How to Design and Report Experiments

by Andy Field

  These results are in line with previous research showing that there is some validity to commonly-held national stereotypes: there appear to be real cross-cultural differences in behaviour which underlie these beliefs (Biggott, 1967; Raciste, Morone and Kruelle, 2000; Wikked and Hartless, 2001). Biggott’s ‘Theory of Racially Induced Patterns of Expression’ (TRIPE) suggests that cultural and racial stereotypes have their origins in early socialization patterns which have become slightly different around the world due to geographical isolation. In Raciste et al.’s ‘Framework Accounting Specificially for Culturally Induced Social Traits’ (FASCIST) theory, cross-cultural differences stem not from childhood experiences, but instead arise when adults of one culture interact with people of another: The stereotypical behaviours are seen as an attempt by the native population to maintain their social identity in the face of threat from a ‘foreigner’ or ‘outsider’. There is evidence in support of both of these theories, and the present study cannot decide between them conclusively. However, it does demonstrate that, contrary to Raciste et al.’s assertions, these cross-cultural differences stem from the behaviour of the ‘foreigner’ rather than the perceptions of the native population in which they find themselves.

  Furthermore, most of the earlier researchers based their conclusions on people’s verbal reports of how they would behave in various situations: for example, even in Raciste et al.’s comparatively recent study, participants were merely asked how acceptable they would find dog-beating. How these participants’ reactions to a real dog-beating would relate to their verbal claims was not investigated, and yet the relationship between overt behaviour and self-report has been shown to be an important issue (Thyn Kin, Sey Ying and Doowing, 1978). In the present experiment, participants’ overt behaviour in a naturalistic situation (sun-bed claiming around a hotel pool) was recorded, without their knowledge that their behaviour was being scrutinized. The present study therefore provides important information on how different nationalities behave in practice, and suggests – contrary to ThynKin et al.’s claims – that cross-cultural differences in overt behaviour are very real and pronounced. The results described here also suggest that these differences have persisted over a comparatively long period of time, given their consistency with the findings of Buonaparte and Nelson (1805).

  However, there are several factors which need to be considered in evaluating the findings of the present research. First, although the observed differences in sun-bed claiming behaviour were statistically significant, they are small in absolute terms: all nationalities were comparatively quick to claim a sun-bed. The maximum difference in latency, between the young Germans and the old Americans was still only approximately 10 seconds; given that the swimming pool was 200 metres from the dining room, it is clear that all participants were in a hurry to claim a sun-bed. The fact that the hotel had only three sun-beds for 200 residents may have had a part to play in this: a future study should include a greater range of hotels, to determine whether the present results generalize to situations in which the resource (i.e. sun-beds) is not in such short supply.

  Secondly, although participants were selected who lacked obvious physical infirmities or disorders, it became apparent during the course of the study that the nationalities were not strictly comparable in terms of physical fitness. The Germans had lithe, firm, fit bodies, in contrast to the American and English tourists, most of whom were somewhat rotund. Although it was not possible to measure fitness objectively in the present study (measurements of waist-size obtained from the video proved unreliable) it was noticeable that many of the English and American tourists waddled to their targeted sun-bed, and then collapsed upon it with a noticeable shortness of breath. Similar behaviour was rarely observed amongst the German tourists. Differences in physical fitness, as opposed to desperation to claim a sun-bed, may therefore have contributed to the observed differences in running speed. Most of the difference between the younger and older participants could be attributable to fitness, rather than due to motivational differences. A future study should take care to ensure that participants are more evenly matched in terms of physical fitness than was the case in the present work.

  A third problem with this study was that most of the German tourists were on a 3-day holiday, whereas the American and English tourists were all booked into the hotel for 14 days. The Germans therefore had less time in which to sun-bathe, a factor which may have contributed to the difference between their behaviour and that of the other two nationalities. However, clearly not all differences in sun-bed claiming behaviour can be attributed to this factor, given that there was also a difference in behaviour between the American and English tourists. Future research should take greater care to ensure that the participants are better matched on any factors such as holiday stay, which might have a significant effect on their motivational level.

  Finally, the present study demonstrates a behavioural basis for a national stereotype in only one domain: sun-bed claiming behaviour. It remains to be determined whether these differences hold true across other situations, or are specific to the hotel pool environment.

  9.8 References

  * * *

  See Chapter 14 for full details; basically, here you provide, in alphabetical order and in a very standardized format, full details of every work that was cited in the body of the text. Here are the references from our fictitious study:

  Biggott, R.S. (1967) National differences in queue-jumping behaviour: an observational study. Journal of Irreproducible Results, 17 (1), 296–305.

  Buonaparte, N. and Nelson, H. (1805, March 21). An analysis of sun-bed claiming behaviour in Western Europe. Seafaring Weekly, 75, 46–49.

  Kebbab, D., Burghur, S. and Schnitzel, Y.P. (1995) ‘Pushiness’ at supermarket checkouts as a function of age and nationality. Journal of Shopping Behaviour, 5 (2), 36–42.

  Krapp, N.A. and Fewtile, T. (1966) Cultural differences in the acceptability of toenail-clipping behaviour. Pedicure, 36 (I), 12–15.

  Raciste, P., Morone, C. and Kruelle, W. (2000) Alsatians and Labradors have different attitudes to dog-beating. Trivia, 4 (3), 1215–1320.

  ThynKin, P.P.O., SeyYing, M.G. and Doowing, D. (1978) An assessment of the validity of measuring cultural variations by questionnaire. In R.S. Biggott, (Ed.) Cultural Variation (pp. 114–190). Ohio: Worthless Books.

  Wikked, H. and Hartless, P. (2001) A survey of poodle-drowning practices amongst the Welsh and Swedish. Marie Eclair, 101 (2), 200–203.

  Box 9.1: Things to check that you have covered in your report

  When you are writing your lab-reports, check that you have covered the following points.

  General:

  Write clearly and simply, but in a formal style, using the passive voice. (e.g. ‘an experiment was performed’ rather than ‘we performed an experiment’).

  Title and Abstract:

  Give your report a clear and informative title, no more than 10–15 words long.

  The Abstract is a clear summary of the study’s aims, methods, findings and conclusions, all in no more than 150 words.

  Introduction:

  Summarize RELEVANT experimental findings and theories which relate to the aims of your experiment. Use this information to provide a justification for why your experiment is worth doing.

  Outline your proposed experiment.

  Make specific predictions about the outcome of the experiment, on the basis of the literature you have reviewed.

  Method:

  Include sub-sections on: Design; Participants; Apparatus; Procedure.

  Make sure there is enough RELEVANT detail for the reader to be able to repeat the experiment purely by reading your Method section.

  In the Design section, identify the independent and dependent variables, and say whether you used independent measures, repeated measures or a mixed design.

  Make sure you give RELEVANT background characteristics of the sample of participants, as well as saying how many took part.

  M
ake sure the Apparatus and Procedure sections are written in proper English, and not like a recipe.

  Results:

  Make sure you have clearly described the results and explained whether the evidence (in general) supports the hypothesis under consideration. Describe them, but leave interpretation (in terms of relationship to theories and previous experimental work) until the Discussion section.

  If you have a fair amount of numerical data, put it in a table or graph, whichever seems clearest.

  Number your tables and figures so that you can refer to them in the text. Figures and tables are numbered independently of each other, so if you have, say, five graphs or diagrams and three tables, these would be numbered as Figures 1 to 5 and Tables 1 to 3. Even if you have just one table or figure, refer to it as ‘Table 1’ or ‘Figure l’ – rather than using phrases like ‘the graph shows . . . ’

  Make sure each table or graph is clearly labelled and has a self-explanatory title.

  Make sure tables and graphs are intelligible without reference to the text, and vice versa.

  Where inferential statistics are used, indicate the statistic that was used (e.g. t, F, etc.). Give the value of the statistic used, the number of degrees of freedom, the level of significance reached, and whether the test was one-tailed or two-tailed (see Box 5.3).

  Put raw data and statistical calculations in an appendix, not in the main text.

  Remember to include means and standard deviations (or medians and ranges or semi-interquartile ranges, if these are more appropriate).

  Discussion:

  Summarize your main results.

  Provide some interpretation of what your results mean, in theoretical terms.

  Indicate clearly whether or not your initial hypothesis has been accepted.

  Discuss your own data with reference to other experimental findings and theories in the area, particularly those summarized in the Introduction.

  Identify potential problems with your study, but don’t produce a litany of trivial criticisms. Make intelligent suggestions for future studies.

  References:

  Give only the surname(s) of the author(s) and the date of the relevant publication in the text, unless you are acknowledging the source of a direct quote – in which case give the number of the page on which the quote can be found.

  In the reference section itself, at the end of the report, give the references in the correct format (see Chapter 14).

  If a source you have used (e.g. Smith, 1991) cites an author to whom you wish to refer (e.g. Bloggs, 1950), it must appear in the text as follows: ‘Bloggs (1950, cited in Smith, 1991)’. Smith (1991) should be the reference which appears in the list at the end of your report, not Bloggs (1950).

  10 General Points When Writing a Report

  * * *

  Having to produce lab-reports is one of the big bugbears about psychology for many students, especially at the beginning of their course. This is especially true for students with non-science backgrounds. Some regard writing a lab-report as a tedious chore; they consider that it’s non-creative and that the formal, jargon-laden ‘scientific’ style is an affront to their writing skills, honed previously on insightful criticisms of ‘King Lear’ and the Metaphysical Poets. Others are simply scared by the whole idea: it appears to be dry, ‘scientific’ and hard, when all they really wanted to do was find out what makes people tick.

  10.1 The Standardized Format of the Report

  Actually, writing a report is none of these things: it can be a highly creative exercise, it should be well written, and it’s not really as difficult as it looks. Reports have a clearly specified format, and much of report-writing is a bit like painting by numbers – you just fill in the blanks with the required stuff. (Using words rather than blobs of paint, preferably).

  The standardized format of the psychology lab-report is based on the format used by academics when they publish their findings in academic journals. Journal articles are one of the main ways in which psychologists communicate their ideas and discoveries to their colleagues. (Another way is in drunken conversations at conferences, but we won’t go into that). Each issue of a journal consists of a number of articles, each written by a researcher or group of researchers. A typical article consists of research findings, prefaced by a brief explanation of the background to the research and followed by a discussion of what the study’s results contribute to our knowledge in that particular area of psychology. The article also provides details about the methods, procedures and apparatus that were used, so that other researchers know exactly how the results were obtained and would be able to replicate the study if they so wished.

  Once you are familiar with the conventions that are used, the standardized format enables you to assimilate the information quickly and efficiently. The title of the article should give you a good idea of the specific topic of research; the article itself is preceded by a summary (‘abstract’) that gives you a quick overview of what was done and what was found; and reading the article itself is facilitated by its standardized presentation. The reader knows exactly where everything is likely to be within the report: if they want to check up on a matter of procedure, or remind themselves of the authors’ findings or conclusions, they can go straight to the appropriate section of the report to find the relevant information.

  Where to Find Guidance on the Formal Aspects of Report Writing

  When in doubt about any aspect of how to write a practical report, there are two ultimate sources of authority. The first and most accessible are journal articles: just go to the library, pick up a journal and see how ‘real’ researchers do it. (Although you don’t have to make your report as tedious as some of the ones you might find!) Don’t worry too much about the technical bits, the scary statistics and so on: just aim to get a feel for the general style in which articles are written. Pretty much any journal will do, as they all use much the same format, but the ‘British Journal of Psychology’ or the ‘Journal of Experimental Psychology’ spring to mind as good role-models.

  The second and definitely the most authoritative source of information are the style guides produced by the British Psychological Society (BPS) and the American Psychological Association (APA). These contain detailed information on every aspect of report writing. Most journals expect submissions to conform to these guidelines: in fact, many American journals will not even look at a manuscript unless it adheres to ‘APA format’. The APA are currently up to the fifth edition of their ‘Publication Manual of the American Psychological Association’. This is, believe it or not, 439 pages long, and it covers every aspect of report writing from Abbreviations to Zero (the latter is covered on pages 126 and 294: I just thought I’d mention that in case you’ve been losing sleep over nothing. Nothing? Zero? Oh, never mind!) While some of it verges on the pedantic, much of it is sound advice, and there are some excellent tips on how to express yourself clearly.

  Obviously we can’t cover everything that the APA manual says, because that would make this book twice as thick as it already is. However, in the sections that follow we will mention some of the manual’s more important suggestions.

  In the rest of this chapter, we will outline the various sections that should be present in a practical report; tell you what should (and shouldn’t!) go in each one; and finally, give you a specimen report that you can use as a rough model for your own write-ups.

  10.2 Some Important Considerations When Writing a Report

  Keep the Audience in Mind

  When writing the report, try to keep your audience in mind. A useful strategy is to imagine that you are talking to someone, and explaining what you did. (Ahah! Finally those voices in your head can be put to some use . . .) . This ‘someone’ is an interested, intelligent person who knows something about psychology, but isn’t familiar with the area in which you have performed your experiment.

  When writing a report, you have to continually bear in mind that the reader only knows what you have told them up to that
point. They shouldn’t have to read later parts of the report to make sense of the earlier sections. This can become a little tricky when it comes to describing the study’s methods and procedures: if they are complicated, it’s sometimes hard to describe them in a way that doesn’t presuppose that the reader already knows what you did before you tell them! In these cases, try giving a draft to a friend who doesn’t know what the study involved, to see if they can make sense of it.

  Substantiate Your Claims and Assertions

  You should imagine that your hypothetical non-specialist psychologist is highly sceptical, so that everything you say needs to be supported. In the Introduction, if you make claims you should back them up with references to previous research. In the Results section, if you say there are differences between conditions in your experiment (or even if you say there are no differences!) you should support these claims by referring to the results of statistical tests that you have performed on the data. In the Discussion, you support any conclusions that you draw about what the results show, by reference to your own results and, where appropriate, to those of previous researchers (in a similar way to how you did it in the Introduction).

  When supporting claims and assertions by referring to previous work, don’t fall into the trap of appealing to authority: sometimes students write statements like ‘but Tardive’s work is rubbish, according to the esteemed and highly respected psychologist Professor Doodle’. Other people’s unsubstantiated opinions are as irrelevant to science as are your own. If Doodle has presented evidence that Tardive’s work is problematic, then by all means cite this in support of your argument. Mere appeals to prestige have no place in science.