Choice consists of the mental process of thinking involved with the process of judging the merits of multiple options and selecting one of them for action. Some simple examples include deciding whether to get up in the morning or go back to sleep, or selecting a given route for a journey. More complex examples (often decisions that affect what a person thinks or their core beliefs) include choosing a lifestyle, religious affiliation, or political position.

Most people regard having choices as a good thing, though a severely limited or artificially restricted choice can lead to discomfort with choosing and possibly, an unsatisfactory outcome. In contrast, unlimited choice may lead to confusion, regret of the alternatives not taken, and indifference in an unstructured existence; and the illusion that choosing an object or a course leads necessarily to control of that object or course can cause psychological problems.

Choice and evaluability in economics

When choosing between options one must make judgments about the quality of each option's attributes. For example, if one is choosing between candidates for a job, the quality of relevant attributes such as previous work experience, college or high school GPA, and letters of recommendation will be judged for each option and the decision will likely be based on these attribute judgments. However, each attribute has a different level of evaluability, that is, the extent to which one can use information from that attribute to make a judgment.

An example of a highly evaluable attribute is SAT score. Everyone knows that an SAT score below 800 is very bad while an SAT score above 1500 is exceptional. Because the distribution of scores on this attribute is relatively well known it is a highly evaluable attribute. Compare SAT score to a poorly evaluable attribute, such as number of hours spent doing homework. Most employers would not know what 10,000 hours spent doing homework means because they have no idea of the distribution of scores of potential workers in the population on this attribute.

As a result, evaluability can cause preference reversals between joint and separate evaluations. For example, Hsee, George Loewenstein, Blount & Bazerman (1999) looked at how people choose between options when they are directly compared because they are presented at the same time or when they cannot be compared because one is only given a single option. The canonical example is a hiring decision made about two candidates being hired for a programming job. Subjects in an experiment were asked to give a starting salary to two candidates, Candidate J and Candidate S. However, some viewed both candidates at the same time (joint evaluation), where as others only viewed one candidate (separate evaluation). Candidate J had experience of 70 KY programs, and a GPA of 2.5, whereas Candidate S had experience of 10 KY programs and a GPA of 3.9. The results showed that in joint evaluation both candidates received roughly the same starting salary from subjects, who apparently thought a low GPA but high experience was approximately equal to a high GPA but low experience. However, in the separate evaluation, subjects paid Candidate S, the one with the high GPA, substantially more money. The explanation for this is that KY programs is an attribute that is difficult to evaluate and thus people cannot base their judgment on this attribute in separate evaluation.