Selecting the question/s you need Selecting a research question lA question about social world that you try to answer by use of empirical data lOne over-arching question lSome sub-questions l lTwo fundamental criteria: l l1) interest l2) researchability Questions that can be researched lA successful research project depends on how researcher develops the research question. lForms the foundation of your study. lMust be specified before you begin your study/data gathering/analysis. lA well-formulated research question requires specificity and preciseness. lA research question guides and centers your research. lIt should be clear and focused. lIt should synthesize multiple sources of information and present your argument as a researcher. l l l Characteristics of good RQs vClear vRelevant vManageable vOriginal, substantial vEmpirically focused vSignificant vAwareness of assumptions vAccessible evidence vRelated to previous research vEthical vInteresting to you Source: recommended by Ingrid Lunt (Oxford university) Clarity lIt should be answerable! l lIt should be understandable! l lIt should make an original contribution to knowledge in some way! Relevant lThe question should be of intellectual interest to people in the field you have chosen to study. lThe question arises from issues raised in the literature or in practice. lYou should be able to establish a clear purpose for your research in relation to the chosen field. l Manageable lYou should be realistic about the aim and scale of your project. lThe question must be within your ability to tackle. lSometimes a RQ appears feasible, but when you start fieldwork or library study, it proves it isn‘t. Substantial, original lRQ is not a simply copy/paste proces for questions asked in other researchers unless it is planned to be conducted on a different sample or in a comparable way. l lIt shows imagination and ability to construct and develop research issues. Empirical focus lGenerate data to answer question. lDetermine methods of enquiry and data collection. lDifferent types of questions will lead to different approaches to research and methods of data collection. l Significant lIs there a clear rationale for the question? l lDoes this question matter? l lWhy is it of interest? To whom is of interest? l Awareness of assumptions lIt should allow your aim to satisfy the outcomes of the study. lExample: a theoretical approach l How do you formulate a good RQ? lChoose a general topic of interest, and conduct preliminary research! l lDetermine what kinds of questions the topic generates! lWho is the audience? l“How?” “What?” and Why?” questions! lTry to evaluate possible responses to those questions. Examples of RQs – WHY? Good RQs Why do students enrolled in Seminar on Research Methods classes perform significantly differently on writing a research proposal than students who do not participate in that seminar? Why some people tend to be more prone to drug consumption than others? How to formulate sociological RQ? lRQs examine the social meaning or patterns of a phenomenon. lKey element – social l lExample: Which of the following RQs is sociological? 1)Does having an abortion increase the chances a woman will be depressed? 2)Does having an abortion vary by social class and ethnicity? How to revise RQ to make it more interesting? lTwo situations: l lRQs are not relevant for social science or concern just a small population l lRQ is not interesting because the results are already known l l 1st situation lYou should enlarge RQ – generalize the question l lHowever, irrelevance is a matter of degree. l lTry to link RQ to theories and other empirical work! 2nd situation lMake sure that the method is not - to some degree - pre-ordaining the results! l lExample: the effect of marital satisfaction on the overall life satisfaction l l- Find independent measures of the two concepts – they might not lead to positive relationship between the two variables. 2nd situation lExistence of many researches does not necessarily mean that scholars have reached a consensus [ a research that finds definitive conclusions and make a significant contribution l lWhen you find prior well-established findings, [ try to work on a new direction or challenge the existing findings l l lExample: prior research have demonstrated that married people are happier than unmarried. l1. Challenge prior research: maybe the key-concepts are not well measured, statistical methods are not good enough, problems with sample, data. l2. Clarify concepts in prior research: eliminating fuzziness in concepts – not new results, but new understanding of the meanings Extend prior research lFocus on a key-subpopulation lUse a new population lUse a different time period lUse new subpopulation or population to test a specific hypothesis lRefine or extend the causal explanations l Refine or extend the causal explanations l1) Mediating variable: l lGender g ??? g earnings l lCausal direction is not obvious! – marriage and happiness??? l l2) Non-mediating variables l Mediators lA mediator variable explains the relationship between a predictor and an outcome l lExample: l We are interested in whether or not males and females have differing unemployment period because of differing levels of social support Moderators lA moderator is a variable that alters the direction or strength of the relationship between a predictor and an outcome lIt is just an interaction – the effect of one variable depends on the level of another lExample: l We are interested not only in the effect of social support on return to labour market, but whether this differs if the person is male or female Moderator model How to structure quantitative RQs - some steps 1)choose the type of quantitative RQ - descriptive, comparative or relationship-based; 2)come up with different types of variables to be measured, manipulated and/or controlled, as well as any groups of interest; 3)choose an appropriate structure for the chosen type of quantitative research question; 4)write down issues you are trying to address in the form of a complete research question Questionnaire What is it? l A set of questions designed to generate the data necessary for accomplishing a research project's goals! Impact on data accuracy lIt must communicate to the respondent what the researcher is asking l l l lIt must communicate to the researcher what the respondent has said Sample size – quantitative approach lYour survey results are representative Û a large number of randomly-selected participants in each group you survey lWhat exactly is "a large number?" Sample Size (N) Margin of Error (percentage) 10 31.6 20 22.4 50 14.1 100 10.0 200 7.1 500 4.5 1000 3.2 2000 2.2 5000 1.4 10000 1.0 Sample Size Calculator lYou can use online sample size calculators to determine the number of people you need to interview for a given precision in your results. l lhttp://www.surveysystem.com/sscalc.htm l lhttp://www.custominsight.com/articles/random-sample-calculator.asp l How many respondents do you need for a pilot study? lConnelly (2008) ⇨ 10% of the sample projected for the larger parent study. lBut, Hertzog (2008) ⇨ it is not a simple or straight forward issue to resolve because these types of studies are influenced by many factors. lPossible: 10 would be a minimum, and 30 maximum when you operate with a sample size of 300. Sample size – qualitative approach lThe answer is that there is no rule of thumb!!! lThere are some types of research where: 1)larger numbers of subjects may be easier to get, 2)more advisable to obtain, 3)more commonly sought. l lA qualitative sample must be big enough to assure that we are likely to hear most or all of the perceptions that might be important. Common sense advice for qualitative approach lNo reasonable answer, no magic number lThe only possible answer ⇨ to have enough interviews to say what you think is true and not to say things you don’t have that number for. lNumber of interviews you need will change from day to day as you learn more and revise your ideas. lTo stop will be somewhat arbitrary (saturation, maybe running out of time or money …). How should you administer the questionnaire? lSelf administered questionnaires (post, email, or online). lInterview administered questionnaires ( telephone or face to face). lAdvantages of self administered questionnaires include: 1)Cheap and easy to administer. 2)Preserve confidentiality. 3)Can be completed at respondent's convenience. l lAdvantages of interview administered questionnaires include: 1)Allow participation by illiterate people. 2)Allow clarification of ambiguity. 3)Targeting individuals who provide specific information. 4)Greater guarantee of a response. 5) lThe exact method of administration also depends on who the respondents are. Questionnaire Construction lAvoiding sensitive issues l lTypes of responses to questions l lDemographic questions l Introduction in the questionnaire lIdentification of the survey or respondent lPurpose of survey lExplanation of respondent selection lRequest for participation/provide incentive lScreening of respondent Sensitive issues lWe cannot ask questions regarding a person’s preferences, desires, and interests on sensitive or private matters l lAny examples? – How many glasses of wine do you drink every evening? l lWe can ask them what they think in a more general way l l l Question form lNonstructured questions lOpen-ended l l l lStructured questions lFixed-response Types of responses lEasy Analysis lRadio Buttons lExactly one response lCheck Boxes lMany responses (none to all) lHard Analysis lText Boxes lSmall areas form response lText Areas lLarge areas form response l Types of responses lRadio responses lRecommended to have most of the questions with this type of response lSelect exactly one of two or more choices lExample: lYes | No lFemale | Male lPrimary education | Lower secondary education | Upper secondary education | Tertiary education lScaled: Totally agree | Agree | Disagree | Totally disagree § Types of Responses lCheck box responses lSome of these lChoose none or more from a list lMay include others options. lExamples: lList of books lList of food products lInclude others so the respondent can fill in what we forgot Types of Responses lFree form responses lText box lSpace for a word or phrase lText area lSpace for an extended response l lKeep these types of responses to a minimum l WHY? lHard to quantify Demographic Questions lQuestions that quantify your population lDemographics provide a standard for measuring other responses lUse radio responses lGender lClass lEducation lMarital status lOccupation lChoose those that are appropriate for your questionnaire Response category content lResponse choices should be: lCollectively exhaustive ltaken together lMutually exclusive lnot overlap l Response category l The response choices are collectively exhaustive but not mutually exclusive!!! lOn the average, how many glasses of wine (100 ml) do you drink per week? l____0 to 3 glasses l____3 to 6 glasses l____More than 6 glasses l l Changing the 3 in the second category to a 4 will make the response choices mutually exclusive! Pretesting the questionnaire •Pretest the entire survey process, including the questionnaire: sampling frame, sample draw, data gathering (mail, phone, online, etc.), editing, coding, file building, data entry, and preliminary analysis •Questionnaire pretest: 20-40 questionnaires; “10 percent change pretest rule” •Changes: add Qs, delete Qs, modify Qs, change order of Qs •Less than 10 percent change no new pretest, 10 percent or more, pretest again Writing a questionnaire for international use lCaution lUse of the same word can have different meanings in different cultures l l lWord Equivalence lDon’t be surprised by your research findings! lSmall effects or no effects – also a finding lSometimes more interesting than large effects! l lNon-effect ≠ non-finding Preparation of qualitative research – the role of the interviewer lRole ⇨ complex and multifaceted l lLocate respondents and ask for their agreement to take part in the study lMotivate respondents to do good job lClarify any confusion/concerns lObserve the quality of responses lConduct a good interview Training the interviewers – when researcher ¹ interviewer lDescribe the entire study lState who is funding/paying for the research lOffer enough information about survey research lExplain the sampling logic and process lExplain interviewer bias lGo through the interview lExplain respondent selection procedures lRehearse interview lExplain supervision lExplain scheduling Interview lOpening remarks: l “Hello. My name is ______ and I am student at _. I’m working on a research project looking at __ (title of your research project) for a course called ‘….’ or my BA/MA/PhD thesis. The aim of my research is_________. Your answers are anonymous and will help me to understand…„ Asking the questions lUse questionnaire carefully, but informally lAsk questions exactly as written lFollow the order given, if there is a continuity between questions (do not jump to another topic) lAsk every question lDon't finish sentences Obtaining proper responses when answer is brief – use of probe lSilent probe lEncouragement lElaboration - "Would you like to elaborate on that?" or "Is there anything else you would like to add?" lAsk for clarification - „Before, you were talking about … Could you tell me more about that?" lRepetition