Reducing Laboratory Dissonance

As part of my research on extreme ritual practices I’ve been trying to reacquaint myself with the theories concerning ‘cognitive dissonance’ pioneered by Leon Festinger back in the 1950s. The basic concept of cognitive dissonance is that when we do or think things which contradict something else that we believe, we experience psychological discomfort and thus become motivated to reduce the dissonance. A classic illustration of these processes was provided in an experiment by Aronson & Mills (1956) in which they varied the ‘severity’ of initiation costs for joining a discussion group.

Aronson & Mills told their young female college student volunteers that they would be joining a discussion group concerning sex and would therefore need to be comfortable with sexual topics. For the control group accepting this was the only requirement but for the experimental conditions they had to demonstrate their comfort with the topic by reading out loud to the experimenters, five words related to sex (mild condition) or twelve obscene words related to sex (extreme condition). Considering this study was conducted in the 1950s this was likely a lot more embarrassing than it would be today. Also, the participants would have been less familiar with psychologist’s shenanigans and hence, more likely to earnestly comply with the instructions. Upon being granted access to the discussion group all participants were treated to a recording of an (intentionally) very dull discussion about animal sexual behavior and then asked to report how interesting they found the group, how much they liked the group and so on. The results showed no difference between the mild and control groups but indicated that those who had performed the extremely embarrassing initiation rated the discussion group as being much more interesting and enjoyable. This counterintuitive finding fits well with the predictions of cognitive dissonance because the participants who endured the severe unpleasant initiation would (in real life) normally only do so for something worthwhile. Consequently, they were the group most motivated to reevaluate their experience and make it more congruent with their expectations.

There have been hundreds of papers on dissonance theory since the 1950s, some criticising it and others testing and refining it. On balance, I would say that the fundamental claims of the theory remain well supported but that they are now surrounded with caveats and qualifications. However, an interesting nugget I found when delving back into this literature was from a transcript of remarks made by Leon Festinger back in the late 80s, in which he reflected on how research on his theory had progressed in the preceding 30 years. This transcript is contained in an Appendix of Cognitive Dissonance: Progress on a Pivotal theory in Social Psychology (Harmon-Jones & Mills, 1999) and reflects quite an ambivalent and critical opinion on how his work has been interpreted. More interestingly, it also contains a far-sighted plea for the value of ‘real world’ field research, which begins with a nostalgic lament for the loss of ability to create ‘real worlds’ in the lab:

One thing that I think has to be done is for more research on dissonance producing situations and dissonance reduction processes as they occur in the “real world”. I put it “real world” because Elliot is quite correct. In the good old days when you did laboratory experiments, we created a real world in the laboratory. I don’t know how we would have gotten anything through ethic committees.

Festinger then goes on to advocate for the need for experiments to move out of the now overly regulated laboratory environment and to start dealing with the ‘messy and difficult’ real world. Providing a very clear explanation, one that I would fully endorse, about why we need research to be ongoing both in the lab and out in the real world.

One of the things about laboratory experiments is that you can only get out the stuff you put in and any good experimenter who is concerned in testing a hypothesis is going to try and eliminate from the laboratory experiment all the unwanted stuff that generally floats around, and dissonance reducing processes are not the only things that affect man, using man in the generic sense. I think we need to find out about how dissonance processes and dissonance reducing processes interact in the presence of other things that are powerful influences on human behavior and human cognition, and the only way to do that is to do studies in the ‘real world’. They’re messy and difficult. You don’t expect the precision out of those studies that you can get in the laboratory. But out of them will emerge more ideas which we can then bring into the laboratory to clarify and help to broaden and enrich the work.

— C. Kavanagh

Did you like this? Share it:

3 thoughts on “Reducing Laboratory Dissonance

  1. Paul Jackson

    I am very interested in this. Have you considered other phenomenon in the cognitive bias umbrella such as confirmation bias, memory errors, etc.?

    These ideas have had a tremendous impact on my world view, and how I perceive current events. Generally, we believe what we want to believe, and are incredibly skilled at rationalizing our positions. The only reasons for optimism that I have found are 1) that people also dislike contradiction and 2) the fact that historically, generation by generation, we do seem to be getting better, slowly, by degrees.

  2. Dominik Lukes

    This is certainly a useful approach. But I’d suggest that you also need a theory of cognitive consonance, i.e. the way in which people make sense of things around them. I’ve been doing research on just that in the context of metaphor negotiation in discourse ( and I think that while cognitive dissonance is a useful concept, it is just a facet of the underlying cognitive apparatus. This cannot be particularly well studied in the lab context which can never give you the one things crucial to sense making which is time. These phenomena play out both at the microscopic and macroscopic timescales (weeks, years, and lifetimes). They also take place in the context of social relationships and the processes of their maintenance. These things can only be hinted at in the lab and often escape the short-term ethnographic probe. But they do sometimes come up when you study discourse.

Leave a Reply