Cognitive Dissonance (as per say Aronson instead of say the original Festinger definition) is a phenomenon whereby if one indulges in a behavior that is not inline with one’s normal schema about oneself, then the discrepancy between the actual behavior and one’s own self-perception leads to a dissonance (or state of unrest/anxiety), such that one modifies one’s internal schema/ self-perception to make it inline with outwardly exhibited behavior. The classical case is a celebrity (who normally smokes) endorsing a public commercial for ‘no smoking’. Just after publicly asking others not to smoke, the celebrity would feel a Cognitive Dissonance and to reduce that uneasy feeling may decide to change his/her self-image to be that of non-smoker and this may lead to him/her smoking less in future. If the incentive to indulge in the hypocritical behavior was less (say the endorsement of commercial was for free and the celebrity did not take an excessive amount of money to endorse the ad, then this would lead to more dissonance and vice versa.
There is an article on BPS related to an observation that drug usage likelihood increases subsequently if people are asked about drug usage likelihood earlier
This study leads to several questions.
How does this observation relate to Cognitive Dissonance effect? When the respondents were asked about likelihood of drug usage and (supposedly) they replied in the negative (to please the surveyor), then this small incentive to lie should have ideally led to a large Cognitive Dissonance and prompted them to stop using drugs and led to change in drug-usage behavior to overcome the dissonance experienced. Is some data available as to whether those who reported more drug usage later had replied in affirmative or in negative to the earlier drug-use-likelihood question? Is there a priming effect so strong that it is shadowing the cognitive dissonance effect? This study raises more questions than it answers!