cause. It became clear during the 1960s that many conclusions could
not be supported by research from other laboratories or from our own
replication. The study of attitudes, beliefs, opinions, and the persuasive
process then experienced the same reduction of generalizability of
results and the accompanying narrowing of experimental focus that
other experimental efforts (e.g., those of Clark L. Hull) suffered in the
late 1940s. It was in the 1960s that a number of social psychological
experimenters began discovering crucial difficulties in their experi-
mental procedures. These difficulties were not of the usual variety,
involving calibrations of various kinds or changes in procedure in
order to better elicit a response from the subject. Rather, social psy-
chologists discovered that the subject entered into the experimental
procedure in a way that had not been expected. In an experiment,
variation in the subject is either controlled directly by the experimenter
by manipulation of the relevant independent variables or is presumed
to vary randomly over a number of subjects. Difficulty arose when
some of those uncontrolled variables that were thought to vary ran-
domly, were responses from the subject directed toward being in the
experiment itself. In short, by interpreting various aspects of his or her
participation in an experiment, the subject introduced a confounding
variable that rendered the results of the study either meaningless or
grossly distorted. Sufficient interest in these distorting variables was
generated to produce a body of research on artifacts in social psycho-
logical experimentation. This research was summarized and discussed
by a number of involved social psychologists in Robert Rosenthal and
Ralph Rosnow now well-known Artifact in Behavioral Research, which
appeared in 1969. The suspiciousness of the subject of the experiment-
er's intent, whether or not a subject had volunteered to appear in a
study, whether or not a pretest was used to evaluate some characteristic
of the subject before an experimental treatment was applied, what the
subject believes is required of him or her in an experiment, what the
subject believes the experimenter expects of him, and whether the
subject is anxious about participating in an experiment, all appeared as
factors that could distort the meaning of the experimental results. In
the decade of the 1970s, research continued on artifacts in the experi-
mental process, further refining the nature of the difficulties more often
than eliminating them. Since the late 1960s, social psychologists have
had a choice; they could continue to perform experiments in order to
eventually minimize or eliminate the effects of unwanted subject influ-
ence in the social psychological experiment, or they could concentrate
on the very nature of the experiment itself and its place in a field that
purported to discover information on the nature of social existence.
One could also, of course, do both, but with separate efforts. This crisis

-3-