The above problems in subject selection begs the question, how biased are social psychologists in choosing subjects to participate in their experimental conditions? And is this biasing a universal problem or is it just associated with research in certain countries.
METHOD
In order to answer the above questions I analysed the types of subjects used in all the articles published in three of the most prominent Social Psychological journals, during the years of 1981 and 1991. The journals used were, British Journal of Social Psychology (BJSP) vol.. 20 and 30, European Journal of Social Psychology (EJSP) vol.. 11 and 21, and the Journal of Personality and Social Psychology (JPSP) vol.. 40 and 60. My reason for examining these three different journals was to see whether there was a variation between the subject population used in different continents. I also wanted to discover whether there had been any changes made in subject selection over a ten year period, which is why I examined all the articles published in both the 1981 and 1991 editions of the above journals.
FINDINGS
I coded the subject population into eight categories (a) university undergraduates; (b) school children; (c) other students, (these included F.E college students, student teachers and nurses etc.); (d) members of the general public; (e) members of the public service sector (mainly police, nurses, health care teams and judges); (f) families; (g) psychiatric patients; (h) others ( writers,U.S Presidents, religious groups, prison inmates etc.). The proportions of the subject populations used in the above categories for all the journals articles surveyed, can be seen in Table I.
TABLE I
The sorts of people used as subjects in articles published
in three prominent Social Psychology Journals.
________________________________________________________________
SUBJECTS BJSP EJSP JPSP
1981 91 1981 91 1981 91
________________________________________________________________
University
undergraduates 53% 49.9% 70.7% 50% 71% 79.2%
School children 23.5% 32.2% 12.5% 21.9% 9.8% 3.9%
Other students 8.8% _ _ 9.4% 1.0% _
Members of the
general public _ 10.7% 6.3% _ 4.1% 9.9%
members of the
public sector _ _ 4.2% 6.2% 5.0% _
Families _ 3.6% _ _ 1.0% _
Psychiatric patients _ _ _ _ _ 2.0%
Others 5.9% 3.6% 6.3% 12.5% 8.1% 4.9%
________________________________________________________________
The percentages were arrived at by adding up the total amount of times subjects were used in each category, then adding up all the category totals and dividing the amount by a 100, then by multiplying each category total by this amount to arrive at a percentage total.
Inspection of Table I reveals that a majority of researchers use a high proportion of students as subjects, with American psychologists using a particularly high percentage. Also the proportions of subjects used in each category has not fluctuated much over the last ten years. On the other hand British psychologists do appear to use far fewer students as subjects than do psychologists from the other countries. Over all, there is an extremely small percentage of studies conducted on members of the general public.
DISCUSSION
I shall now go on to discuss the implications of using such a large proportion of students in research. The first problem I shall address is the fact that such a high percentage of studies are conducted on such a small percentage of the general population, for example, in Britain university students only account for around 3% of the total population yet constitute 50% of the subjects used in psychological research. The group also provides, what is probably, a poor representation of the population as a whole. Members for instance generally possess a higher than average level of intelligence, and the majority belong to a fairly narrow age group. What's more the middle and upper classes are represented to a disproportionately high degree. These factors would also apply to the American and European students who were cited in the journal articles.
It has been suggested that if researchers want to increase the validity of research conducted on students, which is then generalised to the public they need to be able to identify the ways in which students in the laboratory differ from people in everyday life and extrapolate the differences between these two. An alternative, and perhaps a more feasible solution would be to conduct their research on a broader range of subject population.
There is a further biasing effect, which may also limit the degree to which student data can be generalised. This is associated with the differences found between the performance of true volunteers and students who have been coerced into participating in research. During my search through the journals it became increasingly apparent that a majority of the subjects (around 90%) who took part in research papers published in the (JPSP) did so because it was a requirement of their psychology course. This did not seem to apply to subjects used by researchers whose papers were published in the other two journals. It is still not clear to what extent, if at all, this factor might bias psychological findings, but it does need to be considered when generalising research findings.
In summary it appears from my own study that a majority of Social Psychological research is conducted on undergraduate students, who are quite often enrolled in psychology courses. In reply to the questions I proposed at the beginning of this essay I would have to conclude, from the journals I have surveyed, that social psychologists are biased in their choice of subjects, but that this could be blamed on the accessibility of potential subjects. It would also appear that this subject biasing is a widespread problem. Finally it would be unwise of me to generalise my findings to all Social Psychologists research practice, since it might well be a prerequisite of the editorial journals I surveyed that a traditional mode of research should be used in all papers which they publish, hence the large proportion of student/subject population.
BIBLIOGRAPHY
Oakes, W (1972) External Validity and the use of real people as subjects. American Psychologist Vol.27 P959-962.
Sears, D. (1986) College sophomores in the laboratory Journal of Personality and Social Psychology Vol. 51 P515-30
Schultz, D (1969) The human subject in psychological research Psychological Bulletin Vol. 72. P214-218.
REFERENCES
British Journal of Social Psychology vol. 20 and 30.
European Journal of Social Psychology vol. 11 and 21.
Journal of Personality and Social Psychology vol. 40 and 60.
