Cognitive Dissonance
Cognitive dissonance is Leon Festinger's 1957 theory that holding a belief and a behaviour which do not fit together produces a discomfort people move to resolve, and that what usually moves is the belief. Changing your mind about what you did is cheaper than admitting the doing was a mistake. In product design this is the phase where a user stops evaluating a product and starts defending it. The theory is old, enormous, and its most-used laboratory method has a known measurement problem.
What it is
The theory begins with an observation about accounting rather than about feelings. A person holds a great many beliefs and performs a great many actions, and most of the time the two are consistent enough to require no attention. When an action and a belief flatly contradict each other, something has to give, and Festinger's claim was that the pressure to reconcile them is real, unpleasant and motivating in the way hunger is motivating.
The second half of the claim is the one that does the work. Behaviour that has already happened cannot be edited. Beliefs can. So the cheapest available resolution, almost always, is to revise the belief until it fits what was already done. People are considerably better at explaining their past decisions than at concluding those decisions were wrong.
Festinger published A Theory of Cognitive Dissonance with Stanford University Press in 1957. Two years later he and James Carlsmith ran the experiment everyone remembers: participants did an hour of deliberately tedious work, then were paid either one dollar or twenty dollars to tell the next participant it had been enjoyable. The people paid twenty dollars had a reason to lie and did not need to believe it. The people paid one dollar did not have a reason, and the ones who came out saying the task had genuinely been interesting were the badly paid ones. Insufficient external justification, so the belief moved to cover the gap. The paper is "Cognitive consequences of forced compliance", Journal of Abnormal and Social Psychology, 1959.
Nir Eyal brings the concept into Hooked at the investment stage, where it does a specific job: explaining why a user who has already put time or money into a product becomes the product's advocate rather than its critic. That is the reading this vault took the note from, and it is a fair use of the theory, though the book supplies illustrations rather than experiments.
One methodological problem should be stated in the same breath as the theory, because it is real and it is frequently omitted. Dissonance research runs on three main paradigms: induced compliance, effort justification, and free choice. In 2010 M. Keith Chen and Jane Risen showed in the Journal of Personality and Social Psychology that the free-choice paradigm is structurally broken. Because participants sort themselves into conditions by choosing, the standard analysis will produce the signature result, the chosen item rated up and the rejected item rated down, even when nobody's attitude has changed at all. They proved conditions under which the paradigm measures spreading independently of any attitude change.
That is not the end of the story, and this page will not stop where the debunking is most satisfying. In 2021 Enisman, Shpitzer and Kleiman published a meta-analysis in the same journal covering 43 studies using artifact-free versions of the free-choice design, with 2,191 participants. The effect survived, at d = 0.40, with no evidence of publication bias accounting for it. So the criticism was correct about the method and wrong about the conclusion. Choice does appear to change preferences, and a great deal of the older literature cannot be used to show it.