Anthony House
Psychology 002-01
Prof. Howard
11 December 2000

The Perceived Ethical Merit of Psychological Experimentation: Considering Variations in Species Studied and Inolved-Party Pronoun Label.

Abstract

An experiment is reported in which ninety-three students are asked to rate their perception of the ethical merit of another, hypothetical experiment. The vignette read by respondents describes an experiment involving chimpanzees or humans, who are referred to using one of two pronouns: subject or participant. The results are, for the most part, inconclusive, though respondents rated the experiment as less ethical when it involved humans than when it involved chimps.

Text

In the past quarter century, psychology has gone through an ethical revolution of sorts. Regularly and consistently implementing safeguards to protect the rights of those parties studied, the psychological community—along with scientists in other disciplines—has gone to great lengths to ensure fairness to all involved in research. The last twenty-five years have also witnessed the continued ascension of media within American society. Today, mass media operates through outlets such as CNN and the internet, inundating the average person with information. In tandem, these two trends create a dynamic in which the ethical battleground is not restricted to the laboratory; indeed, popular opinion has far-reaching implications in the scientific community. Under such a system, the media plays an undeniable role in shaping the opinions of the masses. It is therefore in the best interest of science to ensure that the language used to describe its research is both accurate and ethically secure.

Though no one has studied the impact of involved-party pronoun choice on the reaction to psychological research, there has been a move in recent years to supplement the term "subject" with the more personal "participant" in certain cases. In the Dictionary of Psychology, Corsini suggests that "the difference between these two terms has to do with the degree of understanding, willingness and agreement on the part of the individuals in an experiment" (Corsini, 1999). He goes on, however, to point out that "some psychologists use the terms interchangeably" (ibid.). Intentionally using the two labels as synonyms nullifies any definitional nuances between the terms. Still, the question remains: do the two terms convey different meanings to the consumers of research—whether peers within the scientific community or lay people? Much research has been done on the way labels affect the general perception of a person. Sitka, for example, found that party label to be the "strongest predictor of voting behavior" (Sitka 1997). Kite’s study on the interaction of gender, age, and occupational roll labels found that a single label often implied other information: "employed targets were most often imagined to be male and young homemaker targets were most often imagined to be female" (Kite, 1996). Similarly, Phelan discovered that the label of homelessness carried as much stigma for the poor as the label of mental illness (Phelan 1997).

Additionally, the world at large has become increasingly aware of the ethical concerns at stake in the treatment of those studied, both animal and human. This poses another question: does species matter to the lay perception of ethical merit, other things being held constant? Looking at these two questions at once requires consideration of the interaction between the two. Specifically, does the species studied in an experiment produce a greater effect (on its perceived ethical merit) than variations in the language used to describe those studies?

The study at hand presented an abbreviated statement of method from a hypothetical psychological experiment. The hypothetical experiment related in the survey included the administration of a "mild electric shock" to a portion of those studied. The vignette read by respondents varied in two ways: the species studied and the pronoun/label used to describe those studied in the hypothetical experiment. About half the respondents read the vignette as involving chimpanzees, while the other half were told the study involved humans (specifically, psychology students). Likewise, about half those responding were presented a study wherein the chimps/students were referred to universally as "participants" while the remainder read a vignette that labeled the chimpanzees/students as "subjects." In all other respects, the vignettes were identical. Respondents were asked to "rate the ethical merit of this study" on a 9-point scale, and were given the opportunity to comment on the hypothetical experiment.

If, in fact, labeling those studied as "subjects" dehumanizes them, we would expect that respondents rate the student-subject vignette as more ethical than the student-participant vignette. We also expect that the chimpanzee vignettes will be rated more ethical than the student vignettes, though we do not anticipate any significant difference between the perceived ethical merit of the chimp-subject and chimp-participant vignettes.

Method

Respondents

Ninety-three students in an introductory psychology class participated voluntarily, en masse. No compensation was offered to the respondents.

Materials

Each respondent was given a survey form relating an abbreviated statement of method from a fictitious experiment. In the fictitious experiment, some of those studied were given mild electric shocks. There were four versions of the experimental vignette, one corresponding to each cell in the 2 × 2 factorial design of the experiment at hand. While identical in all other regards, 47 of the surveys depicted the study as focusing on chimpanzees while 46 used humans (introductory psychology students, in fact.) Similarly, 46 the surveys labeled those studied as "subjects" while the remaining 47 referred to them as "participants." A version of the survey—indicating the points of variation corresponding to the 2 independent variables—is attached as Appendix B: Master Copy of Survey.

Procedure

After being told that the purpose of the study at hand was to explore the general perception of ethical behavior within the psychological community, the respondents were given the surveys. The respondents were asked to disclose their sex and age, and were then asked to "read the following statement of method…and respond to the question below." The question asked the student to "rate the ethical merit of this study" on a 9-point scale—1 representing completely unethical, 5 representing neither ethical nor unethical, and 9 representing completely ethical. The entire experiment required fewer than 5 minutes.

Results

Table 1 Shows the frequency distributions of each of the 4 cells in the study, along with the independent frequency distributions for each of the 2 independent varibles. It is worth noting that the range of ratings provided by respondents was identical for the Chimp/Subject and Student/Subject conditions, and only slightly varied for the other two cells. Table 2 presents descriptive statistics for the data collected. The means for both cells including students are lower than those cells including the chimp level of the species IV. Still, none of the differences in mean perceived ethical merit are statistically significant; c²(24)=18.56, p > 0.05; f(3, 89)=0.17, p=0.68.

Table 2. Means (Standard Deviations) by IV.
	Chimp	Student	Rows
Participant	5.08 (2.08)	4.17 (1.63)	4.64 (1.93)
Subject	5.04 (2.24)	4.48 (2.20)	4.76 (2.24)
Columns	5.06 (2.16)	4.33 (1.95)	4.70 (2.09)

Discussion

While encouraged by the data’s conformity to our predictions, we point out that the differences in perceived ethical merit were not statistically significant. This is, of course, a source of some frustration, and we feel that a more carefully designed study—perhaps including more subjects or utilizing variation within subjects—could result in findings of significance. Nevertheless, our inability to reject the null hypothesis in this experiment may just reinforce Corsini’s acknowledgment that some "use the terms interchangeably" (1999). By reducing the terms to synonimity both in the field journals and in the popular press, Psychology runs the risk of losing the definitional nuances that could otherwise be important.

Further investigation would, of course, be necessary to determine whether, as Corsini suggests, the use of "subject" or "participant" implies a certain degree of consent and understanding granted to those being studied. Regardless, however, it seems as if the development of guidelines to distinguish between the two labels would be helpful, even if such guidelines pull the terms into technical jargon of the field.

Works Cited.

Corsini, Raymond J (1999). The Dictionary of Psychology, Philadelphia: Brunner/Mazel.

Kite, M.E. (1996). Age, Gender, and Occupation Label: A Test of Social Role Theory. Psychology of Women Quarterly, 20, 361-374.

Phelan, J., et al. (1997). The Stigma of Homelessness: The Impact of the Label "Homeless" on Attitudes Toward Poor Persons. Social Psychology Quarterly, 60, 323-337.

Skitka, L.J., Robideau, R. (1997). Judging a Book by Its Cover: The Effects of Candidate Party Label and Issue Stands on Voting Behavior. Journal of Applied Social Psychology, 27, 967-982.