Abstract
Social media “quizzes” that claim to diagnose mental illness are increasingly popular, but may have unintended consequences. In these quizzes, as in many real psychological scales, people report how frequently they experience specific symptoms of a disorder, yielding an aggregate score that represents their overall experience with that collection of symptoms. Evidence from psychological research suggests that the construction of symptom scales—specifically, the order in which items appear—may inadvertently bias people’s symptom reports. We hypothesized that when people see common (or uncommon) symptoms first, the initial feeling of ease (or difficulty) may lead them to form an impression that they experience these kinds of symptoms more (or less) often. Subjects completed the Dissociative Experiences Scale (Bernstein & Putnam, 1986) with items appearing from most-to-least common, the reverse, or in the standard order. Subjects who saw more (or less) common items first inflated (or deflated) their reports of subsequent symptoms relative to their standard-order counterparts, yielding relatively higher (or lower) scale scores. We replicated these findings in a second experiment, but did not find evidence that this order effect extended to subjects’ more general impressions of how often they dissociate. Our findings suggest relatively trivial changes to the construction of a symptom scale may have marked downstream effects on people’s reports of their symptoms and add to the ongoing discussion of how to conceptualize symptoms of psychopathology more generally.