Perceived contingency of skill and chance events: A developmental analysis.

Abstract

Outcomes in skill-controlled activities are contingent on people's attributes and behavior. Outcomes of chance activities are inherently noncontingent. In this study, we explored age differences in recognition of this distinction between skill and chance. Kindergarteners, fourth graders, eighth graders, and college students took part in one game of chance and one game of skill. After each game, subjects predicted winnings of other players who differed in certain attributes (e.g., intelligence) and behavior (e.g., effort) that would influence only skill outcomes. Two developmental trends emerged: (a) On both chance and skill tasks, older subjects expected variations in attributes and behavior to have less impact on task outcomes than did younger subjects; and (b) older subjects were more adept at making predictions that reflected contingency of skill and noncontingency of chance. Kindergarteners showed no ability to make skillchance distinction. Fourth graders were aware of distinction at a gross qualitative level, but their predictions showed that they were unaware of some of most important logical implications of that distinction. Eighth graders and college students were aware of skill-chance distinction and most of its logical implications; yet their predictions revealed a lingering belief that chance outcomes could be influenced slightly by variations in people's attributes and behavior. This last finding is interpreted in light of Rosch's (Rosch & Mervis, 1975) model of category formation. To make accurate judgments about what we can and cannot control, we must accurately judge contingency of outcomes. If an outcome (e.g., rain, a roll of dice) is not contingent on people's behavior, then we will be unable to control it. (For more details on determinants of control, see Weisz & Stipek, 1982.) There is mounting evidence that such noncontingency is very difficult for young children to detect. Piaget (1930) reported that children younger than 6 or 7 years show primitive psychological causality, that is, the belief that any desire whatsoever can influence objects, belief in obedience of external things (p. 303). The young child, as described by Piaget, has illusion that diverse physical events in