Experimentation in social psychology: A reappraisal


  • This article is based on an invited address to Division 8 at the 1975 meetings of the American Psychological Association in Chicago, Illinois. I wish to thank Curt Banks, Michael Basseches, Donald Campbell, Uriel Foa, Mary Gergen, Robert Helmreich, Clyde Hendricks, Ian Lubek, Serge Moscovici, Franz Samelson, Paul Secord, Philip Shaver, Siegfried Streufert, Karl Weick, and Ricardo Zuniga for their valuable commentary on an earlier draft of this paper. Special appreciation is also expressed to Robert Pages for making the facilities of the Laboratoire de Psychologie Sociale, Universitt de Paris VII, available for completing the final manuscript.


Psychological inquiry into social phenomena has become virtually indistinguishable from controlled experimentation. Although the assets and liabilities of psychological experiments have been subject to periodic debate, a continued increase in the reliance placed experiments is evidenced. The present paper re-examines the adequacy of experimentation in light of major features of social interaction. Significant failures of the experiment emerge when the following characteristics of social events are considered: their imbeddedness in broader cultural patterns, their position within extended sequences, their open competition within real-life settings, their reliance on psychological confluences, and their complex determination. The additional consideration of social phenomena within historical context indicates that all reasonable hypotheses are valid and that critical testing between hypotheses about social behaviour is fruitless. Criteria for the productive usage of experiments are detailed.