I just finished reading Cooper’s book on the topic and would very much like to write down what I get out of it.
Cognitive dissonance was first envisioned by Leon Festinger in the 1950s. The theory started out elegant and simple: With consistent cognitions, individuals experience uneasiness and discomfort thus are motivated to reduce the inconsistency. Here cognitions can refer to actions, beliefs, attitudes, and so on. Inconsistency can be due to inconsistent attitude and action. One of the many experimental paradigms to test the theory runs as follows: subjects are asked to evaluate a certain proposition arguing against her/his position. Lastly attitude is elicited again. For instance, subjects (usually college students) are asked about their attitudes towards increasing tuition. Expectedly most of them are against it thus are asked to write an essay/make a speech for tuition increase after which they are asked again about their attitudes on the issue. (All results will be discussed under this paradigm for exposition.) Typical findings on these experiments show that subjects change their attitudes closer to what they (are asked to) advocate. According to Festinger, subjects experience cognitive dissonance: writing an essay for tuition raise is inconsistent with their attitude against a raise. As a result of the dissonance, subjects are motivated to reduce the inconsistency. In this case (as in many other cases), changing one’s attitude is easier than changing a committed action.
The theory is appealing as it explains, elegantly, experiments and observations. However humans are hardly that simple (thus the struggle of social scientists). In the following years, many experiments and theories are raised to clarify and complete the theory. I will first discuss efforts on verifying the physiological manifestation of dissonance and then talk about modifiers found that largely rewrite the whole story.
One of the first questions raised against the theory concerns the motivated drive (Chapter 3 of the book) Festinger highlighted in his original composition. Simply put (especially intuitive to economists), psychologists wondered if people really experience tension states/arousal/emotions or anything with a physiological trait thus are motivated/driven to reduce inconsistent or is it just a metaphor/as-if argument. With the hep of misattribution and two-factor theory, experiments show that the uneasiness is experienced
- Two-factor model of Schachter and Singer (1962): an experience of emotion is based on two factors: arousal and the attachment of a cognitive label to that arousal. They conducted experiments to show that the factors are separable in the sense that how the arousal is experienced is affected by the label provided. In short, the experience of emotion is labile.
- Back to dissonance: subjects who take placebo but think they take a drug which might cause uneasiness and discomfort, do not change their attitudes as much after the essay task. Similarly, subjects who think they take a drug that might cause relaxation change their attitudes more after the task.
- These experiments show that when provided with a convenient label of their discomfort, subjects do not attribute the arousal to inconsistent cognitions thus make no effort in reducing consistency.
These are good news for the theory which solidify the original ‘guess’ of Festinger. Now the ‘bad’ news: New Look Model and Self-Standards Model (SSM henceforth) that largely kill the original idea of the theory. SSM is basically an extension of the former; some years passed before the birth of SSM but I will skip the struggle and talk directly about SSM as if it were a success in one shot (but it is really not).
- The following experimental results forced researchers to rethink about the theory: No/little attitude change is observed if (1) subjects are not presented with a choice, (2) the action does not result in any aversive and foreseeable consequence, or (3) the behavior is expected by some (social/personal) standard. For the essay task,
- Attitude changes if subjects are asked if they would like to write such an essay (high-choice treatments) and no changes if subjects have to do so (low-choice treatments).
- Attitude changes if the essay is to be presented to the college committee and no if it is destroyed after completion. Attitude changes if the essay might be viewed by the committee and no if it comes as a complete surprise (unforeseeable).
- This point cannot be demonstrated with the paradigm and plays an important role in turning Festinger’s theory up-side-down. So I will discuss it in more details in the next entry.
Just as a side note: the experiments become more intriguing as the theory gets more complicated. The levels (plural really) of deception used in these experiments stunned me, an economist who sleeps with the dogma that deception is not allowed in experiments.
Okay this entry is getting too long to bear even for myself. I am gonna end here.
Cooper, Joel. Cognitive dissonance: 50 years of a classic theory. Sage, 2007.
P.S. This is a very nicely written book: well-organized and fun to read. Perfect as an introduction into the theory.