Developmental Psychology1981, Vol. 17, No. 4,481-489Copyright 1981 by the American Psychological Association, Inc.0012-1649/81/1704-048 U00.75Illusory Contingency in Children at the State FairJohn R. WeiszUniversity of North Carolina at Chapel HillAccurate judgments about personal control depend in part on accurate judgmentsabout the contingency of outcomes because noncontingent outcomes are inher-ently uncontrollable. Yet children often fail to recognize noncontingency whenthey see it. In a developmental study of such failures, children's contingencyjudgments were assessed following their participation in chance activities at astate fair. Younger children (aged 6-10 years) regarded the outcomes of theseactivities as controllable—that is, most of them saw the outcomes of their ownperformance as caused by skill-related factors and regarded such factors (age,intelligence, effort, and practice) as significantly influencing the outcomes ofother children's performance. Older children (aged 11-14 years), by contrast,generally identified the outcomes of their own performance as caused by luck,and they minimized the role of skill-related factors in the performance outcomesof others. Yet even the older children regarded such factors as somewhat relevantto outcomes they predicted for others. This was true even of children at formaloperational age levels and of children who explicitly identified outcomes as causedby "luck." Moreover, children at both age levels showed evidence of self-servingbias: Those who had won prizes they wanted saw outcomes as strongly affectedby effort, but those who had failed to win did not. The findings support Piaget'sviews on the pervasiveness of perceived contingency in young children. Consistentwith adult literature, however, the findings suggest that neither the attainmentof formal operations nor the recognition that luck causes outcomes ensures ac-curate judgments about control.In theory and research on the psychologyof control, perceived contingency plays acentral role (see, e.g., Abramson, Seligman,& Teasdale, 1978; Seligman, 1975; Weisz,1979). For an individual to exercise controlover an outcome, two conditions must exist.First, the outcome must be contingent onvariations in the behavior of persons similarto that individual; a noncontingent outcomeis inherently uncontrollable. Second, the in-dividual must have sufficient competence tocapitalize on the contingency that exists—that is, he or she must be able to producethe behavior on which the desired outcomeThis study was supported by grants from the NationalInstitute of Mental Health and from the ResearchCouncil of the University of North Carolina.Special thanks go to Dale Webb, Mary Gratch, NanHarmon, Bobby Atwell, Mary Gwyn, and Lora Marvinfor assistance in data gathering and analysis. I am alsograteful to George Holden, John Schopler, and KeithYeates for their helpful comments on an earlier draft.Requests for reprints should be sent to John R. Weisz,Department of Psychology, Davie Hall 013-A, Univer-sity of North Carolina, Chapel Hill, North Carolina27514.is contingent. (For further details, see Weisz& Stipek, in press.) Mistaken perceptionsof contingency can lead to futile attemptsto influence noncontingent outcomes (seeLanger, 1975), inappropriate losses in self-esteem following uncontrollable failure (cf.Abramson et al., 1978), and even the blam-ing of innocent victims for adverse noncon-tingent outcomes (see Lerner, 1977).Literature from the Piagetian traditionsuggests that erroneous perceptions of con-tingency are to some extent a function ofdevelopmental level. In Piaget and Inhelder's(1975) developmental analysis of chanceconcepts, for example, they argue that chil-dren understand the noncontingency of ran-dom events only as they develop a grasp ofreversible operations. The outcome of athrow of the dice, for example, is not logi-cally reversible. Its fortuitous nature canonly be understood when it is contrasted withoutcomes that can be reversed by invertinga causal sequence. That sort of contrast,however, only begins to make sense duringthe concrete operations period (elementary481482JOHN R. WEISZschool age levels) and is not fully understooduntil formal operations (adolescence).Early in development, according to thePiagetian perspective, children reason intu-itively; they often infer contingency on thebasis of mere contiguity of events (for data,see Siegler & Liebert, 1974). Lacking log-ical operations, the children have no inter-nalized standard representing truly contin-gent relations against which to test newexperience. One result of this is that someevents that occur by chance, both in gamesand in nature, are perceived as contingenton potentially identifiable causes. Piaget andInhelder (1975) saw evidence of such a per-ception in many of the "why" questions withwhich young children so often frustrate theirparents. Questions such as "Why isn't therea spring in our garden?" or "Why are youso tall and yet have small ears?" are said toreveal an assumption that identifiable con-tingencies exist for what are actually chanceevents in nature. Piaget and Inhelder did notgive examples drawn from games of chance,but they would presumably interpret youngchildren's questions about why dice fall intoa certain pattern or how a younger child candraw winning cards more often than an olderchild can (both observed in my own re-search) as reflecting similar contingency as-sumptions.Other examples of erroneous perceivedcontingency are given in Piaget's (1930)studies of causal reasoning per se. In oneexample, a child maintains that "the moongets bigger because we are growing bigger"(p. 304). Unlike adults, who might say thisanalogically, the young child, according toPiaget, "means that we actually make themoon grow bigger" (p. 304). Piaget does notgive an exhaustive explanation of such er-rors. They seem likely, however, to be stim-ulated partly by a limited understanding ofthe natural world and partly by difficultiesin comprehending the relation between causeand effect (for related evidence in theachievement domain, see Nicholls, 1978).Whatever their causes, the kinds of illusorycontingency described in the causality lit-erature (e.g., Piaget, 1930, 1976), like thosedescribed in the literature on chance (Piaget& Inhelder, 1975), are depicted as decliningduring development and as largely disap-pearing with the appearance of formal op-erational thought.A third body of Piagetian evidence maybe relevant to