to their own interview and
consent to us approaching the child to invite their participation in the child interview.
- The child will also be asked to give informed consent to the child for their own interview. Ipsos Mori’s
experience leads them to recommend that the child is asked to confirm their consent verbally rather than in
writing. Asking children to sign a formal document is not necessarily conducive to engaging participation and
putting them at ease for the interview. Instead, the interviewer is asked to sign to confirm that they have
obtained informed consent verbally (see Annex 2).
- Both parent and child will be clearly informed that they may leave any question unanswered and they may
stop the interview at any point. The interviewers are trained to provide a calm and confidential context within
which children can express hesitation and be reassured or permitted to withdraw as appropriate.
- The consent process includes introductory wording tailored for parents and for children of different ages;
however, interviewers will also be instructed to tailor their approach for each respondent and work to ensure
that each respondent understands the nature of research in their own terms.
- Anonymity and confidentiality of responses is guaranteed to both parents and children, with one exception.
 
57
As shown in Annex 2, the small but possible risk that the child reports that they are being harmed in some
way will be handled as an explicit condition limiting the promise of confidentiality.
- If either parent or child denies consent, the interview will not take place. The interviewer will not enter a
home without a parent present and without express parental permission.
1.2. How has the study been discussed or are there plans to discuss the study with those likely to be involved, including
potential participants or those who may represent their views?
The study has been extensively discussed by those who represent the views and experiences of children in relation to
the internet. This includes meetings of the EU Kids Online network and with the EC’s Safer Internet Programme. It has
been designed partly in response to a series of focus groups the EC Safer Internet Programme held with children (aged
9-10 and 12-14) during 2007. INSAFE (on the advisory panel, below) maintains a Youth Panel which also advises the
Safer Internet Programme, including EU Kids Online.
EU Kids Online’s International Advisory Panel has been fully involved at all stages from the initial proposal draft to
the design of the survey and thereafter. Its purpose is to ensure that the project benefits from the best research practice
internationally and that its findings can be of maximum benefit to children. Its members are:
- Will Gardner, of Childnet International, the leading UK child welfare charity focused on internet-related risk
and safety issues;
- Professors David Finkelhor and Janis Wolak, of the Crimes against Children Research Center, University of
New Hampshire, USA – they conduct the leading American surveys examining internet-related risks to
children;
- Dr Ellen Helsper, formerly of the Oxford Internet Institute, now at the Department of Media and
Communications, LSE, experienced in the World Internet Project;
- Amanda Lenhart, Senior Research Specialist in teens and social networking at the Pew Internet & American
Life Project;
- Annie Mullins, Corporate Social Responsibility, Vodafone;
- Janice Richardson, director of INSAFE, the network of safety awareness-raising nodes for the Safer Internet
Programme, EC;
- Dieter Carstensen, Save the Children Denmark, and director of ENASCO, the European network of child
welfare NGOs in relation to internet safety;
- Agnieszka Wrzesień, of the Nobody′s Children Foundation, Poland;
- Maria José Cantarino, Corporate Social Responsibility, Telefonica;
- Professor Eileen Munro, Professor of Social Policy, LSE, expert in risk assessment and management in child
protection and welfare.
Now that the survey questionnaire is finalised and the sampling procedures and processes of administration are
determined, the development of the questionnaire will undergo cognitive testing with parents and children from a range
of ages across all of those countries involved in the survey. This will explore question wording, responses, themes and
the process of the interview including interpretations of the consent form. Furthermore the fieldwork will undergo a
piloting phase which will assess the success of the recruitment process and methods for conducting the questionnaire.
Technical Report and User Guide: The 2010 EU Kids Online Survey
58
1.3. Has information (written and oral) about the study been prepared in an appropriate form and language for potential
participants? (see Informed Consent guidance which lists questions to be considered). At what point in the study will this
information be offered?
As noted in section 1.3, information about the study has been prepared in an appropriate form and language for
potential participants. Information about the study will be provided orally and in written form as a letter to the parent
when the fieldwork interviewer from Ipsos Mori first visits the home to invite participation in the study.
If the parent wishes for more time to decide or if the timing is inconvenient for an interview, the interviewer will leave
a copy of the information letter with them and re-visit them on another day.
The letter will contain both LSE and Ipsos branding, plus contact details of the local fieldwork agency and the local EU
Kids Online network representative. It will also (as noted below) contain a url and date by which an accessible
summary of the findings will be posted.
An explanation of the nature and purposes of the study will be given orally to the child by the fieldworker. The child
will be left also with an information leaflet on useful child-friendly sources of help and guidance on matters concerning
online risk and safety.
As noted earlier, everything will take place in the national (official) language(s) of the country concerned. Thus there
will be careful translation into all languages of the interviewer protocols, the letter of project introduction, the parent
and child survey questionnaires, the information leaflet and the final posting of accessible findings on the project
website.
1.4 How will potential participants be informed of whether there will be adverse consequences of a decision not to
participate? Or of a decision to withdraw during the course of the study?
There are no adverse consequences of participating in the study. It is purely voluntary, there is no incentive payment,
and the survey is entirely anonymous.
At the point when the researcher first visits, potential participants will be advised that there will be no adverse
consequences if they decide not to participate and they can withdraw at any point, or choose not to answer specific
questions. Interviewers will be sensitive to the child’s mood or possible hesitation, and will remind the child of their
right to omit a question or to withdraw if appropriate.
1.5 What provision has been made to respond to queries and problems raised by participants during the course of the
study?
During the interview, the fieldwork interviewer will be the main point of contact for any explanation needed or to
address any concerns regarding the study. The letter of introduction, to be left with parents, will provide clear contact
details of national (and Coordinating) team of EU Kids Online II, plus contact details for the national fieldwork agency
(contracted by Ipsos Mori).
At the end of the interview, the child’s attention will be carefully drawn to further sources of information (in the form
of a child-friendly leaflet containing advice, contact information to national agencies and the national child helpline for
confidential advice).
The child will also be urged to discuss with a parent or trusted adult any concerns they have regarding things that may
have or could happen in relation to the internet (see end of Child Survey, attached to this application).
As explained below, interviewers are carefully trained, will be briefed on the particularities of this project, and are
supervised closely by the approved national fieldwork agency contracted to Ipsos Mori. They remain in close contact
with their supervisors and are required to report any problems to their supervisor.
In turn, the national fieldwork agency remains in close contact with the coordinating agency, Ipsos Mori in Belgium.
Ipsos Mori has appointed one key contact, Rosario Spadaro, to remain in weekly contact with the LSE coordinating
 
59
team (see Ipsos Mori’s proposal regarding fieldwork processes, attached to this application, for details of line
management and team coordination both within the Ipsos Mori network across Europe and for their communication
with LSE.).
It is anticipated that most if not all ethical issues (regarding sensitive questions or survey administration) will be
resolved during the cognitive testing and piloting phases of the research process. However, Ipsos Mori and LSE (Sonia
Livingstone) will remain in close contact throughout fieldwork, with weekly reporting and discussion planned and
more frequent or immediate communication possible if needed.
2. Research methodology.
2.1. How does the research methodology justify the use deception?
Not applicable
2.2. If the proposed research involves the deception of persons in vulnerable groups, can the information sought be
obtained by other means?
Not applicable
2.3. How will data be collected during the project? Please provide details of data analysis.
The data to be collected is largely quantitative survey responses from parents (plus one or two open ended questions
addressed to children). CAPI interview data is uploaded daily by fieldworkers to a national data base. PAPI interview
data is entered by the fieldworkers manually into the database. National fieldwork agencies will upload the national
data sets using a secure password-protected intranet, especially built for this project, to a single multinational data set
held by the Brussels coordinator (Ipsos Mori) shared with LSE. This means that weekly reports on progress (and any
problems) with data collection and fieldwork are shared with LSE and we are alerted early to any issues.
As explained in the original research proposal (see the Description of Work attached to this application), it was decided
that in home face to face interviews with children, in the comfort and privacy of their own home, offered the best
chance of obtaining reliable and valid information on sensitive issues.
Thus, data will be collected by face-to-face interviews conducted in home with parents and children in each of the
countries participating in the project.
The project will be explained in turn to the parent and the child, and informed consent will be obtained from the child
and young person and the parent for their own interviews.
Interviewers will be fully trained to ensure that consent is fully informed, in line with ESOMAR guidelines and the
core principles contained in LSE informed consent guidance.
Interviews will last, on average, 10 minutes for the parent and 30 minutes for the child.
Every effort will be made to ensure respondents are at ease in their domestic setting and the interviewer will be at pains
to create a comfortable situation in which questions can be asked, explained and/or refused without awkwardness.
Interviews will be administered via CAPI where possible, and by PAPI otherwise, with the highest priority given to
collection of high quality data in an ethical and sensitive manner. Specifying these requirements was central to the
public call for tender issued in spring 2009. As a result of this process, Ipsos Mori was appointed to conduct the
fieldwork in all 25 countries.
This process of selecting and approving Ipsos Mori is detailed below for it is important: though LSE is the coordinator
of the project, the fieldwork is entirely contracted out to Ipsos Mori. Hence the quality control process adopted by LSE
to make this contract, and the expertise of Ipsos Mori themselves is noted below. A full record of the tender process is
maintained by Margaret Newson, purchasing manager at LSE and will be reported to the European Commission. A
Technical Report and User Guide: The 2010 EU Kids Online Survey
60
lengthy document detailing the curriculum vitae of Ipsos Mori staff working on the project, plus their prior experience
in this field, was submitted to LSE as part of the tender process. Both documents are available to the REC on request.
The group evaluating the public tender process which appointed Ipsos Mori ensured ethical considerations were a key
criterion in awarding the contract. Those on the evaluation panel were:
- Professor George Gaskell, Deputy Director and Academic Governor, LSE;
- Professor Uwe Hasebrink, Hans Bredow Institute For Media Research, Hamburg;
- Dr Cristina Ponte, New University of Lisbon, Portugal;
- Dr Bojana Lobe, University of Ljubljana, Slovenia;
- Dr Brian O’Neill, Dublin Institute of Technology, Ireland;
- Margaret Newson, Finance Department, LSE;
- Bhimla Dheermojee, Research and Project Development Division, LSE
- Professor Sonia Livingstone, Project Director for EU Kids Online, LSE;
- Dr Leslie Haddon, Senior Research Fellow, EU Kids Online, LSE.
Ipsos MORI, successful winners of the tender, has a long and established tradition of social and government research.
They have a large team of around 200 experienced, specialist researchers in our Social Research Institute. Ipsos MORI
works extensively for both central and local government, conducting more research for this sector than any other UK
company. This, together with their national reputation among the public from our work as opinion pollsters, means that
they have additional credibility among a wide range of audiences. The UK based team from this project is drawn from
our specialist children and families research team.
They have considerable expertise in delivering large-scale random probability government surveys for numerous
government departments – including Department for Children, Schools and Families (DCSF), Home Office,
Department for Communities and Local Government, Department for Trade and Industry, Department for Work and
Pensions, Scottish Executive and the Welsh Assembly as well as for Agencies such as the Commission for Racial
Equality and Child Support Agency. Projects for DCSF involving similar surveys with children and/or parents include
the Longitudinal Survey of Young People in England; evaluation of Play Pathfinders, the Extended Schools Survey, the
evaluation of Find Your Talent. Much of their work has included researching those living in deprived communities and
those who are perceived as ‘hard-to-reach’ – experience which is key for enabling us to minimise non-response bias
and reach groups that are key for policy.
They also have a strong track record in delivering large scale surveys to target, time and to budget and their approach is
supported by the work of the Quantitative Research Methods Unit, chaired by Patten Smith, which not only supports
best practice internally, but contributes new methodological thinking of value industry wide regarding best practise
approaches to survey. Furthermore, IPSOS MORI has large experience in the coordination of international surveys.
Below, we offer examples of international research they have conducted in connection with children/parents, family,
young people and also use of internet.
- Particularly pertinent to the current research, the following surveys were carried out within the framework of
the Eurobarometer (15 Member States). Eurobarometer surveys were conducted by IPSOS (previously
INRA) among the population aged 15+; n=1.000 face-to-face interviews (except Germany: 2000,
Luxembourg: 600, United Kingdom 1300 including 300 in Northern Ireland).
- ‘Illegal and harmful content on the Internet’ (Eurobarometer 60.2) : This Eurobarometer 60.2 focused on the
following: places locations where child uses the Internet, setting rules for child on the use of various
entertainment applications, rules guidelines set for children on the use of Internet, the need for information on
protecting child from illegal and harmful content and contact on the Internet, awareness of amongst children
on what to do in case ain the event a situation on the Internet make him or her feel uncomfortable, preferred
sources and format of information on the safe use of the Internet, preferred information format on safe use of
the Internet, and awareness of where to report illegal or harmful content on the Internet.
- ‘Youth and drugs: TO YOUNG PEOPLE aged 15-24 only’ (Eurobarometer 57.2): This Eurobarometer 57.2
focused on: main reasons for experiencing experimenting with drugs, main reasons to find it hard to stop
using drugs, barriers to giving up drug use, Consequences of drugs using drugs, most effective ways of
tackling drug-related problems, how information is obtained on drugs, obtaining information about drugs;,
personal situation in relation to drugs, dangerousness of drugs, whether respondent personally takes drugs
 
61
and perceived dangers of drug use.
- ‘Internet usage’ (Eurobarometer 56.2): This Eurobarometer 56.2 focused on the use of Internet and the
periodicity of nternet usage and frequency of usage.
- ‘Young Citizens: TO YOUNG PEOPLE aged 15-24 only’ (Eurobarometer 55.1): This Eurobarometer 55.1
focused on: reasons why young people live longer in their parent’s home, leisure time activities, source of
money and use of ICT equipment.
In terms of data analysis, the EU Kids Online network, coordinated by LSE, bears sole responsibility for analysing and
disseminating the findings. LSE has appointed a postdoctoral survey research officer, from January 2010 to June 2011
(the official end of the project) to implement the analysis, as led by Sonia Livingstone and Leslie Haddon and as
advised by a team of survey experts within the network (and its international advisors).
The initial reporting of top line findings is timed for the EC’s major meeting of stakeholders in Luxembourg at the
Safer Internet Forum in October 2009. Thereafter, a series of reports, focusing on pan European similarities and
differences, is planned as specified in the Description of Work (attached to this application). The purpose is to balance
academic and policy ambitions by maximising the value of this unique and large data set in as timely a manner as
possible. This means prioritising policy and public dissemination in the short term and academic publication in the
longer term.
Three months after the final report (June 2011), the full data set will be deposited in a public archive (in October 2011)
to ensure maximum exploitation of the data set in the future. The project is intended not only to report on the state of
European children’s internet risk and safety experiences in 2010-11 but also to establish a benchmark against which
future trends can be measured.
2.4. How have ethical concerns arising from data collection been addressed?
The project participants and advisors have compared research practice across a series of recent projects focused on
asking children about risk and safety matters on the internet.
Our approach is set out in detail in section 3.1 below. Our intention is to draw on the best practice available in relation
to three research challenges – working with children, working in multiple countries and languages, and addressing
sensitive matters of risky experience.
These have been a core focus of the early network discussions which shaped the research proposal, a central theme in
the project’s kick off meeting (in a discussion led by Professor Eileen Munro, LSE, advisor to the project, along with Dr
Janis Wolak, who conducts the leading American surveys on internet risk to children. Since then, in additional to lively
electronic communication within the network, the network has met in full, with its advisors and with Ipsos Mori, in a
workshop in Hamburg in October 2009, at which survey sampling, design, administration, sensitive questions and
research ethics were all central topics. The advisors to the project are all active, expert and constructive.
The EC’s Safer Internet Programme also takes a close interest in the progress and design of the project and Sonia
Livingstone visits them in Luxembourg regularly and remains in frequent contact with the Project Officer.
3. Research design.
3.1 What concerns have been taken into account with regard to the design of the research project? If agencies,
communities or individuals are directly affected by the research (e.g. participants, service users, vulnerable communities
or relations), what means have you devised to ensure that any harm or distress is minimized and/or that the research is
sensitive to the particular needs and perspectives of those so affected?
Research importance
We note first, that at present there is no comparable, reliable data on children’s experience of online risks in Europe.
Indeed, there is no survey of children’s use of the internet in Europe that asks questions of any kind. At present, the
research and policy community is guided by existing surveys conducted in America, by pan-European surveys of
parents who then report on (their perceptions of) their child’s internet use, and by piecemeal surveys conducted with
Technical Report and User Guide: The 2010 EU Kids Online Survey
62
children that ask similar but not identical questions in different ways and to different samples in some countries only.
Hence, we address the ethical issues that arise in asking children about online risk in a wider context in which robust
evidence is almost wholly lacking, and in which a sizeable policy community of multi sector stakeholders, is
developing educational, industry, awareness raising and other initiatives which sorely need an evidence base to guide
them.
Survey design
The survey questionnaires (attached to this proposal) will ask a range of questions of children and parents, a central aim
being to develop a realistic assessment of the risks (range, severity, responses) experienced by children online. A
further aim is to identify the subset of children who are in some sense vulnerable –whether in their lives generally
and/or in their experiences of the internet in particular.
The areas covered in the children’s interview that relate to sensitive areas are:
- Range of activities engaged in online/varieties of sites and services used;
- The child’s experience of a wide range of specific risks;
- The nature, severity and consequences of specific risks experienced, including child’s risk responses and/or
coping;
- Possible mediators of risk (for example, measures of self-esteem, skills, vulnerability).
The areas covered in the parent’s interview that relate to sensitive areas will be:
- Their child’s experience of a wide range of specific risks;
- Parental regulation strategies (social, technical) in relation to perceived online risks experienced by children.
A crucial part of the project design is to ask matched questions of children and parents (particularly regarding
assessment of risk and nature of parental mediation). This will permit interesting forms of analysis comparing parents
and children who see things similarly or differently. It will also provide a much needed check on the widespread use of
parents to report on their children’s experience.
The purpose of the measures of child vulnerability (mainly here relying on the internationally used SDQ) is to permit
the study to go beyond standard demographic measures of risk. It is expected that, for a range of online experiences,
most children are sufficiently resilient to encounter risk with no distress. It is also expected that the minority of children
who do encounter distressing content or contact on the internet, their identification will be better pinpointed with a
subtle combination of social and psychological vulnerability factors rather than a simple demographic characterisation.
This, however, remains to be discovered.
It is also an important part of the research that we identify the incidence of online risk in relation to possible risks
encountered elsewhere (through other media or face to face experiences), the purpose being to enable a proportionate
response to online risk in the future by putting online risk in the context of other risky experiences.
Last, the project team are committed to identifying ways in which children may be resilient, to cope well, or to support
each other in addressing online risk. A series of questions will permit findings on these possibilities insofar as they do
exist, thus enriching public and policy discussions which are, at times, too simplistic in portraying all children as naïve
or vulnerable.
Interviewer training
Ipsos Mori is a member of ESOMAR and all local agencies also work within national industry ethical and legal codes.
All fieldwork will be conducted in line with stipulated ethical guidelines for conducting research with children and
young people, as well as those specified by the LSE Research Committee.
All fieldworkers will be experienced interviewers, including specific experience with conducting interviews with
children. They will receive a project-dedicated briefing, overseen by national members of the EU Kids Online network,
regarding specific issues for this project. CRB checks or equivalent (in line with local procedures, such as police
certificates of character and documents stating no criminal convictions in the past) will be required of all fieldworkers
(see also Ipsos Mori’s agreed proposal to LSE for details of interviewer training and experience with children, attached
 
63
to this proposal).
Before the interview takes place, respondents will be notified of their right to withhold answers to particular questions
or stop the interview at any point with no adverse consequences. This will also be reiterated at key stages during the
interview process. Confidentiality/anonymity will be guaranteed where there is not a disclosure of risk of harm.
To reassure both parents and children that it is safe for an adult interviewer to interview the child, the interview itself
the survey is administered in the child’s home with the parents in the vicinity, whilst care will also be taken to avoid
physical contact with children.
Sensitive questions
The flow of questions and use of gateway questions will aim to ensure that the interview does not introduce the child
for the first time to ideas or material that may be ethically problematic. Specifically, questions which ask about ‘risky’
behaviour will have introductory wordings where appropriate to forewarn of the nature of the next questions and to
clarify that the research does not condone such behaviour but that we are not passing any judgement on their response.
All questions will undergo thorough cognitive testing in each country – this means that while the survey is planned to
take 30 minutes on average, in cognitive testing fieldworkers will take up to two hours per child in order to clarify
misunderstandings, understand any hesitations, and so identify any problems. Only after this has been completed in all
languages/countries will be survey questionnaire be finalised. The network is, during November, constructing a table of
sensitive terminology by language to guide the translators and fieldworkers).
Further, to minimise distress, some questions will only be asked of children aged 11-16 and not those aged 9-10. If
required, more questions will be restricted to the older age groups only, as revealed by pilot testing.
In some countries the survey will be administered via CAPI and CASI (Computer Assisted Personal Interview and
Computer Assisted Self interview) whilst in other countries the interview will be completed on paper (by interviewer
and respondent). The interviewer will ask many questions in person, but participants will be asked to complete the most
sensitive questions (identifying their own risky behaviour) in a self-completion format and thus will not be asked to
disclose this information to the interviewer. This will help reassure the respond of confidentiality and anonymity
thereby encouraging honest answers.
The CASI approach will involve the interviewer handing the computer to the respond, explaining what they need to do
and then allowing them to complete the section. The self completion script will be user friendly, using formats tried and
tested with children and parents. It will start with a practice question. Answers will be stored electronically so that it is
clear to the respondent that they do not see their answers afterwards. The paper self-completion approach will be
similar, except that the respondent will be provided with a paper form, and an envelope into which they will put their
completed form to help reassure of confidentiality and that the interviewer won’t see the answers. The interviewer will
be on hand to answer queries if the respondent gets stuck at any point. .
Since the survey will collect data from parents and children, it is important to ensure confidentiality within as well as
beyond the family. Hence, it is important that, as far as possible, the parent does not oversee the child’s answers to
sensitive questions. Such privacy may be achieved by asking the parent to leave the room, by occupying the parent in
conversation while the child completes a self-completion portion of the questionnaire (written or on the computer) for
sensitive items, or by requesting the child to complete the self-completion portion and return to the interviewer in a
sealed envelope (or closing that section of a computer-assisted interview). The interview will note if the parent (or
other household members) are present or intrusive.
We will encourage parents to be absent from the room, but on hand near by during interviews, but the comfort and
wellbeing of children and parents will be paramount, and we will be flexible on this. If the parent does remain present
we will ask them to keep as low a profile as possible, and refrain from prompting the child or inputting into the survey
responses in any way.
Where there is a disclosure of a child being at a risk of serious harm that ‘no reasonable person could ignore’ steps will
be undertaken – considered on a case by case basis - by the research team to follow local procedures, laws and contact
national agencies.
Technical Report and User Guide: The 2010 EU Kids Online Survey
64
After the interview
The interviewer will thank the respondent and reassure/clarify once again about confidentiality, but also the value of
the study in helping policies to improve children’s safety on the internet.
The interviewer will also explain to the child that if they have experienced anything that has upset or worried them on
the internet that they should talk to a parent or other trusted adult so that they can help.
All respondents, parents and children, will be provided with information about online risk and safety, including local
help lines (or other appropriate provision for children identified through the conduct of the survey as in some way ‘at
risk’), whereby the respondent can access private, confidential help and advice.
If a child is considered possibly at risk
Given the important non-interventionist principles of social research, intervention will only be triggered on the basis of
relatively serious harm being identified. In general we will work according to the broad principle that this is
“something any reasonable person could not ignore”. Importantly we will follow national laws regarding the
types/levels of harm that should be acted upon.
Below we have summarised our approach to responding to (potential) harm if identified (i) from survey questions and
(ii) during the wider fieldwork process.
(i) Action that will be taken if a participant’s response to a survey question indicates that they may be potentially at risk
from harm.
- Some questions on experience of risks are included in the questionnaire. However, they ask about exposure to
risks in the past and do not directly identify current issues, although they may indicate the possibility of
current potential risk.
- Questions on risk will be asked within self completion modules and as such interviewers will not know the
child’s responses. We will therefore take a universal approach to responding to possible risk for all children.
The interviewer will explain to all children interviewed that if they have they have experienced harm, they
should tell a trusted adult,
- The interviewer will leave with the child a leaflet with helpline numbers and ‘top-tips’ to safety. These
leaflets are being developed for the project by the national Insafe nodes of the EC’s Safer Internet
Programme, with input also from Child Helpline International (see www.childhelplineinternational.org). The
leaflet (attached to this application) will provide safety tips, contact information (phone, email, url) for the
national Insafe node (the national child/internet safety organisation) and the main national child helpline
(members of the Child Helpline International Organisation).
- In addition, fieldwork agencies will abide by local laws regarding actions required to protect children.
(ii) Action that will be taken if a participant makes a disclosure to the interviewer outside their response to a survey
question and/or the interviewer witnesses something in the household suggesting that a child is at risk.
- If the interviewer becomes aware of risk of harm to a child that no reasonable person could ignore, or that
requires action within national laws, appropriate action will be taken.
- Given that disclosure of harm in this scenario is outside the main interview questions, this approach does not
conflict with guarantees of respondent confidentiality with regards to survey responses.
- The interviewer will report the "incident" to the project manager/field supervisor. Action will be taken by the
Institute, according to national law. Where institutes are not competent to make a decision of this kind, a
legal person will be consulted before action is decided upon.
- In such cases, the interviewer will also tell the child that they are concerned and talk to them about the action
that they will be taking. It will be preferable to gather the child’s consent, although in cases of serious cause
for concern there are exemptions (in some countries) where it appropriate to act with out this.
- As mentioned above, the interviewer will also encourage the child to talk to a trusted adult (if they have not
already done so) and provide them with the leaflet of top tips/help line support services.
 
65
3.2. How has the methodology addressed how sensitive information, data or sources will be handled?
Data from the parent will not be revealed to the child. Data from the child will not be revealed to the parent. The
sensitive portion of the questionnaire to the child, which is to be asked using self-completion methods (if a CAPI
interview, the screen is turned to the child only; if a PAPI interview, the child completes a paper and pen questionnaire
and places it themselves in a sealed envelope to give to the interviewer) is kept confidential to the child (ie neither
parent nor fieldwork knows of their responses).
The participants themselves will be advised during the introductory stages that data will be held securely and kept
confidential, and that the final data will stored, analysed and reported in a completely anonymised format. The contact
details of respondents will be kept linked to the survey data for just a very short time after the interview, to enable some
quality control call backs (15% of parent respondents are recontacted by telephone to check the conduct and content of
the interview, for purposes of quality control). However, after this process, all personal identifiers will be removed and
deleted on finalisation of the complete data set. The details of each interview case will be fully anonymised so that
anyone analysing that database will not be able to trace the participants.
All data will be held securely in line with data protection legislation and professional industry in each country.
Appropriate mechanisms for ensuring secure transfer of data between local agencies and the co-ordination centre and in
turn with the LSE will also be in place.
The data set to be delivered to LSE (EU Kids Online) will therefore be wholly anonymised. The quantitative data could
not be traced back to any individual. The inclusion of open-ended questions is currently subject to timing (i.e. the length
of the questionnaire overall) but should this be included still in the final version, all text will be checked by the national
EU Kids Online members so that any identifying information is removed. Only the wholly anonymised version of the
data set will be retained.
3.3. Have you been able to devise a timetable of research?
The project timetable as planned is set out in the Description of Work (attached) on p.30.
The timetable that follows provides a more detailed breakdown of fieldwork tasks to be completed by Ipsos Mori.
Since the cognitive testing phase, designed to ensure the questionnaire is thoroughly understood by children, was added
during contract negotiations with Ipsos Mori, the cognitive testing begins earlier than initially planned, and the main
fieldwork phase begins later than initially planned.
Overall, the timetable is very tight, but the ‘real’ deadline is to report key findings at the EC’s Safer Internet Forum in
October 2010, an event which all stakeholders across Europe and beyond attend each year.
Technical Report and User Guide: The 2010 EU Kids Online Survey
66
Milestones Sub-tasks Number
of weeks
Start date End date
Contract start date 4th
week of
September
Set-up meeting with LSE project team in London 29th September
Finalisation of the questionnaire and sampling schemes 3 weeks 1st October 22nd October
Meeting in Hamburg 16th October 18th October
LSE send out new draft of questionnaire 20th October
Input from Ipsos sent to LSE 22nd October
Final questionnaire 23rd October
LSE communicate questions that need testing 23rd October
Cognitive testing including feedback from LSE and
questionnaire finalisation
13 weeks 23rd October 22nd Jan 2010
Design of interview guide for cognitive testing 23rd October 29th October
Feedback from LSE on cognitive testing guide 4th November
Final cognitive testing guide 6th November
Briefing of interviewers 6th November
Recruitment in the UK 28th October 6th November
Fieldwork cognitive testing phase 1 7th November 16th November
Analysis and reporting 17th November 23rd November
Report sent to LSE 23rd November
Feedback from LSE 27th November
New version of questionnaire after 1st
phase of cognitive testing
27th November 4th December
Translation of questionnaire 4th December 18th December
Recruitment in 23 countries 4th Jan 2010 7th Jan 2010
Briefing of interviewers 4th Jan 2010 7th Jan 2010
Fieldwork cognitive testing phase 2 8th January 13th January
Analysis and reporting 14th January 21st January
Report sent to LSE 21st January
Feedback from LSE 25th January
New version of questionnaire after 2nd
phase of cognitive testing
26th January 28th January
CAPI Scripting 1.5 weeks 29th January 5th February
Pilot testing 3 weeks 5th February 1st March
Briefing of interviewers 5th February
Fieldwork pilot testing 6th February 22nd February
Pilot report 23rd February 1st March
Pilot report sent to LSE 1st March
Feedback from LSE on the pilot 8th March
Finalisation of the national questionnaires 2 weeks 9th March 18th March
Amendments to national questionnaires 9th March 15th March
4. Ethical questions arising from financial support/the provision of incentives
4.1 Are there any real or perceived conflicts of interest which could compromise the integrity and/or independence of the
research due to the nature of the funding body?
No, none
4.2 Have any incentives to the investigator been declared?
No, none apply
 
67
4.3 Are there any restrictions on the freedom of the investigator(s) to publish the results of the research?
No, none
4.4 Are any incentives being offered to participants?
No, none
5. Research Subjects
5.1 Who do you identify as the participants in the project? Are other people who are not participants likely to be directly
impacted by the project?
The participants will be children aged 9-16 who use the internet and one of their parents. Other people who are not
participants are not likely to be impacted by the project.
The decision to define the sample of children as those from 9 to 16 years old has been carefully taken. Ever younger
children are now accessing the internet – across the EU27, 75% of 6-17 year olds now uses the internet, this including
60% of 6-10 year olds.
Almost nothing is known of young children’s use, so it would be preferable to start with qualitative rather than
quantitative research methods for younger children. In a previous project, Children and their Changing Media
Environment, a 12 nation comparison conducted by Sonia Livingstone a decade ago, the youngest children surveyed
were 9 years old. This proved satisfactory in terms of the collection of reliable and valid data, though questions were
carefully pretested in terms of their comprehensibility and the appropriateness of response options provided.
Other researchers’ experience in this field concurs that interviews with those as young as nine are feasible (for
example, the SAFT - Safety Awareness Facts and Tools - project funded by the EC Safer Internet Programme as the
precursor of the present survey.
5.2 What arrangements have been made to preserve confidentiality for the participants or those potentially affected?
This has been addressed in detail in sections 3.1 and 3.2 above.
Confidentiality and anonymity will be guaranteed for participants in the survey, only limited in cases where a young
person makes a disclosure of risk or harm (see above and below). Participants’ names will not be recorded so it will not
be possible to link responses to individual children.
It may be that the questions will uncover a child possibly at risk. Such an eventuality must be anticipated when briefing
the interviewers and when obtaining informed consent from respondents. While generally confidentiality will be
preserved, in such cases specific actions appropriate to the circumstances would then be taken in line with the relevant
child protection policy of the country. The interviewers will be instructed to bring such situations to the attention of
their supervisor at the national fieldwork organisation who will then review the nature of the risks and options. If the
latter determines the risk is real, the appropriate agencies will be contacted.
The level and nature of any such contacts will be included in the full field work report to be submitted as part of its
work by Ipsos Mori to LSE.
Technical Report and User Guide: The 2010 EU Kids Online Survey
68
5.3. What are the specific risks to research participants or third parties?
We identify four possible risks, and have addressed these in the foregoing:
- The risk that the child will be distressed by sensitive questions – addressed in 3.1.
- The risk that the parent will find out the answers given by the child - addressed in 3.1.
- The risk that others will find out answers given by the parent and the child – addressed in 3.1 and 3.2.
- The situation where the child is ‘at risk’ – addressed in 3.1 and 5.2.
5.4. If the research involves pain, stress, physical or emotional risk, please detail the steps taken to minimize such effects?
Explain why this is reasonable within the context of the project?
Although we do not anticipate ‘unacceptable stress’, since we may uncover or occasion some stress, the following
procedures will be in place.
The interviewer will ensure that the child is genuinely happy to take part and that the child is entirely clear they don’t
have to answer any questions they don’t wish to answer and can end the interview at any time.
The interviewers will be trained to be very neutral and phase questions in a way that make the children feel comfortable.
They will reassure the child that the survey is informal, non-judgemental and that there are no right or wrong answers.
In their training organised by the national survey firm interviewers will be advised on the signs of any discomfort they
should be aware of (e.g. in terms of body language) when dealing interviewing the children and on how to cope with
any immediate distress shown by the child.
Interviewers’ experience and training
For a survey of this size, the quality of interviewing will be absolutely vital, and there is no substitute for interviewers
who are thoroughly experienced with this kind of work. Ipsos MORI is one of the most experienced organisations when
it comes to large scale social surveys, and we regard the experience of the field force used in each country to be as
critical as that of the executive teams.
Each fieldwork institute member of the Ipsos MORI network is committed to allocate to this project experienced
professional interviewers in opinion face to face interviewing, with a very minimum of six months experience. In most
cases, interviewers are considerably more experienced, usually at least one year and often over 10 years experience. In
addition, to general survey research experience, interviewers selected to conduct fieldwork will have particular skills in
conducting public opinion research among children.
New interviewers are hired after having successfully passed a strict selection procedure:
- Analysis of the applicant’s curriculum vitae.
- Face to face discussion with the fieldwork manager about the applicant’s professional background, motivation
and skills are carefully analysed.
- The interviewer’s skills are tested through a role play.
In addition to Ipsos MORI’s standard vigorous interviewer training, before an interviewer works on this project, they
will have to go through intensive project-specific training via a thorough combination of both written and classroom
based briefings, further details of which are outlined below.
Briefing of the interviewers
Ipsos MORI Coordination Centre will provide all national operators with detailed and uniform instructions for
conducting fieldwork. The Coordination Centre will prepare these instructions, with the assistance of the Quality
Control Committee.
In addition to these procedures, we will set up, for the attention of Project Managers in each country, a Training Book
which will comprise all instructions regarding the survey and instructions on how to brief interviewers. In addition,
individual project managers from each country will also receive an interactive telephone briefing. The aim of these
 
69
measures is to further ensure the uniformity of fieldforce training across all countries covered by the survey.
The briefing given to interviewers will cover the following main topics:
- Overall brief on EU Kids Online Survey:
o background
o purpose
o importance of the survey
o international dimension, ensuring the essential consistency of fieldwork across countries.
- Detailed description of the random route sampling procedures:
o Definition of the population to be sampled
o Concept of starting address, location on a map
o Focus on random-walk rules
o Child and parent selection: implementation of the ‘next birthday’ procedure
o Management of failed contacts: recall procedure (number and timing of visits), letter in mailbox when
relevant, etc.
o Sampling follow-up: review of contact sheets and how to use them
o Explanation of over-sampling when relevant
o Contact sheet procedures.
- Full questionnaire review:
o Overall structure of the questionnaire
o Review of the various topics
o Explanation of complex questions, concepts or words
o Detailed presentation of questionnaire routine and specifics: filters, split samples, show cards, etc.
- Briefings on key aspects of approach relevant to interviewing children, including consent, ethics, child
protection, and interviewing techniques
- Fieldwork management rules:
o Reminder of interviewing techniques: general behaviour and presentation, contact techniques to limit/avoid
refusals and maximise the response rate, interview flow, techniques to maintain respondents’ attention,
techniques for interviewing children and young people, etc.
o Handling of survey materials
o Survey schedule: fieldwork dates and hours
o Detailed and thorough reminders of the importance and procedures of reporting (requirements and how to
meet them): mode and frequency of contacts with the survey supervisor or manager, interim returns of
questionnaires and contact sheets, rules of replacement of interviews if quality controls reveal mistakes made,
mode and date of debriefing at the end of fieldwork.
- Specific techniques to convert refusals and maximise the response rate
- Review of ESOMAR ethical rules
- A reminder of how the quality of their work will be supervised and managed, including back-checking
procedures
In summary, in each country/territory, the following briefing methods will be used:
- Detailed briefings on paper as outlines above, detailing objectives, usage of show cards, specific backgrounds
per topic (if deemed necessary), using examples of completed questionnaires (if deemed necessary).
Interviewers will receive these written instructions in their Interviewing Pack.
- Local supervisors and interviewers attend face-to-face briefing sessions. These half-day or one-day sessions
are organised centrally or at regional level. These briefing sessions end with role plays where interviewers
work in pairs on the questionnaire.
- The country coordinator at the coordination centre will have a debriefing session over the phone with the
project managers and fieldwork supervisors to clarify any problem/question raised during the interviewers
Technical Report and User Guide: The 2010 EU Kids Online Survey
70
briefing.
- Continuous availability of the field management team and supervisors for whenever questions arise. A
dedicated phone line will be available to the interviewers.
6. Risk to researchers
6.1 Are there any risks to the researcher(s)? Please provide details if risk identified.
We do not foresee any risks for the interviewers. However, some cities/neighbourhoods are safer for male interviewers
than for women. In some areas, there may be a concentration of ethnic minorities who could be less inclined to let
someone from another community entering their homes.
In these cases, Ipsos Mori pays particular attention to allocating the right interviewer to the right area (e.g. try to match
the ethnic origin of the interviewer to that of the surveyed area).
In addition, in the interviewers’ briefing, all interviewers are reminded of elementary rules of behaviour such as
neutrality, respect, politeness. All stay in close contact with their supervisor and with the national field work agency
which monitors their quality of their work, including consideration of their personal safety.
7. Confidentiality
7.1 Explain the mechanisms in place to ensure confidentiality, privacy and data protection.
See 3.2
8. Dissemination
8.1 Will the results of the study be offered to those participants or other affected parties who wish to receive them? If so,
what steps have been taken to minimize any discomfort or misrepresentation that may result at the dissemination level.
The project is designed to inform multiple stakeholders, including children and parents as well as educators, awareness
raisers, child welfare workers, governments and industry.
The participants in the study will be offered access to the findings and resulting recommendations. Specifically, in the
LSE letter introducing the project to each household, the name a url will be provided as well as the date by which we
will post a family-friendly summary of the results (November 2010).
The leaflet to be left with all interviewees will include helpful safety information and further sources of information for
them in their country.
 
71
Information letter to parents
April 2010
Dear Parent
EU Kids Online survey
Thank you very much for agreeing to participate in our survey. At the London School of Economics we lead
this important project for the European Commission’s Safer Internet Programme.
We are working with university researchers in 25 different countries, plus international expert advisors who
make sure that the results will be useful for initiatives to make the internet safer for children. These
advisors include Save the Children, European Schoolnet, and a European network for safety awareness raising
(Insafe).
We have designed this survey for parents and children from all over Europe, and the findings will be
important for advising schools, child welfare, youth workers and others who work to enable children to get
the best out of the internet while minimising online risks.
For example, knowing what children do online can help teachers to devise cyberbullying programmes. It
will also help governments in deciding whether parts of the internet should be better regulated. Youth
workers and other professionals who work with children also need to know what to warn or advise children
about. And our work will also provide guidance for parents, so they can learn ways to help and support their
children when using the internet.
The survey also aims to get the risks faced by some children into perspective, by discovering the beneficial
things children do on the internet and the great ways children ar e learning to use the internet sensibly and
well. This is why our survey asks lots of questions – so that we can understand the different kinds of
experiences that children of different ages and backgrounds may have in different countries.
Information about the researchers and advisors in each country is available on our website at
www.eukidsonline.net. The findings will be reported by the European Commission on 21
st
October 2010 in
Luxembourg. We will post the findings on our website on that date – please visit the website if you would
like to know the results.
Again, many thanks for participating in this survey.
Yours sincerely,
Professor Sonia Livingstone
Director, EU Kids Online project
Department of Media and Communications
London School of Economics and Political Science
Houghton Street, London WC2A 2AE, UK
Telephone +44(0)2079557710 Email s.livingstone@lse.ac.uk
Technical Report and User Guide: The 2010 EU Kids Online Survey
72
Parental consent letter
LOGO’s – university and agency Date
Dear Parent/Guardian
Research to help make the internet safe for children and young people
I am writing to ask for your help with an important study that is being conducted with children aged 9-16 who use the
internet and their parents across <INSERT COUNTRY NAME> as well as in twenty three other countries across Europe.
The Independent research organisations Ipsos and <INSERT FIELD AGENCY NAME> are carrying out this research on
behalf of the London School of Economics, funded by the European Commission.
I would like to invite both you and your child to take part in an interview about your views and experiences of
your child’s use of the internet. Your household has been selected at random to take part in the research. The
questionnaire will ask about your own experiences of the internet and your child’s experiences – this will include
discussions about how often your child uses the internet, where they go online, how they spend time on the internet, and
their exposure to potentially harmful or inappropriate material and behaviour. The survey results will be used by
governments across Europe to help ensure that children are safe when they go online and to support parents in helping to
protect their children from online risks.
The interviews will be relaxed and informal and you and your child would be free to skip questions that you don’t feel
comfortable with, but whatever information you feel able to provide will really help the governments across Europe to
understand the risks that children currently face and how best they can work with parents to protect children.
Your survey answers would be treated in absolute confidence, in accordance with the Data Protection Act. Your name or
personal details will not be passed on to anyone outside the Ipsos/LOCAL AGENCY research team nor be identified in
any research findings. Once the research is complete, your responses will be anonymised, and your name and address
will be securely deleted from Ipsos’s/LOCAL AGENCY records.
The interviews would take place in your home at a time convenient for you. We would like to talk to your child for around
30 minutes and to you for around 10 minutes. Taking part is voluntary but we hope that you will take part so we can hear
the views of a range of people.
The interviewer will carry a photo identification card.
If you have any questions about the research or do not want to take part please call XXXX at Ipsos on XXXX or <LOCAL
AGENCY> who will be happy to answer any questions you might have. If you do get in touch, please remember to give
your name and the reference number at the top of this letter.
I do hope that you will be able to take part in this important survey.
Yours sincerely,
<NAME OF MANAGER>, Study Manager, <COUNTRY NAME>
 
73
Consent forms
This consent form is usually integrated into the contact sheet so that the interview completes a single form for each
household at the stage of initial contact.
Introduction
Good morning/afternoon/evening. My name is …… from Ipsos MORI, the independent research company.
I would like to ask your help with a survey we are carrying out among young people and their parents - the survey
is about young people using the internet safely. The questionnaire will ask about your own experiences of the
internet and your child’s experiences – this will include discussions about how often they use the internet, where
they go online, how they spend time on the internet, and their exposure to potentially harmful or inappropriate
material and behaviour, such as content that would normally be for adults. The survey results will be used by
governments across Europe to help ensure that children are safe when they go online and support parents in
helping to protect their children from online risks.
Your household has been selected completely at random from a list of addresses in this area. All information will
be treated in the strictest of confidence; the reporting of findings will not identify individuals or families and the
names of those who take part will not be passed on to anyone outside Ipsos MORI and <Local agency>, or used
for any purpose other than this research project. You do not have to answer any questions that you do not want
to and you can stop the interview at any point.
Screening
If more than one parent, select parent to take part (random method).
If more than one child, select child to take part (random method).
We would like to carry out an interview with you that will last 10 minutes and an interview with your child/one of
your children that will last around 30 minutes.
Parent consent
If necessary, repeat intro to parent to gain parent consent/participation:
Are you able to take part in this research?
Yes – would it be convenient to conduct the interview now (If not arrange appointment)?
No (close)
I would also like to conduct an interview with [selected child] are you happy for me to invite him/her to take part?
Yes (proceed to consent)
No (close)
Complete if consent given
Parent name ………………………………………………………………
Signature…………………………………………………………………..
Relationship to young person………………………………………….
Technical Report and User Guide: The 2010 EU Kids Online Survey
74
Intro for 9-12 year olds
Hello, my name is XXX and I am from Ipsos MORI, a company that asks people questions about lots of different
things.
We'd like to ask you what you think about using the internet and the types of things you do and see online
including things you have liked but also things that you have not liked. We are speaking to lots of other young
people like you, from across lots of different counties.
The findings will be used to help make the internet safer for young people to use. There aren't any right or wrong
answers, and nobody will know what you have said - we just want to find out what you think. If there's a question
you don't like, you don't have to answer it and you can stop the interview at any time. The only thing we would
have to tell someone about is if you said that you or someone else was being hurt, but we would talk to you about
that first, ok?
Would you be able to help us? It will take about 30 minutes.
Yes
No
Interviewer to sign that informed consent has been obtained
Signature…………………………………………………………………..
Intro for 13-16 year olds
Hello, my name is XXX and I am from Ipsos MORI, the research company (we find out what people think about
things using questionnaires and surveys). We'd like to ask you what you think about using the internet and the
types of things you do and see online including things you have liked but also things that you have not liked.
The research is being carried out across Europe and the findings will be used help make the internet safer for
young people to use.
There aren't any right or wrong answers, and nobody will know what you have said - we just want to find out what
you think. If there's a question you don't like, you don't have to answer it and you can stop the interview at any
time. The only thing we would have to tell someone about is if you said that you or someone else was being hurt,
but we would talk to you about that first, ok?
Would you be able to help us? It will take about 30 minutes.
Yes
No
Interviewer to sign that informed consent has been obtained
Signature…………………………………………………………………..
 
75
Information leaflet to be given to the child at the end of an interview
Technical Report and User Guide: The 2010 EU Kids Online Survey
76
 
77
ANNEX 4: TRANSLATION
OF DIFFICULT WORDS
Academic representatives in every country in the EU Kids Online network also reviewed translations to double
check that the meaning of key terms was as intended. In particular, a list of concepts for which there were
challenges ensuring translation generated identical meaning across countries was drawn up (“upset” is one
example) and network members input to ensure the most comparable terminologies were used. The list of these
concepts can be found below for each country.
Austria
Concept TRANSLATION
Bothered beunruhigt
Upset beschäftigt
Social worker Sozialarbeiter
Adviser Berater
Try to get back at the other person Ich habe versucht, mich an der anderen Person zu rächen
Privacy settings Einstellungen für die Privatsphäre
Contact settings Einstellungen für die Kontakte
face to face persönlich
Sexual image Bilder mit sexuellem Inhalt
Sexual message Nachrichten mit sexuellem Inhalt
Sexual act Geschlechtsverkehr
An adult/X-rated website Eine Seite für Erwachsene
Peer to peer file-sharing Auf einer Seite, wo Daten mit anderen Personen geteilt werden
Private parts Geschlechtsteile
Social networking site sozialen Netzwerk-Seite
Instant messaging Sofortnachrichtendienst
Chatroom Chatroom
Gaming website Spiele-Seite
Pop-up Fenster, das auf einmal aufgegangen ist (per Zufall)
Desktop computer Computer am Schreibtisch
Virtual world virtuelle Welt
Filter preferences Filtereinstellungen
Parental controls Kindersicherung
Spam/junkmail Spam
Technical Report and User Guide: The 2010 EU Kids Online Survey
78
Belgium - French
Concept TRANSLATION
Bothered Tracassé
Upset Perturbé
Social worker Travailleur social
Adviser conseiller
Try to get back at the other person essayer de se venger de l'autre personne
Privacy settings paramètres de confidentialié
Contact settings coordonnées
face to face Face à face
Sexual image image à caractère sexuel
Sexual message Message à caractère sexuel
Sexual act relation sexuelle
An adult/X-rated website Site pornographique
Peer to peer file-sharing Site d'échange de fichier
Private parts Sexe
Social networking site Site de réseau social
Instant messaging Messagerie instantanée
Chatroom Chat
Gaming website Site de jeux
Pop-up Fenêtre qui s'ouvre sur l'écran
Desktop computer ordinateur de bureau
Virtual world monde virtuel
Filter preferences Filtres de préférence
Parental controls Contrôle parental
Spam/junkmail spam/mail indésirable
Belgium - Flemish
Concept TRANSLATION
Bothered Je zorgen maken
Upset Geschokt zijn
Social worker Sociaal werker
Adviser hulplijn voor kinderen
Try to get back at the other person Proberen het de andere persoon betaald te zetten
Privacy settings Privacy instellingen
Contact settings
face to face persoonlijk
Sexual image sexueel getint beeld
Sexual message sexueel getinte boodschap
Sexual act sexuele handeling
An adult/X-rated website een site voor volwassenen/niet geschikt voor kinderen
Peer to peer file-sharing bestanden die je deelt met andere internet gebruikers
Private parts intieme lichaamsdelen
Social networking site site waar je een sociaal netwerkprofiel hebt
Instant messaging instant messaging (MSN, Windows Live Messenger,…)
Chatroom Chatroom
Gaming website spelletjeswebsite
Pop-up pop-ups (kleine venstertjes de opeens op je scherm verschijnen)
Desktop computer PC
Virtual world virtuele wereld
Filter preferences Instellingen veranderen
Parental controls ouderlijke contrôle
Spam/junkmail ongewenste e-mail (spam)
 
79
Bulgaria
Concept TRANSLATION
Bothered Притеснен(а)
Upset Разстроен(а)
Social worker Социален работник
Adviser Съветник
Try to get back at the other person Опитах се да си го върна на другия човек
Privacy settings Настройки за поверителност и защита
Contact settings Настройки за поверителност и защита
face to face Лице в лице
Sexual image Сексуално изображение
Sexual message Сексуално съобщение
Sexual act Правене на секс
An adult/X-rated website Забранен за под 18 г. сайт
Peer to peer file-sharing Торент сайт
Private parts Интимни части на тялото
Social networking site Онлайн социална мрежа
Instant messaging Програма за разговори в реално време
Chatroom Чат-рум
Gaming website Геймърски сайт
Pop-up Поп-ъпс (нещо, което се появява случайно)
Desktop computer РС (настолен компютър)
Virtual world Виртуален свят
Filter preferences Предпочитания за филтриране
Parental controls Родителски контрол
Spam/junkmail Нежелана поща/спам
Cyprus
Concept TRANSLATION
Bothered Ενοχλημένος
Upset Αναστατωμένος
Social worker Κοινωνική λειτουργός
Adviser Σύμβουλος
Try to get back at the other person Προσπαθώ να εκδικηθώ
Privacy settings ρυθμίσεις ασφαλείας
Contact settings ρυθμίσεις επαφών
face to face πρόσωπο με πρόσωπο
Sexual image εικόνα σεξουαλικού περιεχομένου
Sexual message Μήνυμα σεξουαλικού περιεχομένου
Sexual act σεξουαλική συνεύρεση
An adult/X-rated website site (ιστοσελίδα) ενηλίκων
Peer to peer file-sharing site ανταλλαγής αρχείων
Private parts γεννητικά όργανα
Social networking site site (ιστοσελίδα) κοινωνικής δικτύωσης
Instant messaging στιγμιαίο μήνυμα
Chatroom Chatroom (ηλεκτρονικό δωμάτιο συζητήσεων)
Gaming website ιστοσελίδα παιχνιδιών
Pop-up εικόνες που εμφανίζονται ξαφνικά (pop up)
Desktop computer Υπολογιστής (επιτραπέζιος)
Virtual world εικονικός κόσμος
Filter preferences ρυθμίσεις φιλτραρίσματος
Parental controls Γονικός έλεγχος
Spam/junkmail ενοχλητική αλληλογραφία
Technical Report and User Guide: The 2010 EU Kids Online Survey
80
Czech Republic
Concept TRANSLATION
Bothered rozrušený
Upset rozhozený
Social worker sociální pracovník
Adviser linka bezpečí
Try to get back at the other person snažit se pomstít
Privacy settings nastavení soukromí
Contact settings kontaktní údaje
face to face osobně, tváří v tvář
Sexual image něco se sexuální tematikou
Sexual message zpráva se sexuální tematikou
Sexual act sex
An adult/X-rated website stránky pro dospělé přístupné od 18 let
Peer to peer file-sharing stránky pro sdílení souborů
Private parts intimní partie
Social networking site stránka sociální sítě
Instant messaging komunikační aplikace
Chatroom chatovací místnost
Gaming website stránky pro hráče
Pop-up pop-up webové okno (okno, které se objeví samo)
Desktop computer stolní počítač
Virtual world virtuální svět
Filter preferences nastavení filtru
Parental controls rodičovská kontrola
Spam/junkmail spam
Denmark
Concept TRANSLATION
Bothered Bekymret over/føles sig generet af
Upset Chokeret eller rystet
Social worker Socialrådgiver
Adviser Rådgiver
Try to get back at the other person Forsøger at hævne sig på den anden person
Privacy settings Personlige indstillinger
Contact settings Kontaktoplysninger
face to face Ansigt til ansigt (personligt)
Sexual image Seksuelle billeder
Sexual message Seksuelle beskeder
Sexual act Gøre noget seksuelt
An adult/X-rated website Hjemmesider kun for voksne
Peer to peer file-sharing Ven-til-ven fildeling (f.eks. Limewire)
Private parts Kønsdele
Social networking site Sociale netværkssteder
Instant messaging Messenger/MSM
Chatroom Chatroom
Gaming website Hjemmeside med spil
Pop-up Pop-up vindue (noget, der vises tilfældigt)
Desktop computer PC (stationær PC)
Virtual world Virtuel verden
Filter preferences Foretrukne filterindstillinger
Parental controls Forældrekontrol
Spam/junkmail Spam
 
81
Estonia - Estonian
Concept TRANSLATION
Bothered häiritud olema
Upset endast väljas olemine
Social worker sotsiaaltöötaja
Adviser nõustaja
Try to get back at the other person Teisele inimesele samaga vastata proovima/ tagasi teha
Privacy settings privaatsusseaded
Contact settings kontaktandmed
face to face silmast silma
Sexual image seksipilt
Sexual message Seksisõnum
Sexual act seksakt
An adult/X-rated website täiskasvanute veebileht
Peer to peer file-sharing isikult isikule faili jagamine
Private parts initiimsed kehaosad
Social networking site suhtlusportaal
Instant messaging MSN, messenger
Chatroom jututuba
Gaming website mängulehekülg
Pop-up hüpikaken
Desktop computer lauaarvuti
Virtual world virtuaalmaailm
Filter preferences filtrieelistused
Parental controls vanemakontroll
Spam/junkmail spämm/rämpsmail
Estonia - Russian
Concept TRANSLATION
Bothered быть обеспокоенным, обескураженным
Upset расстроиться, огорчиться
Social worker социальный работник
Adviser консультант, советчик
Try to get back at the other person ответить другогму человеку тем же, отомстить
Privacy settings Настройки безопасности
Contact settings Контактные данные
face to face с глазу на глаз
Sexual image изображение сексуального характера
Sexual message сообщение сексуального характера
Sexual act половой акт
An adult/X-rated website вебсайт только для взрослых, Х-вебсайт
Peer to peer file-sharing обмен файлами между пользователями
Private parts интимные части тела
Social networking site социальная сеть
Instant messaging MSN, мессенджер
Chatroom чат
Gaming website игровой сайт
Pop-up всплывающее окно
Desktop computer настольный (стационарный) компьютер
Virtual world виртуальный мир
Filter preferences настройки фильтра
Parental controls родительский контроль
Spam/junkmail спам
Technical Report and User Guide: The 2010 EU Kids Online Survey
82
France
Concept TRANSLATION
Bothered Tracassé
Upset Perturbé
Social worker Travailleur social
Adviser conseiller
Try to get back at the other person essayer de se venger de l'autre personne
Privacy settings paramètres de confidentialié
Contact settings coordonnées
face to face Face à face
Sexual image image à caractère sexuel
Sexual message message à caractère sexuel
Sexual act relation sexuelle
An adult/X-rated website
site web classé X/
Pornographique
Peer to peer file-sharing Site d'échange de fichier
Private parts parties intimes /sexe
Social networking site Site de réseau social
Instant messaging Messagerie instantanée
Chatroom Chat
Gaming website Site de jeux
Pop-up Une fenêtre qui s’est ouverte sur l’écran sans que tu le veuilles
Desktop computer ordinateur de bureau
Virtual world monde virtuel
Filter preferences Filtres de préférence
Parental controls Contrôle parental
Spam/junkmail Spam/courrier indésirable
Finland
Concept TRANSLATION
Bothered vaivata / vaivaantunut
Upset järkyttää / järkyttynyt
Social worker sosiaalityöntekijä
Adviser nuorisoneuvoja
Try to get back at the other person Yritin kostaa tälle henkilölle
Privacy settings yksityisyysasetukset
Contact settings yhteydenottoasetukset
face to face kasvokkain
Sexual image seksuaalinen kuva
Sexual message seksuaalinen viesti
Sexual act seksuaalisten asioiden tekeminen
An adult/X-rated website aikuisten/lapsilta kielletty sivusto
Peer to peer file-sharing vertaisverkon tiedostojen jako
Private parts intiimit alueet
Social networking site verkkoyhteisö
Instant messaging pikaviesti
Chatroom chat-huone
Gaming website pelisivusto
Pop-up pop-up/ponnahdusikkuna
Desktop computer pöytäkone
Virtual world virtuaalimaailma
Filter preferences filtteri-/estoasetukset
Parental controls suodatinohjelma /lapsilukko-ohjelma
Spam/junkmail roskaposti
 
83
Germany
Concept TRANSLATION
Bothered beunruhigt
Upset unangenehm berührt
Social worker Sozialarbeiter
Adviser Betreuer
Try to get back at the other person Habe versucht, mich an der anderen Person zu rächen
Privacy settings Privatsphäre-Einstellungen
Contact settings Kontaktdaten-Einstellungen
face to face "persönlich" or "von Angesicht zu Angesicht"
Sexual image Bilder sexueller Art
Sexual message Nachrichten sexueller Art
Sexual act sexuelle Dinge/Handlungen
An adult/X-rated website Internetseite für Erwachsene / nicht jugendfreien Internetseite
Peer to peer file-sharing Peer-to-peer-Netzwerken oder Tauschbörsen (z.B. RapidShare)
Private parts Geschlechtsteile
Social networking site Soziales Netzwerk
Instant messaging Instant messaging
Chatroom Chatroom
Gaming website Spiele Webseite
Pop-up Pop-up
Desktop computer PC (Festinstallierter PC)
Virtual world Zeit in einem virtuellen Raum verbracht (z.B. Second Life, SIMS usw.)
Filter preferences Filtereinstellungen
Parental controls Elterliche Kontrollen
Spam/junkmail Spam oder Junk-Mail
Greece
Concept TRANSLATION
Bothered ενόχλησε/ αναστάτωσε/ απασχόλησε
Upset στεναχώρησε/ ανησύχησε
Social worker κοινωνικός λειτουργός
Adviser σύμβουλος
Try to get back at the other person προσπάθησε να εκδικηθεί το άλλο άτομο
Privacy settings ρυθμίσεις ιδιωτικότητας
Contact settings στοιχεία επικοινωνίας
face to face πρόσωπο με πρόσωπο
Sexual image εικόνες με σεξουαλικό περιεχόμενο
Sexual message μήνυμα με σεξουαλικό περιεχόμενο
Sexual act σεξουαλική πράξη
An adult/X-rated website ιστοσελίδες που είναι μόνο για ενήλικες
Peer to peer file-sharing προγραμμάτων ανταλλαγής αρχείων από υπολογιστή σε υπολογιστή
Private parts γεννητικά όργανα
Social networking site ιστοσελίδα κοινωνικής δικτύωσης
Instant messaging άμεσα μηνύματα
Chatroom δωμάτια επικοινωνίας
Gaming website ιστοσελίδας με διαδικυακά παιχνιδιών
Pop-up Από εικόνες που εμφανίζονται ξαφνικά στην οθόνη
Desktop computer σταθερό/ προσωπικό υπολογιστή
Virtual world εικονικός κόσμος
Filter preferences επιλογές φίλτρων
Parental controls γονικός έλεγχος
Spam/junkmail ανεπιθύμητες διαφημίσεις ή ανεπιθύμητη αλληλογραφία
Technical Report and User Guide: The 2010 EU Kids Online Survey
84
Hungary
Concept TRANSLATION
Bothered felzaklat, zavar, valami rossz történik
Upset felzaklat, zavar
Social worker családsegítő
Adviser tanácsadó szakember
Try to get back at the other person bosszút áll
Privacy settings személyes biztonsági beállítások
Contact settings személyes biztonsági beállítások
face to face személyesen
Sexual image szexuális tartalmú felvételek, képek
Sexual message szexuális tartalmú üzenetek
Sexual act szex
An adult/X-rated website korhatáros (felnőtteknek szóló) honlap
Peer to peer file-sharing peer-to-peer fájlmegosztó
Private parts nemi szervek
Social networking site közösségi oldal
Instant messaging üzenetküldő program (msn)
Chatroom chat
Gaming website játék oldal
Pop-up felugró ablak
Desktop computer asztali számítógép (PC)
Virtual world virtuális világ
Filter preferences családsegítő
Parental controls tanácsadó szakember
Spam/junkmail bosszút áll
Italy
Concept TRANSLATION
Bothered Infastidito
Upset turbato
Social worker assistente sociale
Adviser tutor (9-10)/educatori (11-16)
Try to get back at the other person Cercare di vendicarsi dell'altra persona
Privacy settings Impostazioni sulla privacy
Contact settings Impostazioni del mio contatto
face to face Faccia a faccia
Sexual image Immagine a sfondo sessuale
Sexual message Messaggio a sfondo sessuale
Sexual act Cose a sfondo sessuale/activita sessuale
An adult/X-rated website Sito per adulti/ vietato ai minori
Peer to peer file-sharing Programma di condivisione di file
Private parts Parti intime
Social networking site (Sito di) social network
Instant messaging messaggi istantanei
Chatroom Chat
Gaming website Sito di giochi
Pop-up Pop-up
Desktop computer Computer da tavolo
Virtual world Mondo virtuale
Filter preferences Cambiare le preferenze dei filtri
Parental controls Controllo genitori/ parental control
Spam/junkmail spam
 
85
Lithuania
Concept TRANSLATION
Bothered
Sunerimęs (QA6; QF12); sukėlė nerimą (QA7);
sutrikti/sutrikdyti (QD8; QD9, QD15, QF11, QF21, QG5)
Upset Nuliūsti
Social worker Socialinis darbuotojas
Adviser
Try to get back at the other person Mėginau atsilyginti tuo pačiu tam asmeniui
Privacy settings slaptumo nustatymai
Contact settings kontaktiniai duomenys
face to face tiesioginis bendravimas
Sexual image seksualinio turinio atvaizdas
Sexual message seksualinio turinio žinutė
Sexual act
seksualiniai dalykai (QG3 C), užsiiminėti seksu (QG3 e),
seksualiniai veiksmai (QH3)
An adult/X-rated website Suaugusiems skirtas puslapis
Peer to peer file-sharing
Per P2P (peer to peer) keitimąsi duomenimis
(pvz., „Torrent“, „Linkomanija“, RC)
Private parts Intymios kūno dalys or intymios kūno vietos in different questions
Social networking site Socialinis tinklas
Instant messaging
Naudotis tiesioginio susirašinėjimo programomis
(pvz. Skype, Google talk, MSN...)
Chatroom Pokalbių svetainė (pvz. Chat.lt, zebra.lt)
Gaming website Žaidimų puslapis
Pop-up Iškylantys reklaminiai langai (pop – ups)(kurie kartais netikėtai iškyla)
Desktop computer stalinis kompiuteris
Virtual world virtualus/virtualusis pasaulis
Filter preferences Filtrų nustatymai
Parental controls Tėvų kontrolė
Spam/junkmail
Brukalas (SPAMas)-decided to leave SPAM as it is called like that pretty much
offten than "brukalas", junkmail-nepageidaujama reklama
Netherlands
Concept TRANSLATION
Bothered van streek zijn
Upset van slag, geschrokken
Social worker maatschappelijk werker
Adviser adviseur
Try to get back at the other person iemand terugpakken
Privacy settings instellingen voor mijn privacy
Contact settings contact gegevens
face to face persoonlijk
Sexual image seksuele foto/ plaatje of video
Sexual message seksueel bericht
Sexual act seksuele handeling
An adult/X-rated website niet geschikt voor minderjarigen
Peer to peer file-sharing file sharing sites gebruikt (peer to peer)
Private parts intieme lichaamsdelen
Social networking site sociale netwerk site
Instant messaging instant messaging (MSN)
Chatroom Chatroom
Gaming website spelletjes website
Pop-up pop-up
Desktop computer computer
Virtual world virtuele wereld
Filter preferences filter voorkeuren
Parental controls not translated litterally
Spam/junkmail Spam
Technical Report and User Guide: The 2010 EU Kids Online Survey
86
Norway
Concept TRANSLATION
Bothered plaget
Upset lei seg
Social worker sosialarbeider
Adviser rådgiver (skolerådgiver)
Try to get back at the other person Prøv å komme tilbake til den andre personen
Privacy settings Personvern innstillinger
Contact settings kontakt innstillinger
face to face ansikt til ansikt
Sexual image seksuelt bilde
Sexual message seksuell melding
Sexual act seksuell handling
An adult/X-rated website pronoside
Peer to peer file-sharing fildeling mellom datamaskiner
Private parts kjønnsorganer
Social networking site sosialt nettverksted
Instant messaging Direktemeldinger
Chatroom Chattested/ pratested
Gaming website Nettside for dataspill
Pop-up pop-up
Desktop computer Skrivebord på datamaskin/ bord datamaskin
Virtual world virtuell verden
Filter preferences filter innstillinger
Parental controls Foreldrekontroll
Spam/junkmail Spam/søppelmail
Poland
Concept TRANSLATION
Bothered zaniepokojony
Upset przejąć się czymś, w innym miejscu: zrobiło Ci się nieprzyjemnie
Social worker pracownik socjalny
Adviser doradca
Try to get back at the other person zemścić się, odegrać na tej osobie
Privacy settings ustawienia prywatności
Contact settings ustawienia kontaktów
face to face twarzą w twarz
Sexual image obraz, zdjęcie lub film związany z seksem
Sexual message wiadomość związana z seksem
Sexual act czynność seksualna
An adult/X-rated website strona przeznaczona dla dorosłych
Peer to peer file-sharing portale umożliwiające dzielenie się plikami (tzw. peer-to-peer)
Private parts intymne części ciała
Social networking site portal społecznościowy
Instant messaging komunikator
Chatroom czat (chatroom)
Gaming website strona z grą/grami
Pop-up wyskakujące okienko (pop-up)
Desktop computer komputer stacjonarny
Virtual world świat wirtualny
Filter preferences ustawienia filtrów
Parental controls programy kontroli rodzicielskiej
Spam/junkmail spam/niechciane wiadomości
 
87
Portugal
Concept TRANSLATION
Bothered Incomodado
Upset Perturbado, Chateado;
Social worker Assistente Social
Adviser Conselheiro
Try to get back at the other person Vingar-se
Privacy settings Definições de Privacidade
Contact settings Definições de Contactos
face to face Cara-a-Cara
Sexual image Imagem de teor sexual/ imagem sexual
Sexual message Mensagem de teor sexual / mensagem sexual
Sexual act Acto Sexual / Fazer sexo
An adult/X-rated website Website Conteúdos para Adultos
Peer to peer file-sharing Partilha de Ficheiros PtP
Private parts Zonas intimas / partes intimas
Social networking site Site de Rede Social
Instant messaging Mensagens Instantâneas
Chatroom Sala de Chat
Gaming website Jogos Online
Pop-up Janelas Pop-Up
Desktop computer Computador de secretária
Virtual world Mundo Virtual
Filter preferences Preferências de Filtragem
Parental controls Controlo Parental
Spam/junkmail Correio Electrónico Não Solicitado/Lixo Electrónico
Romania
Concept TRANSLATION
Bothered Deranjat
Upset Supărat
Social worker Asistent social
Adviser Persoană a cărei ocupaţie este să dea sfaturi
Try to get back at the other person A încerca să te răzbuni pe persoana respectivă
Privacy settings Setări de protecţie a identităţii
Contact settings Setări de contact
face to face faţă în faţă
Sexual image Imagine cu conţinut sexual
Sexual message Mesaj cu conţinut sexual
Sexual act Act sexual
An adult/X-rated website Un site pentru adulţi
Peer to peer file-sharing Site-uri de "share-uit" fişiere , adica puse la comun (dc++, odc, torrente)
Private parts Părţi intime
Social networking site Reţea socială
Instant messaging Messenger
Chatroom Cameră de chat
Gaming website Site de jocuri
Pop-up Pop-up
Desktop computer Calculator (desktop)
Virtual world Lume viruală
Filter preferences Preferinţe de filtrare
Parental controls Control parental
Spam/junkmail Spam
Technical Report and User Guide: The 2010 EU Kids Online Survey
88
Slovenia
Concept TRANSLATION
Bothered vznemirjen
Upset razburiti
Social worker socialni delavec
Adviser svetovalec
Try to get back at the other person maščevati se
Privacy settings nastavitve zasebnosti
Contact settings kontaktne informacije
face to face osebno, v živo
Sexual image podoba s spolno vsebino
Sexual message sporočila s spolno vsebino
Sexual act spolni odnos, spolno početje
An adult/X-rated website vsebine za odrasle
Peer to peer file-sharing stran za izmenjavo dokumentov
Private parts spolovila, intimni deli
Social networking site spletna stran za socialno mreženje
Instant messaging takojšnje sporočanje
Chatroom klepetalnica
Gaming website spletna stran z igrami
Pop-up pop-up okno, nekar kar se pojavi samo od sebe
Desktop computer namizni računalnik
Virtual world virtualni svet
Filter preferences lastnosti filtrov
Parental controls starševski nadzor
Spam/junkmail nezaželjena pošta, spam
Spain –Castilian
Concept TRANSLATION
Bothered Molestar
Upset Disgustar
Social worker Asistente social
Adviser Asesor
Try to get back at the other person Reaccionar en contra de otra persona
Privacy settings Condicones de privacidad
Contact settings Condiciones de contacto
face to face Cara a cara
Sexual image Imágenes de contenido sexual
Sexual message Mensajes de contenido sexual
Sexual act Acto sexual
An adult/X-rated website Página calificada como X / para adultos
Peer to peer file-sharing Redes P2P para compartir archivos
Private parts Partes intimas
Social networking site Red social
Instant messaging Mensajería instantánea - Messenger
Chatroom Chat
Gaming website Página de juegos
Pop-up Pop up o ventana emergente
Desktop computer Ordenador de sobremesa
Virtual world Mundo virtual
Filter preferences Preferencias de filtrado
Parental controls Controles paternales
Spam/junkmail E-mail spam o no deseado
 
89
Spain - Catalan
Concept TRANSLATION
Bothered Molestar
Upset Disgustar
Social worker Assistent social
Adviser Assessor
Try to get back at the other person Reaccionar en contra d’una altra persona
Privacy settings Condicions de privacitat
Contact settings Condicions de contacte
face to face Cara a cara
Sexual image Imatges de contingut sexual
Sexual message Missatges de contingut sexual
Sexual act Acte sexual
An adult/X-rated website Pàgina qualificada com X/ per a adults
Peer to peer file-sharing Xarxes P2P per a compartir arxius
Private parts Parts íntimes
Social networking site Xarxa social
Instant messaging Missatgeria instantània - Messenger
Chatroom Xat
Gaming website Pàgina de jocs
Pop-up Pop up o finestra emergent
Desktop computer Ordinador de sobretaula
Virtual world Món virtual
Filter preferences Preferències de filtratge
Parental controls Controls paternals
Spam/junkmail E-mail spam o no desitjat
Sweden
Concept TRANSLATION
Bothered Oroad
Upset Upprörd
Social worker Socialarbetare
Adviser Rådgivare
Try to get back at the other person Försöka ge igen
Privacy settings Sekretessinställningar
Contact settings Kontaktinställningar
face to face öga mot öga
Sexual image Erotisk bild
Sexual message Erotiskt meddelande
Sexual act Sex
An adult/X-rated website En barnförbjuden webbplats
Peer to peer file-sharing Fildelning
Private parts Könsdelar
Social networking site Hemsidor för socialt nätverkande", t.ex. Hamsterpaj eller Facebook
Instant messaging Snabbmeddelanden, chattmeddelande
Chatroom Chattrum
Gaming website Spelwebbplats
Pop-up Poppuppfönster
Desktop computer Stationär dator
Virtual world Virtuell värld
Filter preferences Filterinställningar
Parental controls Spärrfunktion
Spam/junkmail Skräppost
Technical Report and User Guide: The 2010 EU Kids Online Survey
90
Turkish
Concept TRANSLATION
Bothered Rahatsız etmek
Upset Üzücü
Social worker sosyal görevli
Adviser danışman
Try to get back at the other person Diğer kişiden öç almak
Privacy settings Gizlilik ayarları
Contact settings İletişim bilgileri
face to face Yüz yüze
Sexual image Cinsel içerikli resim
Sexual message Cinsel içerikli mesaj
Sexual act Cinsel içerikli davranış
An adult/X-rated website Yetişkinlere yönelik site
Peer to peer file-sharing Dosya paylaşım sitesi aracılığıyla (örn. Kazaa, Limewire, Rapidshare)
Private parts Vücuttaki mahrem /ayıp yerler
Social networking site Sosyal paylaşım sitesi (Facebook gibi)
Instant messaging Hızlı/anlık ileti (MSN gibi)
Chatroom Sohbet odası
Gaming website Oyun sitesi
Pop-up Kazara açılan pencereler (Pop-ups)
Desktop computer Masaüstü bilgisayarı
Virtual world Sanal dünya
Filter preferences Filtre seçenekleri
Parental controls Aile kontrolü
Spam/junkmail İstenmeyen reklam ya da e-posta, spam
 
91
ANNEX 5: KEY
VARIABLES
Use and activities
Concept Questions / Response options Summaries /
variable names
At school or college
Living room (or other public room) at home
At a friend's home
Own bedroom (or other private room) at home
At a relative's home
In an internet café
In a public library or other public place
Number of places
where the internet is
used
When 'out and about'
The number out
of eight response
options/
DPplaceNM
Shared PC
Own PC
Television set
Mobile phone
Games console
Own laptop
Shared laptop
Number of devices
used to access the
internet
Other handheld or portable device (e.g. iPod Touch, iPhone or Blackberry)
The number out
of eight response
options/
DPdeviceNM
Estimated minutes
online each day
About how long do you spend using the internet on a normal school day / normal
non-school day?
DCtimeuse
Technical Report and User Guide: The 2010 EU Kids Online Survey
92
Digital literacy
Concept Questions / Response options Summaries /
variable names
Bookmark a website
Block messages from someone you don’t want to hear from
Find information on how to use the internet safely
Change privacy settings on a social networking profile
Compare different websites to decide if information is true
Delete the record of which sites you have visited
Block unwanted adverts or junk mail/spam
Digital skills
Change filter preferences
The number out
of eight response
options/
DPskillsNM
Used the internet for school work
Played internet games on your own or against the computer
Watched video clips
Visited a social networking profile
Used instant messaging
Sent/received email
Read/watched the news on the internet
Played games with other people on the internet
Downloaded music or films
Put (or posted) photos, videos or music to share with others
Used a webcam
Put (or posted) a message on a website
Visited a chatroom
Used file sharing sites
Created a character, pet or avatar
Spent time in a virtual world
Range of online
activities
Written a blog or online diary
The number out
of 17 response
options/
DCactNM
Belief about internet
abilities
I know lots of things about using the internet. 1 (not true) to 3 (very true) DCwebableB
 
93
Risky activities
Concept Questions / Response options Summaries /
variable names
Missed school lessons without my parents knowing
Been in trouble with my teachers for bad behaviour
Been in trouble with the police
Had so much alcohol that I got really drunk
(only asked of children aged 11+)
Risky offline
activities
(adapted from the
Health Behaviour in
School-aged
Children survey;
Currie et al., 2008) Had sexual intercourse (only asked of children aged 11+)
The number out
of three response
options for 9-10
year olds and out
of five response
options for
children aged
11+ /
DCROB1NM
DCROB2NM
Looked for new friends on the internet
Added people to my friends list or address book that I have never met face-to-
face
Pretended to be a different kind of person on the internet from what I really am
Sent personal information to someone that I have never met face-to-face
Risky online
activities
(adapted from the
UK Children Go
Online survey;
Livingstone &
Helsper, 2010). Sent a photo or video of myself to someone that I have never met face-to-face
The number out
of five response
options/
DCriskactNM
Online risks
Concept Questions / Response options Summaries /
variable names
Online contacts
Online contacts Can I just check, have you ever had contact on the internet with someone you
have not met face to face before? yes/no
QC147
Meeting online
contacts offline
And have you ever gone on to meet anyone face to face that you first met on the
internet in this way? yes/no
QC148
Number of online
contacts met offline
And how many new people have you met in this way in the last 12 months, if
any? 1 to 2, 3 to 4, More than 10
QC149
Seeing and receiving sexual messages
Receiving sexual
messages
In the PAST 12 MONTHS, have you seen or received sexual messages of any
kind on the internet? yes/no
QC167
Frequency of
receiving sexual
messages
How often have you seen or received sexual messages of any kind on the
internet in the PAST 12 months?
Every day or almost every day
Once or twice a week
Once or twice a month
Less often
QC168
I have been sent a sexual message on the internet,
I have seen a sexual message posted where other people could see it on the
internet,
I have seen other people perform sexual acts,
I have been asked to talk about sexual acts with someone on the internet,
Types of sexual
messages received
The number out of
five response
options
I have been asked on the internet for a photo or video showing my private parts
The number out
of five response
options/
QC169A-E
Technical Report and User Guide: The 2010 EU Kids Online Survey
94
Sexual images
Seeing sexual
images
Have you seen these kinds of things [images that are obviously sexual] on any
websites in the past 12 months? yes/no
QC131
Types of sexual
images
Which types of website have you seen things like this [ANY KIND OF SEXUAL
IMAGES] on in the LAST 12 MONTHS?
The number out
of five response
options/
QC133A-E
Images or video of someone naked,
Images or video of someone's 'private parts',
Images or video of someone having sex,
Images or video of movies that show sex in a violent way,
Something else
Bullying
BULLYING
(introduction)
Sometimes children or teenagers say or do hurtful or nasty things to someone and this can
often be quite a few times on different days over a period of time, for example. This can
include:
• teasing someone in a way this person does not like
• hitting, kicking or pushing someone around
• leaving someone out of things
When people are hurtful or nasty to someone in this way, it can happen:
• face to face (in person)
• by mobile phones (texts, calls, video clips)
• on the internet (e-mail, instant messaging, social networking, chatrooms)
Cyberbullying (victim of)…
Being cyberbullied Has someone acted in this kind of hurtful or nasty way to you in the past 12
months? At any time during the last 12 months, has this happened...By mobile
phone calls, texts or image/video texts? yes/no [AND/OR] At any time during the
last 12 months, has this happened on the internet? yes/no
QC114B and/or
QC115
Online bullying (victim of)…
Being bullied online Has someone acted in this kind of hurtful or nasty way to you in the past 12
months? At any time during the last 12 months, has this happened on the
internet? yes/no
QC115
Types of being
bullied online
And in which ways has this [SOMEONE HAS DONE NASTY OR HURTFUL
THINGS TO YOU ON THE INTERNET] happened to you in the LAST 12
MONTHS?
The number out
of five response
options/
QC117A-E
Nasty or hurtful messages were sent to me,
Nasty or hurtful messages about me were passed around or posted
where others could see,
I was left out or excluded from a group or activity on the internet,
I was threatened on the internet,
Other nasty or hurtful things on the internet
 
95
Hate messages that attack certain groups or individuals
Ways to be very thin (such as being anorexic or bulimic)
Ways of physically harming or hurting themselves
Talk about or share their experiences of taking drugs
Number of items
reflecting negative
user generated
content
Ways of committing suicide
The number out
of five response
options/
DC142NM
Somebody used my password to access my information or to pretend to be me
Somebody used my personal information in a way I didn't like
Number of items
reflecting data
misuse
I lost money by being cheated on the internet
The number out
of three response
options/
DC143NM
Has experienced any
of seven online risks
Online contacts, Meeting online contacts offline, Receiving sexual messages,
Seeing sexual images, Being bullied online, Has come across one or more
negative user generated content, Has experienced personal data misuse of any
kind
DCirisk2
Online perpetrators
Concept Questions / Response options Summaries /
variable names
Cyberbullying others Have you acted in a way that might have felt hurtful or nasty to someone else in
the PAST 12 MONTHS? In which of the following ways have you acted like this
in the past 12 months…? By mobile phone calls, texts or image/video texts
[AND/OR] On the internet yes/no
QC127B and/or
QC127c
Online bullying
others
Have you acted in a way that might have felt hurtful or nasty to someone else in
the PAST 12 MONTHS? In which of the following ways have you acted like this
in the past 12 months…? On the internet yes/no
QC127c
Sending sexual
messages
In the PAST 12 MONTHS, have you sent or posted a sexual message (example:
words, pictures or video) of any kind on the internet? This could be about you or
someone else. yes/no
QC179
Has done either of
the two things
associated with
being a perpetrator
Online bullying others, Sending sexual messages DCiperp2
Technical Report and User Guide: The 2010 EU Kids Online Survey
96
Harm from online risks
(overall, sexual images, sexual messages, meeting online contacts offline, being bullied online)
Concept Questions / Response options Summaries /
variable names
Experience of
harm on the
internet (overall)
In the past 12 months, have you seen or experienced something on the
internet that has bothered you in some way? For example, made you
feel uncomfortable, upset, or feel that you shouldn’t have seen it. Yes/no
QC110
Experience of harm
(specific risk)
And in the LAST 12 MONTHS has [the risk] bothered you in any way? For
example, made you feel uncomfortable, upset […]yes/no
QC134, QC152,
QC171
Intensity of harm
(specific risk)
Thinking about the last time you were bothered by [experiencing the risk], how
upset did you feel about it (if at all)? 0 (not at all upset) to 3 (very upset)
QC118, QC135,
QC160, QC172
Duration of harm
(sexual images,
sexual messages,
being bullied
online)
How long did you feel like this [upset] for? 1 (I got over it straight away) to 4 (I
thought about it for a couple of months or more).
QC119, QC136,
QC173
Duration of harm
(meeting online
contacts offline)
How long did you feel like this [upset] for? 1 (I got over it straight away) to 3
(I felt like that for a few weeks).
QC161
Harm index
(sexual images,
sexual messages,
being bullied
online)
Intensity x duration0 (low) – 12 (high) QC118*QC119,
QC135* QC136,
QC172* QC173
Harm index
(meeting online
contacts offline)
Intensity x duration0 (low) – 9 (high) QC160* QC161
 
97
Mediation
Concept Questions / Response options Summaries
/ variable
names
Does your parent/do either of your parents sometimes...
sit with you while you use the internet?
stay nearby when you use the internet?
encourage you to explore and learn things on the internet on your own?
do shared activities together with you on the internet?
Does your parent/do either of your parents sometimes.../ Have any teachers at your school
ever done any of these things?
Active
mediation of
internet use
talk to you about what you do on the internet?
Either
number of
available
response
options OR
if at least
one of them
was chosen
or not/
DC327NM
DP220NM
Does your parent/do either of your parents sometimes.../ Have any teachers at your school
ever done any of these things? Have your friends ever done any of these things?
Helped you when something is difficult to do or find on the internet
Explained why some websites are good or bad
Suggested ways to use the internet safely
Suggested ways to behave towards other people online
Helped you in the past when something has bothered you on the internet
Does your parent/do either of your parents sometimes.../ Have any teachers at your school
ever done any of these things?
Active
mediation of
internet safety
In general, talked to you about what to do if something on the internet bothered you
Either
number of
available
response
options OR
if at least
one of them
was chosen
or not/
DC329NM
DP222NM
Parents CURRENTLY allow them to do them only with permission/supervision, or never
allow.
Use instant messaging
Download music or films on the internet
Watch video clips on the internet
Have your own social networking profile
Give out personal information to others on the internet
Upload photos, videos or music to share with others
Have any teachers at your school ever done any of these things?
Restrictive
mediation
Made rules about what you can do on the internet at school
Either
number of
available
response
options OR
if at least
one of them
was chosen
or not/
DC328NM
DP221NM
Does your parent/either of your parents sometimes check any of the following things
afterwards?
Which websites you visited
The messages in your email or instant messaging account
Your profile on a social networking or online community
Parental
monitoring
Which friends or contacts you add to your social networking profile/instant messaging
service
Either
number of
available
response
options OR
if at least
one of them
was chosen
or not/
DC330NM
DP223NM
Technical Report and User Guide: The 2010 EU Kids Online Survey
98
Does your parent/do your parents make use of any of the following…?:
Parental controls or other means of blocking or filtering some types of website
Parental controls or other means of keeping track of the websites you visit
A service or contract that limits the time you spend on the internet
Parents
Technical
mediation
Software to prevent spam or junk mail/viruses
Either
number of
available
response
options OR
if at least
one of them
was chosen
or not/
DC331NM
DP224NM
 
99
Psychological measures
SELF-EFFICACY (variable: DCSEMN)
Adapted from Schwarzer and Jerusalem (1995; 4 items, α = .65)
Item Property Analyses, Selection and Re-phrasing for the Adapted Self-Efficacy Scale
Item Original item phrasing ITC
original items
ITC
selected items
Adapted item phrasing
for EU Kids Online II
1 I can always manage to solve difficult problems if I try hard enough. .39 - -
2 If someone opposes me, I can find means and ways to get what I want. .54 - -
3 It is easy for me to stick to my aims and accomplish my goals. .62 .60 It’s easy for me to stick to my aims
and achieve my goals.
4 I am confident that I could deal efficiently with unexpected events. .58 .60 I am confident that I can deal with
unexpected problems.
5 Thanks to my resourcefulness, I know how to handle unforeseen situations. .59 .64 I can generally work out how to
handle new situations.
6 I can solve most problems if I invest the necessary effort. .31 - -
7 I can remain calm when facing difficulties because I can rely on my coping
abilities.
.54 - -
8 When I am confronted with a problem, I can usually find several solutions. .53 - -
9 If I am in trouble, I can usually think of something to do. .55 .51 If I am in trouble I can usually think
of something to do.
10 No matter what comes my way, I’m usually able to handle it. .62 .61 I can generally work out how to
handle new situations.
Cronbach’s α .84 .80
Notes: A 3-point response scale was used (1 = Not true, 2 = A bit true, 3 = Very true), ITC: Corrected item-total correlation, original items 5 and 10 were combined for adapted
item phrasing, all analyses were performed on selected cases of children 12- 15 years from a public data set (Schwarzer, 2006; N = 1254).
Technical Report and User Guide: The 2010 EU Kids Online Survey
100
PSYCHOLOGICAL DIFFICULTIES (variable: DCSDQMN)
Adapted from Strength and Difficulties Questionnaire (SDQ; Goodman, 1998; 16 items, α = .71) using items
measuring psychological difficulties only.
Item Property Analyses and Selection for the Psychological Difficulties Scale (adapted from SDQ)
Item Item phrasing and variable names by subscale
ITC
Pilot
ITC
selected items in
EU Kid O li II
Emotional symptoms (DCSDQepMN)
1 I get a lot of headaches, stomach-aches or sickness. .40 .36
2 I worry a lot. .48 .35
3 I am often unhappy, sad or tearful. .34 .48
4 I am nervous in new situations, I easily lose confidence. .36 .37
5 I have many fears, and I am easily scared. .23 .40
Conduct problems (DCSDQcpMN)
1 I get very angry and often lose my temper. .61 .42
2 I usually do as I am told. (reversed) .07 .06
3 I fight a lot, I can make other people do what I want. .17 .27
4 I am often accused of lying or cheating. .40 .41
5 I take things that are not mine from home, school or elsewhere. .48 .26
Peer relationship problems (DCSDQppMN)
1 I am usually on my own, I generally play alone or keep to myself. .43 .26
2 I have at least one good friend. (reversed) .20 .12
3 Other people my age generally like me. (reversed) .32 .21
4 Other children or young people pick on me. .52 .42
5 I get on better with adults than with people my own age. .40 .28
Hyperactivity (DCSDQhpMN)
1 I am restless, I cannot stay still for long. .36 -
2 I am easily distracted, I find it difficult to concentrate. .46 .37
3 I think before I do things. (reversed) .34 -
4 I finish the work I’m doing, my attention is good. (reversed) .19 Cronbach’s
α .77 .71
Notes: A 3-point response scale was used (1 = Not true, 2 = A bit true, 3 = Very true); ITC: Corrected item-total correlation;
ITCs and Crobach’s αs were computed for the full psychological difficulties scale; the full sample of 9-16 year olds was
used for both analyses (NPilot = 76, NData = 25142).
SENSATION SEEKING (variable: DCsensationMN)
From Stephenson, Hoyle, Palmgreen, and Slater (2003; 2 items, r = .64, p < .001).
Item Item phrasing
1 I do dangerous things for fun
2 I do exciting things, even if they are dangerous
Notes: A 3-point response scale was used (1 = Not true, 2 = A bit true, 3 = Very true)
 
101
EXCESSIVE USE (variable: DCaddictMN)
Adapted from Šmahel, Vondráčková, Blinka, and Godoy-Etcheverry (2009; 5 items, α = .77).
Item Item phrasing
1 I have gone without eating or sleeping because of the internet
2 I have felt bothered when I cannot be on the internet
3 I have caught myself surfing when I'm not really interested
4 I have spent less time than I should with either family, friends or doing schoolwork because of
the time I spent on the internet
5 I have tried unsuccessfully to spend less time on the internet
Notes: A 4-point response scale was used (1 = Never/almost never, 2 = Not very often, 3 = Fairly often, 4 = Very often);
items were only asked of 11-16 year olds.
References:
Currie, C., Gabhainn, S, N., Godeau, E., Roberts, C., Smith, R., Currie, D., Picket, W., Richter, M., Morgan,
A., & Barnekow, V. (Eds.). (2008). Inequalities in young people's health: Health behaviour in schoolaged
children (HBSC) international report from the 2005/2006 survey. Copenhagen, Denmark: WHO
Regional Office for Europe.
Goodman R., Meltzer H., Bailey V. (1998) The Strengths and Difficulties Questionnaire: A pilot study on the
validity of the self-report version. European Child and Adolescent Psychiatry, 7, 125-130.
doi:10.1007/s007870050057
ITU (2010). Partnership on measuring ICT for development, core ICT indicators.
http://www.itu.int/dms_pub/itu-d/opb/ind/D-IND-ICT_CORE-2010-PDF-E.pdf
Livingstone, S., and Helsper, E. J. (2007) Taking risks when communicating on the internet: The role of offline
social-psychological factors in young people’s vulnerability to online risks. Information,
Communication and Society, 10(5), 619-643.
Šmahel, D., Vondráčková, P.,Blinka, L. & Godoy-Etcheverry, S. (2009). Comparing addictive Behavior on the Internet in
the Czech Republic, Chile and Sweden. In G. Cardosso, A. Cheong, J. Cole (Eds.), World wide internet:
Changing societies, economies and cultures (pp. 544-582). Macao : University of Macau.
Schwarzer, R. (2006). SPSS RAW DATA WITH 18,000 Participants. Retrieved from http://userpage.fu-
berlin.de/~health/world_24nations_25nov2006.sav
Schwarzer, R. & Jerusalem, M. (1995). Generalized Self-Efficacy scale. In J. Weinman, S. Wright & M.
Johnston (Eds.), Measures in health psychology: A user’s portfolio. Causal and control beliefs (pp. 35-
37). Windsor, UK: NFER-NELSON.
Stephenson, M. T., Hoyle, R. H., Palmgreen, P., & Slater, M. D. (2003). Brief measures of sensation seeking
for screening and large-scale surveys. Drug and Alcohol Dependence, 72(3), 279-286.
doi:10.1016/j.drugalcdep.2003.08.003
UNESCO (2006). International Standard Classification of Education I S C E D 1997, Re-edition.
http://www.uis.unesco.org/Library/Pages/DocumentMorePage.aspx?docIdValue=144&docIdFld=ID