SKINNER 1974 comments on Rogers

SAVE OFFLINE

Psychology Is About" (Hebb[mal], February 1974). The latter article mentioned the early Russian psychologist Shenger-Krestovnikova, of whom I have never heard. I am nevertheless sure that this psychologist was female because the Russian language distinguishes between male and female last-name endings. This suggests a much-needed repair to our method of citation. We should add, at the end of each name, a suffix to indicate sex. The suffixes that immediately come to mind are -ma, for male, and -fe, for female. But perhaps my male chauvinism is showing, for the female suffix is quite similar to the Yiddish expletive for dirt. Alternatively, we could use -mal and -fern, although the male suffix would suffer from "bad" connotations. The second set of suffixes seems preferable, and I urge their adoption. REFERENCES Bernstein, M. D., & Russo, N. F. The history of psychology revisited: Or, up with our foremothers. American Psychologist, 1974, 29, 130-134. Hebb, D. O. What psychology is about. American Psychologist, 1974, 29, 7179.

E. J. KAY [MAL] Lehigh University

Comment on Rogers I was surprised to hear Carl Rogers (February 1974) say that I have "cheated the profession" by refusing "to permit the nine-hour confrontation we held at the University of Minnesota to be released." As I remember, it was not my permission but my advice that was asked for, and I said that the quality of the recording was so wretched that I did not feel it fair to offer the tapes for rental or sale. I have no objection whatsoever to the tapes being made available if anyone is willing to listen to them. I have heard a recording made by a member of the audience which is much clearer and might be a better candidate for release.

I do not, however, relish the prospect of editing my share of a ninehour transcription, and I should not want a transcription circulated without editing, particularly considering the bad recording. That was certainly never part of our original agreement. REFERENCE Rogers, C. R. In retrospect: Forty-six years. American Psychologist, 1974, 29, 115-123.

B. F. SKINNER Harvard University

Some Ambiguities in the Escape from the Ambiguous This note is a comment on Sheehan's (November 1973) article. This comment is divided into two parts: The first part is based on what would ostensibly appear to be a reasonable interpretation of Sheehan's account; the second part is based on a consideration of the experimental techniques employed by both Barber (1969) and Sheehan. Predominantly, our concern is with Sheehan's critique of what he called the operational paradigm or model. His critique was essentially an analysis of Barber's (1969) experimental procedure. Sheehan maintained that Barber's model contains an artifact; we maintain that the term artifact was used too loosely by Sheehan. Artifact refers to a specific kind of error, and, although it was evident that Sheehan had legitimately found a fault in Barber, this fault is of a different nature. First, artifact arises typically in the following situation: An experiment is conducted in which there is an experimental group and a control group. These groups by definition differ only with respect to the presence of the experimental factor in the experimental group. All other factors in the two groups are supposed to be the same. If experimentation demonstrates a difference on the dependent variable between the two groups, the difference is

640 • AUGUST 1974 • AMERICAN PSYCHOLOGIST

attributed to the experimental factor. If it turns out that the experimenter, in exemplifying (Campbell, 1969, p. 358) the experimental factor, unwittingly introduced some aberrant or unspecified independent variable such that the difference in the dependent variable can be shown to be produced by this ignored factor, we say that the original result was artijactitious. Thus, experimenter bias, or demand characteristics, or the Hawthorne effect are elements unintentionally introduced into experimental stituations and have subsequently been shown responsible for differences in dependent variables. Heuristically, one might view Barber's recent studies (for a summary, see Barber, 1969) as purporting to expose an artifact. Barber pointed out that results that appeared to be due to hypnotic trance were artifactitious. Among the factors prerequisite to the production of a trance, one finds something called task motivation. In numerous experiments, Barber presumed to show that when task motivation was added to the control condition, the superiority of the hypnotized subjects disappeared. Orne's real-simulating model discussed by Sheehan (1973) is a typical example of a situation in which artifact arises. Although both Sheehan and Orne referred to the simulators as "quasi-controls," this seems an unnecessary qualification once it is made clear that the quasi-control is really the experimental group. What makes the situation look different is that Orne wanted to show a faking instruction to be an inhibiting factor so that the "experimental" simulators would do "worse." The simulators were treated in every way like the hypnosis subjects except for the additional instruction to fake. The instructions to fake can be regarded as the experimental factor. Furthermore, the artifact, the so-called simulation effect (Sheehan, 1973, p. 984), was found to be an element unwittingly introduced with the instruction to fake, again fulfilling the desideratum for normal artifact situations.