Metacognition of agency across the lifespanIntroductionExperiment 1MethodParticipantsApparatusProcedureDesignProgramResultsPerformanceJudgments of performanceRelation of judgments of performance to performance: metacognitive accuracyJudgments of agencyDiscussionExperiment 2: older adultsMethodParticipantsProcedureResultsPerformanceJudgments of performanceRelation of judgments of performance to performanceJudgments of agencyAdditional analysesDiscussionExperiment 3: childrenMethodResultsPerformanceJudgments of performanceRelation of judgments of performance to performanceJudgments of agencyAdditional analysesDiscussionGeneral discussionAcknowledgementsReferencesMetacognition of agency across the lifespanJanet Metcalfea,*, Teal S. Eicha,b, Alan D. CastelbaDepartment of Psychology, 401B Schermerhorn Hall, Columbia University, New York, NY 10027, United StatesbDepartment of Psychology, 1285 Franz Hall, Box 951563, University of California, Los Angeles, CA 90095-1563, United Statesarticle infoArticle history:Received 8 November 2009Revised 13 March 2010Accepted 15 May 2010Keywords:AgencyMetacognitionOlder adultsChildrenabstractMetacognitions of agency were investigated using a computer task in which X’s and O’sstreamed from the top of a computer screen, and the participants moved the mouse toget the cursor to touch the X’s and avoid the O’s. After each 15 s trial, participants madejudgments of agency and judgments of performance. Objective control was either undis-torted, or distorted by (1) Turbulence (i.e., random noise), (2) a Lag between the mouseand cursor movements (of 250 or 500 ms), or (3) ‘Magic,’ (i.e., an increased radius aroundthe X’s for which credit was given). In Experiment 1, college students’ judgments of agencyshowed that they were sensitive to all three manipulations. They also indicated that theyfelt more in control in the Lag conditions, where there was a rule on which they couldpotentially capitalize, than in the matched Turbulence conditions. In Experiment 2, olderadults were also sensitive to all three manipulations, but less so than the college students.They were not sensitive to the difference between the Lag and Turbulence manipulations.Finally, in Experiment 3, 8–10 year-old children were sensitive to their loss of controlequally in the Lag and Turbulence conditions. However, when performance was artificiallyimproved, in the Magic condition, children took full credit and showed no evidence thatthey realized that the results were due to an external variable. Together, these findingssuggest that people’s metacognition of agency changes in important ways across thelifespan.Ó 2010 Published by Elsevier B.V.1. IntroductionUntil recently, the idea that people might not know thatthey were the agents behind their own actions was almostinconceivable. The ‘‘I” who was doing the thinking, inDescartes’ (1637/1969) meditations, became his incontro-vertible basis of all other knowledge. His metacognitionabout his own agency was the one and only thing Des-cartes could not doubt or deny. His thoughts could bewrong; his perceptions distorted; his knowledge inaccu-rate. But Descartes was unable to conceive of the possibil-ity that it was not he who was doing the thinking anddoubting. As Jeannerod and Pacherie (2004) put it: ‘‘Anumber of contemporary thinkers acknowledge that whenI judge ‘‘I think X”, I may be mistaken about X and thus thatthe mind is not wholly transparent of itself. But they main-tain, with Descartes, that when I judge: ‘‘I think X”, I cannotbe mistaken about who the subject of the thought is. ‘‘ (p.114). So, too, by this view, when actions are taken, weknow, unmistakably and in a uniquely privileged way, thatwe are doing them ourselves.This brand of privileged access, which Ryle (l949) calledthe ‘official doctrine’, has special status in the law. Eyewit-nesses accounts—the report from a witness that they sawthe perpetrator—hold enormous weight, both in court, injuror’s decisions, and even in the face of conflictingevidence (see, Fox & Walters, l986). However, even eyewit-ness reports pale by comparison to an individual’s confes-sion. There is simply no more incriminating thing that aperson can do than assert that they did it. Their attributionof self-agency with respect to the act of the crime is para-mount. And, although everyone acknowledges that confes-sions might be coerced and hence not be valid (c.f., Kassin0010-0277/$ - see front matter Ó 2010 Published by Elsevier B.V.doi:10.1016/j.cognition.2010.05.009* Corresponding author. Tel.: +1 212 854 7971; fax: +1 212 865 0166.E-mail address: [email protected] (J. Metcalfe).Cognition 116 (2010) 267–282Contents lists available at ScienceDirectCognitionjournal homepage: www.elsevier.com/locate/COGNIT& Sukel, l997), if the confession is made freely and there isevery reason to suppose the confessor believes it, then wethe jury take this to be the most sure evidence that existsthat the person indeed committed the crime.And yet, there are cases of uncoerced false self-confes-sion, and that people may believe they did something theydid not do (Kassin & Gudjonsson, 2004). Indeed, in labora-tory situations, Kassin and Kiechel (1996) and Redlick andGoodman (2003) report that adults will confess falsely tohaving hit a forbidden computer key that crashed a com-puter, under circumstances that are far from what wouldbe considered legally coercive. Furthermore, Redlick andGoodman (2003) found that younger children, 10–12 yearolds, were more likely to confess than were older childrenand college students. And, when Candel, Merckelbach,Loyen, and Reyskens (2005) said to 8–10 year-old children,‘‘You hit the SHIFT-key, didn’t you?” 36% of the children—none of who had actually hit the key—said yes. There are anumber of possible explanations. Perhaps the participants,and especially the younger children, were just being com-pliant or suggestible. But it is also possible that peoplehave considerable uncertainty about their own actions.Perhaps the younger children were genuinely less able todiscern their own agency than were the adults, and theyreally did not know that they had not done it. Childrenare also impaired, when compared to young adults, in theirmemory for source, as, indeed, are older adults (Schacter,Kagan, & Leichtman, 1995; Spencer & Raz, l995). These dif-ferences in source memory might be ascribable to a mem-ory differences. But
or
We will never post anything without your permission.
Don't have an account? Sign up