COVARIATION INFORMATION AND ABILITY ASSES SMENTS

The assumption is tested that it is possible to draw ability inference s solely on the basis of consistency information. In three studies, th e relative weight of consistency information and consensus information is studied by means of providing subjects with either consistency inf ormation alone or by adding consensus information to it (and in some c ases, distinctiveness information): in Study 1, it is assessed whether subjects think that they are being enabled to infer their own ability ; in Study 2, free responses are recorded in addition to this, and the wish for active information search is also measured; in Study 3, subj ects choose the information class they find themselves most relevant. In all three studies with the diverse measures, the relative impact of consensus information and the negligible role of consistency informat ion was consistently found.