If you are a bit interested in psychology and especially in social psychology, I think that the first thing you will think of when hearing the name of Leon Festinger is “cognitive dissonance”. It is the central concept in a theory that he developed with his team. In a nutshell, the theory says that we try to adapt our interpretation of the facts to our beliefs if the facts don’t fit the beliefs, while to an outsider the other way round would seem more rational. Of course, adapting your beliefs (and the actions that follow from them) to the facts is also a kind of dissonance reduction, but adapting the interpretation of the facts to the beliefs happens so often and is so remarkable since it seems so illogical, that the theory of cognitive dissonance has become almost synonymous with a theory that explains this irrationality. To give an example, when a smoker reads a research report on the bad effects of smoking, of course, he can say “I’ll quit”, but there is a big chance that he’ll think that the research is not right or that there are also positive effects of smoking, for instance because his grandfather, who was a fervent smoker, has become hundred years old, or which other positive reasons for smoking may come to his mind. For this blog I’ll understand cognitive dissonance in this limited way.
Festinger is not only known for the theory of cognitive dissonance but also for having promoted the use of laboratory experiments in social psychology and for his methodological contributions to this approach. However, one experiment that brings to light a certain phenomenon is only one experiment, and since an experimenter can make mistakes, in many handbooks on methodology it is recommended to repeat experiments in one way or another. This can be done by replicating the original research as exactly as possible or by trying to get the same results by using a different design or otherwise. If the new research confirms the original results, it has become more likely that the theory tested is true. If it doesn’t, we have a problem, and we have to find an explanation for the difference (the cognitive dissonance has to be reduced, so to speak, if we use the concept in its broad sense, so including the idea that the original theory may be revised as well). Of course, it is possible that the repetition of the original research in one form or another was not correct, but this is only one of the options that may explain the difference with the original results.
Be it as it is, as Ruud Abma notes in an article on replication in psychology, just the latter, namely seeing a replicatory study as imperfect in case of non-confirming results, has become tradition in social psychology. So what did Festinger write in his article “Laboratory experiments” (published a few years before the famous When Prophecy Fails by Festinger et al., in which the theory of cognitive dissonance was expounded)? Indeed, that negative results not conforming to the expectations probably mean that the experiment had not been done in a careful way and that the manipulation of the research variables by the project leader had not been effective. In other words: Adapt the facts to the theory. Is there a better proof of the theory of cognitive dissonance?L. Festinger, “Laboratory Experiments”, in: L. Festinger and D. Katz (eds.), Research methods in the behavioral sciences New York: Dryden, 1953; pp. 137-172.