Abstracts
The abstract, although it heads the article, is often written last, together
with the title. This is partly because writers know what they have achieved,
and partly because it is not easy to write an abstract. Abstracts have to
summarise what has been done, sometimes in as few as 150 words.
It is easier to write an abstract if you remember that all abstracts have a
basic structure. Indeed, the phrase ‘structured abstracts’ says it all. This kind
of abstract, common in medical research journals and now appearing in
many social science articles, can be adapted for most normal purposes.
STRUCTURED ABSTRACTS
Structured abstracts are typically written using ﬁve sub-headings – ‘background’,
‘aim’, ‘method’, ‘results’ and ‘conclusions’. Sometimes the wording
of these sub-headings varies a little – ‘objectives’ for ‘aim’, for example, but
the meaning is much the same.
Structured abstracts were introduced into medical research journals in the
1980s. Since then they have been widely used in medicine and other areas
of research (Nakayama et al., 2005). In 2004, I published a narrative review
of their effectiveness based upon thirty-one research papers available at that
time (Hartley, 2004). I concluded that, compared with traditional abstracts,
structured abstracts:
• contained more information
• were easier to read
• were easier to search
• facilitated peer review for conferences
• were generally welcomed by readers and by authors.
Figure 2.3.1a below shows a typical structured abstract. Figure 2.3.1b
shows the same abstract written with the sub-headings removed. It can be
seen that both abstracts are clear, and so it is useful to write an abstract in
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
Chapter 2.3
a structured form ﬁrst, and then to adjust it for the journal you are writing
for if this journal does not use them.
Figures 2.3.1a and b illustrate some of the virtues of structured abstracts.
Using the sub-headings and the appropriately spaced typographical layout
makes the content clearer (Hartley and Betts, 2007). Furthermore, structured
abstracts are easier for readers to scan, as every abstract follows the same
format. The sub-headings thus allow the readers to go to the same place
each time in an abstract to ﬁnd out what it says. Furthermore, as the information
required has to be provided by the author under each sub-heading,
nothing gets missed out. With traditional abstracts, it is all too common
to ﬁnd that some elements are missing – the background, the method or
the results, for example. Often one is left saying, ‘So, what happened?’ or
‘So what?’.
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
32 The academic article
Background. In 1997 four journals published by the British Psychological
Society began publishing structured abstracts.
Aims. The aim of the studies reported here was to assess the effects of
these structured abstracts by comparing them with original versions written
in a traditional, unstructured format.
Method. The authors of the articles accepted for publication in the four
journals were asked to supply copies of their traditional abstracts (written
when the paper was submitted for publication) together with copies of their
structured abstracts requested by the editor when their paper was accepted.
Forty-eight such requests were made, and thirty pairs of abstracts were
obtained. The abstracts were then compared on a number of measures.
Results. Analysis showed that the structured abstracts were signiﬁcantly
more readable, signiﬁcantly longer and signiﬁcantly more informative than
the traditional ones. Judges assessed the contents of the structured abstracts
more quickly and with signiﬁcantly less difﬁculty than they did the traditional
ones. Almost every respondent expressed positive attitudes to structured
abstracts.
Conclusions. The structured abstracts fared signiﬁcantly better than the
traditional ones on every measure used in this enquiry. We recommend,
therefore, that editors of other journals in the social sciences consider adopting
structured abstracts.
Figure 2.3.1a An original abstract in structured form.
Adapted from Hartley and Benjamin (1998), and reproduced with permission of the British Journal
of Educational Psychology. © the British Psychological Society.
Many people think that structured abstracts are only suitable for empirical
papers – those with ‘methods’ and ‘results’. As one of my correspondents
put it:
It seems to me that the format you have chosen imposes a unitary
conception of research, at a time when educational research in particular,
and social science more widely, has at last broken away from narrow
strictures of method and procedure.
However, I believe that the underlying characteristics of a structured abstract
can apply to many other forms of enquiry. Figure 2.3.2a, for example, shows
an original abstract written to accompany a review paper. Figure 2.3.2b
shows a revision of it that, in my view, makes the background, aims and
conclusions of the study more explicit.
Bayley and Eldredge (2003) provide references to a variety of papers in
the health sciences that have structured abstracts. These include qualitative
studies, narrative reviews, systematic reviews, meta-analyses and randomised
controlled trials. Table 2.3.1 similarly lists some more recent papers in the
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
Abstracts 33
In 1997 four journals published by the British Psychological Society began
publishing structured abstracts. The aim of the studies reported here was
to assess the effects of these structured abstracts by comparing them with
original versions written in a traditional, unstructured format. The authors
of the articles accepted for publication in the four journals were asked to
supply copies of their traditional abstracts (written when the paper was
submitted for publication) together with copies of their structured abstracts
requested by the editor when their paper was accepted. Forty-eight such
requests were made and thirty pairs of abstracts were obtained. The abstracts
were then compared on a number of measures. Analysis showed that the
structured abstracts were significantly more readable, significantly longer
and signiﬁcantly more informative than the traditional ones. Judges assessed
the contents of the structured abstracts more quickly and with signiﬁcantly
less difﬁculty than they did the traditional ones. Almost every respondent
expressed positive attitudes to structured abstracts. In short, the structured
abstracts fared signiﬁcantly better than the traditional ones on every measure
used in this enquiry. We recommend, therefore, that editors of other journals
in the social sciences consider adopting structured abstracts.
Figure 2.3.1b The same abstract in unstructured form.
health and social sciences that have used structured abstracts with a variety
of research methods.
After the title, the abstract is the most frequently read part of any paper.
Writing it in a structured format (with or without the headings) ensures
that it is informative and complete.
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
34 The academic article
There is something of a controversy taking place over how best to theorise
human learning. In this article we join the debate over the relationships
between sociocultural and constructive perspectives on learning. These two
perspectives differ in not just their conceptions of knowledge (epistemological
assumptions) but also in their assumptions about the known world and the
knowing human (ontological assumptions). We articulate in this article six
themes of a nondualist ontology seen at work in the sociocultural perspective,
and suggest a reconciliation of the two. We propose that learning involves
becoming a member of a community, constructing knowledge of various levels
of expertise as a participant, but also taking a stand on the culture of one’s
community in an effort to take up and overcome the estrangement and division
that are consequences of participation. Learning entails transformation of both
the person and the social world. We explore the implications of this view
for thinking about schooling and for the conduct of educational research.
Figure 2.3.2a An original abstract for a review paper.
Reproduced with permission from Packer and Goicoechea (2000) and Taylor & Francis, www.
informaworld.com.
Table 2.3.1 Examples of studies with structured abstracts published in the health and
social sciences
Method Example
Literature review Mayhew and Simpson (2002)
Observational study Lauth et al. (2006)
Survey Wilding and Andrews (2006)
Longitudinal study Flouri (2006)
Statistical paper Prosser and Trigwell (2006)
Simulation Wright (2006)
Experimental study Clariana and Koul (2006)
Epidemiological study Evans (2000)
Meta-analysis Bunn et al. (2006)
Systematic review Duperrex et al. (2006)
Qualitative study Maliski et al. (2002)
REFERENCES
Bayley, L. & Eldredge, J. E. (2003). The structured abstract: An essential tool for
researchers. Hypothesis, 17(1), 1 and 11–15. Or: http://research.mlanet.org/structured_
abstract.html. (Retrieved 1 August 2006.)
Bunn, F., Collier, T., Frost, C., Ker, K., Roberts, I. & Wentz, R. (2003). Trafﬁc calming
for the prevention of road trafﬁc injuries: Systematic review and meta-analysis. Injury
Prevention, 9(3), 200–4.
Clariana, R. B. & Koul, R. (2006). The effects of different forms of feedback on fuzzy
and verbatim memory of science principles. British Journal of Educational Psychology,
76(2), 259–70.
Duperrex, O., Bunn, F. & Roberts, I. (2002). Safety education of pedestrians for injury
prevention: A systematic review of randomised controlled trials. British Medical Journal,
324(7348), 1129–34.
Evans, L. (2000). Risks older drivers face themselves and threats they pose to other road
users. International Journal of Epidemiology, 29(2), 315–22.
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
Abstracts 35
Background. An interesting debate is currently taking place among
proponents of different ways of thinking about human learning. In this article
we focus on that portion of the debate that addresses sociological and
constructive perspectives on learning. These two perspectives differ in not
just their conceptions of knowledge (epistemological assumptions) but also
in their assumptions about the known world and the knowing human
(ontological assumptions).
Aims and approach. We wish to try and reconcile these two different
approaches ﬁrst by examining the ontological assumptions of them both. We
then consider six key themes of a nondualist ontology seen at work in the
sociocultural perspective. Finally we propose that the constructive perspective
attends to epistemological structures and processes which the sociological
perspective must place in a broader historical and cultural context.
Conclusions. We conclude that learning involves becoming a member of a
community, constructing knowledge of various levels of expertise as a
participant, and taking a stand on the culture of one’s community in an effort
to take up and overcome the estrangement and division that are consequences
of participation. Learning entails transformation of both the personal and the
social world. We explore the implications of this view for thinking about
schooling and the conduct of educational research.
Figure 2.3.2b The same abstract in structured form.
Flouri, E. (2006). Parental interest in children’s education, children’s self-esteem and
locus of control, and later educational attainment: Twenty-six year follow-up of the
1970 British Birth Cohort. British Journal of Educational Psychology, 76(1), 41–56.
Hartley, J. (2004). Current ﬁndings from research on structured abstracts. Journal of the
Medical Library Association, 92(3), 368–71.
Hartley, J. & Benjamin, M. (1998). An evaluation of structured abstracts in journals
published by the British Psychological Society. British Journal of Educational Psychology,
68(3), 443–56.
Hartley, J. & Betts, L. (2007). The effects of spacing and titles on judgments of the
effectiveness of structured abstracts. Journal of the American Society for Information Science
& Technology, 58(14), 2335–40.
Lauth, G. W., Heubeck, B. G. & Mackowiak, K. (2006). Observation of children with
attention-deﬁcit hyperactivity (ADHD) problems in three natural classroom contexts.
British Journal of Educational Psychology, 76(2), 385–404.
Maliski, S. L., Heilemann, M. V. & McCorkle, R. (2002). From ‘death sentence’ to
‘good cancer’: Couples’ transformation of a prostate cancer diagnosis. Nursing Research,
5(6), 391–7.
Mayhew, D. R. & Simpson, H. M. (2002). The safety value of driver education and
training. Injury Prevention, 8 (Suppl. II): ii3-ii8.
Nakayama, T., Hirai, N., Yamazaki, S. & Naito, M. (2005). Adoption of structured
abstracts by general medical journals and format for a structured abstract. Journal of
the Medical Library Association, 93(2), 237–42.
Packer, M. J. & Goicoechea, J. (2000). Sociocultural and constructivist theories of learning:
Ontological, not just epistemology. Educational Psychologist, 35(4), 227–41.
Prosser, M. & Trigwell, K. (2006). Conﬁrmatory factor analysis of the Approaches to
Teaching Inventory. British Journal of Educational Psychology, 76(2), 405–19.
Wilding, J. & Andrews, B. (2006). Life goals, approaches to study and performance in
an undergraduate cohort. British Journal of Educational Psychology, 76(1), 171–82.
Wright, D. B. (2006). Comparing groups in a before-after design: When t test and
ANCOVA produce different results. British Journal of Educational Psychology, 76(3),
663–75.
FURTHER READING
Hartley, J., Rock, J. & Fox, C. (2005). Teaching psychology students to write structured
abstracts: An evaluation study. Psychology Teaching Review, 1(1), 2–11.
Kamler, B. & Thomson, P. (2004). Driven to abstraction: Doctoral supervision and
writing pedagogies. Teaching in Higher Education, 9(2), 195–209.
Kelly, A. E. & Yin, R. K. (2007). Strengthening structured abstracts for education
research: The need for claim-based structured abstracts. Educational Researcher, 36(3),
133–8.
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
36 The academic article
Introductions
It is but a short step from structured abstracts to structured texts. In the
following chapters we shall see how each part of the structure of a scientiﬁc
article (the introduction, method, results, discussion and conclusion) can
indeed be subdivided into ﬁner structures.
Swales and Feak (2004) describe what they characterise as ‘moves’ in
the various sections of academic articles. Basically, a ‘move’ is a stage in the
argument that all writers go through. The ‘moves’ for the introduction are
typically as follows (p. 244):
• Move 1: The authors establish a research territory:
(a) by showing that the general research area is important,
central, interesting, problematic or relevant in some way
(optional);
(b) by introducing and reviewing items of previous research
in the area (obligatory).
• Move 2: They then establish a ‘niche’ by indicating a weakness in the
account so far:
(a) by indicating a gap in the previous research, raising a
question about it or extending previous knowledge in some
way (obligatory).
• Move 3: They then occupy the niche by saying they are going to put
this right:
(a) by outlining the purposes or stating the nature of the
present research (obligatory);
(b) by listing research questions or hypotheses to be tested
(optional);
(c) by announcing the principal ﬁndings (optional).
Swales and Feak argue that most introductions to academic articles follow
this basic structure. Lewin et al. (2001) offer a similar, but more detailed,
analysis that readers might also ﬁnd useful.
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
Chapter 2.5
AN EXAMPLE
While writing this section of Academic Writing and Publishing, I coincidentally
received a copy of a paper by Slatcher and Pennebaker (2006). This paper
was about the effects of one of the partners of a dating couple writing either
neutral or strongly emotional letters to the other one about their relationship.
The paper concluded that the participants who wrote the emotional letters
were signiﬁcantly more likely to be dating their romantic partners three
months later than were the writers of the neutral letters. Be that as it may,
I was intrigued to observe that the introduction to this paper followed
almost exactly the generic structure described by Swales and Feak.
Slatcher and Pennebaker’s introduction contains ﬁve paragraphs. Here are
some examples of how the moves appear:
Move 1: Establishing a research territory
The paper starts (paragraphs 1 and 2) with describing the background and
setting the scene. Key phrases are: ‘Researchers are now . . .’, ‘Preliminary
ﬁndings suggest . . .’, ‘There are a number of ways in which one could measure
the effects of expressive writing . . .’.
Move 2: Establishing a niche
The paper continues (in paragraphs 3 and 4) with the following key phrases:
‘Although previous studies have addressed . . . none have . . .’, ‘One potential
mediator is . . .’, ‘There are various ways to measure . . .’, ‘The use of emotional
words may be particularly relevant . . .’, ‘One way is to analyse the texts
used in instant messaging . . .’.
Move 3: Occupying the niche
The introduction concludes (in paragraph 5) with the following key phrases:
‘In the present study we sought to investigate the social effects of expressive
writing . . .’, ‘Three predictions were tested. First . . .’.
Slatcher and Pennebaker thus follow Swales and Feak’s analysis almost
line by line. It is also worth noting, in passing, that the literature review
in this paper is quite short, and there are only nine references. Day and
Gastel (2006) comment that, ‘Introductions should supply sufﬁcient information
to allow the reader to understand and evaluate the results of the
present study without (them) needing to refer to previous publications on
the topic’ (pp. 57–8).
Of course many papers are written with more detailed substructures.
Three types of structure typical in introductions are:
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
42 The academic article
1 The one listed above – where the authors establish their niche by
indicating limitations or omissions in the previous research.
2 One where two (or more) different areas of research are reviewed – and
the authors establish their niche by bringing them together.
3 One where some previous research has provided support for a particular
ﬁnding or theory, and some has not – and the authors establish their
niche by seeking to resolve and explain this.
Further, there are disciplinary variations: Haggan (1998), for example,
examined the introductions for twenty-six articles in the sciences, twentysix
in linguistics and twenty-six in the arts. She found that the introductions
in the science papers were less likely to contain a plan for the paper than
were the introductions in linguistics, and that they lay midway in their use
of impersonal language between introductions in the arts (the least personal)
and introductions in linguistics (the most personal). Introductions in the
sciences were more personal, however, when there was more than one author.
Such disciplinary formulaic introductions enhance the clarity of a paper
and ensure that the readers’ expectations about the format and the purpose
of an introduction are maintained. Such devices keep the reader reading.
REFERENCES
Day, R. A. & Gastel, B. (2006). How to write and publish a scientiﬁc paper (6th edn).
Cambridge: Cambridge University Press.
Haggan, M. (1998). In search of the linguistics niche: A study of research article
introductions in Linguistics, Literature and Science. Arab Journal for the Humanities,
61, 345–77.
Lewin, B., Fine, J. & Young, L. (2001). Expository discourse: A genre-based approach to social
science research texts. London: Continuum.
Slatcher, R. B. & Pennebaker, J. W. (2006). How do I love thee? Let me count the
words. Psychological Science, 17(8), 660–4.
Swales, J. M. & Feak, C. B. (2004). Academic writing for graduate students (2nd edn). Ann
Arbor, MI: University of Michigan Press.
FURTHER READING
Kendall, P. C., Silk, J. S. & Chu, B. C. (2000). Introducing your research report:
Writing the introduction. In R. J. Sternberg (Ed.), Guide to publishing in psychology
journals (pp. 41–57). Cambridge: Cambridge University Press.
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
Introductions 43
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
Methods
Method sections vary in journal articles, but rather less so than introductions.
This is because the ‘moves’ in the method sections generally involve working
through a series of subsections. Most method sections are usually subdivided
(with subheadings) into three sections, as follows:
1 participants
2 measures
3 procedure(s).
If no participants are involved, then the method simply describes the measures
and procedure(s). In the Slatcher and Pennebaker (2006) example, there are
three subheadings in the method section: Participants, Procedure and
Linguistic Analysis (or measures).
Method sections may be brief and succinct – when the methods used are
well known and standardised – or quite lengthy, when the methods used
are new or different and thus require careful elaboration.
Students and authors are typically instructed to write their method sections
in such a way that readers can repeat the method from the descriptions
given. Day and Gastel (2006, p. 64) recommend that colleagues unfamiliar
with what was done should be asked to read the account to see if they can
follow it. Authors are sometimes too close to what they did and thus tend
to forget to mention tiny but – sometimes – key details.
A useful device for clarifying the procedure or the method for the reader
– especially if it is complicated – is to summarise it in a table or ﬁgure
(e.g. see Gotzsche, 2006). Figure 2.6.1 gives a schematic version of Slatcher
and Pennebaker’s prose description of their method. Such procedures, though,
are rarely used. None of the authors of ﬁfty-six articles in the 2005 volume
of the Journal of Educational Psychology used this strategy, and only two
provided illustrations of the equipment used. However, eleven (i.e. twenty
per cent) of these articles did include ﬁgures to illustrate either the theoretical
models underlying the reasoning for their experiments or the analyses that
they were going to use.
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
Chapter 2.6
REFERENCES
Day, R. A. & Gastel, B. (2006). How to write and publish a scientiﬁc paper (6th edn).
Cambridge: Cambridge University Press.
Gotzsche, P. C. (2006). Believability of relative risks and odds ratios in abstracts: A
cross sectional study. British Medical Journal, 333, 231–4.
Slatcher, R. B. & Pennebaker, J. W. (2006). How do I love thee? Let me count the
words. Psychological Science, 17(8), 660–4.
FURTHER READING
Reis, H. T. (2000). Writing effectively about design. In R. J. Sternberg (Ed.), Guide to
publishing in psychology journals (pp. 81–97). Cambridge: Cambridge University Press.
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
46 The academic article
ProcedureParticipants Measures
86 couples
Relationship
Assessment
Scale
Experimental group (N = 44 couples)
Write text messages for 20 mins
on days 4, 5 and 6 about their
deepest thoughts and feelings
about their current romantic
relationship
Control group (N = 42 couples)
Write text messages for 20 mins
on days 4, 5 and 6 about their
daily activities
Figure 2.6.1 A schematic illustration of the prose version of the Method used in the study
by Slatcher and Pennebaker (2006).
Results
A ‘moves’ analysis of the results sections of academic articles either looks
like this:
• Move 1: State the main ﬁndings in order – relating them in turn to the
hypotheses and methods used.
• Move 2: State the subsidiary ﬁndings – relating them in turn to the
hypotheses and methods used.
or it is an interweaving of the two – the ﬁrst set of main ﬁndings and related
subsidiary ones, followed by the second set, and so on.
Again these subsections may be cued by subheadings. Slatcher and
Pennebaker (2006), for example, divide their results section into two main
parts (separated by the subheadings, ‘Relationship stability and language
use’, and ‘Mediation effects of changes in use of emotional words’). They
provide a description of the results obtained, mainly in prose, in each part,
indicating that the partners who wrote the romantic letters were signiﬁcantly
more likely to be dating their romantic partners three months later than
were the partners who wrote the neutral ones.
It is typical in results sections to present the main data that support (or
reject) the hypotheses in the form of tables and graphs. Indeed, it is quite
common to ﬁnd that the ﬁrst sentence of a results section begins, ‘Table 1
shows that . . .’. Slatcher and Pennebaker’s paper is unusual here in that
they provide only one such table, near the start of their second section of
results, and this table is not used to illustrate their main ﬁndings. Because
tables and graphs are so important in academic and scientiﬁc writing, I shall
discuss them separately, in more detail, in Chapter 3.5.
Salovey (2000) argues that the art of writing a good results section is to
take the readers through a story. This does not mean working step by step
through the results obtained, but rather – as implied above – articulating
what happened and illustrating it clearly, usually with data. In my view,
this story is clearer if the sequence of topics addressed in the results section
is the same as that articulated in the introduction and the method(s) sections.
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
Chapter 2.7
Swales and Feak (2004) comment that the distinction between the results
and the subsequent discussion section is not always as sharp as one might
think. They cite a study by Thompson (1993) that showed that the authors
of papers in biochemistry used a variety of rhetorical devices in their results
section to justify their methodology, to interpret and comment on the ﬁndings,
and to relate them to previous research. Indeed, the only thing that they
did not do in their results sections was to call for further research – this
was left for the discussion.
REFERENCES
Salovey, P. (2000). Results that get results. Telling a good story. In R. J. Sternberg
(Ed.), Guide to publishing in psychology journals (pp. 121–32). Cambridge: Cambridge
University Press.
Slatcher, R. B. & Pennebaker, J. W. (2006). How do I love thee? Let me count the
words. Psychological Science, 17(8), 660–4.
Swales, J. M. & Feak, C. B. (2004). Academic writing for graduate students (2nd edn). Ann
Arbor, MI: University of Michigan Press.
Thompson, D. K. (1993). Arguing for experimental ‘facts’ in science: A study of research
article results sections in biochemistry. Written Communication, 10(1), 106–28.
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
48 The academic article
Discussions
Discussions, like introductions, have a typical structure. Lewin et al. (2001)
and Swales and Feak (2004) describe typical ‘moves’ in the discussion sections
of academic research papers. Putting these descriptions together suggests
the following moves:
• Move 1: Restate the ﬁndings and accomplishments.
• Move 2: Evaluate how the results ﬁt in with the previous ﬁndings – do
they contradict, qualify, agree or go beyond them?
• Move 3: List potential limitations to the study.
• Move 4: Offer an interpretation/explanation of these results and ward off
counter-claims.
• Move 5: State the implications and recommend further research.
Discussions, then, go beyond a summary of the ﬁndings and, indeed,
there may be disciplinary differences in how they are approached. Holmes
(1997), for instance, found that the discussion sections of papers in sociology
and political science were similar in format to those in the sciences, whereas
those in history were less complex. Swales and Feak (2004) state that some
scientists believe that a long discussion implies weak methods and results,
whereas social scientists and people in the arts may well believe the opposite.
AN EXAMPLE
Lewin et al. (2001) provide numerous quotations from the discussion sections
of several research articles to support the above ‘moves’ analysis. In terms
of Slatcher and Pennebaker’s (2006) paper referred to earlier, we may note
the following sentences contained in the six paragraphs of their discussion
section:
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
Chapter 2.8
• Move 1: Restating the ﬁndings and accomplishments:
– Par. 1: ‘The very simple act of writing about their romantic relationship
changed the way in which participants communicated
. . .’;
– Par. 2: ‘Taken together these findings shed light on processes
underlying interactions in close relationships . . .’;
– Par. 3: ‘An advantage of the current design is that . . .’;
– Par. 6: ‘Unlike previous expressive-writing studies, this is the ﬁrst
to demonstrate . . .’.
• Move 2: Evaluating how the results ﬁt in with previous research:
– Par. 3: ‘In particular, the ﬁndings relating to increases in emotion
words illuminate previous research [3 references provided]’.
• Move 3: Stating the limitations:
– Par. 5: ‘There are some potential limitations in this study. First
. . . Second . . .’.
• Move 4: Warding off alternative explanations:
– Par. 5. ‘. . . make this an unlikely possibility’.
• Move 5: Stating implications:
– Par. 4: . . . [this ﬁnding] ‘has clear implications for clinicians’;
– Par. 5: ‘. . . future studies should address this issue’.
These quotations illustrate that the ﬁve moves are present, but they are not
as clearly sequenced or indicated as might be implied from the list above.
Authors seem more ﬂexible in how they tackle their discussions, although
the moves listed are usually present.
Discussion sections are difﬁcult to write because their aim is to discuss
and comment on the ﬁndings, rather than just to report them. Day and
Gastel (2006) suggest that journal editors reject many papers because of
their weak discussions. They recommend that discussions should end with
a short summary regarding the signiﬁcance of the work, which, they claim,
is not always adequately considered.
Woods (1999) recommends:
1 that writers should keep notes about what it might be useful to include
in the discussion as ideas occur to them when they are writing other
sections; and
2 that it might be wise to set aside a day or two to tackle this section of
the paper.
This, he says, will make the task less daunting.
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
50 The academic article
REFERENCES
Day, R. A. & Gastel, B. (2006). How to write and publish a scientiﬁc paper (6th edn).
Cambridge: Cambridge University Press.
Holmes, R. (1997). Genre analysis, and the social sciences: An investigation of the
structure of research article discussion sections in three disciplines. English for Speciﬁc
Purposes, 16(4), 321–37.
Lewin, B., Fine, J. & Young, L. (2001). Expository discourse: A genre-based approach to social
science research texts. London: Continuum.
Slatcher, R. B. & Pennebaker, J. W. (2006). How do I love thee? Let me count the
words. Psychological Science, 17(8), 660–4.
Swales, J. M. & Feak, C. B. (2004). Academic writing for graduate students (2nd edn). Ann
Arbor, MI: University of Michigan Press.
Woods, P. (1999). Successful writing for qualitative researchers. London: Routledge.
FURTHER READING
Calfee, R. (2000). What does it all mean? The discussion. In R. J. Sternberg (Ed.),
Guide to publishing in psychology journals (pp. 133–45). Cambridge: Cambridge University
Press.
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111
Discussions 51
1111
2
3
4
5
6
7
8
9
1011
1
2
31
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44111