Three measures of accuracy-confidence correspondence (A-CC) for eyewitness verbal testimony (i.e., within-subjects A-CC, between-subjects A-CC measured across witnesses exposed to the same witnessing conditions, and between-conditions A-CC (a term derived by the author that refers to between-subjects A-CC measured across witnesses exposed to different witnessing conditions)) were identified from the literature and examined in the light of Deffenbacher's (1980) optimatity hypothesis, Leippe's (1980) theoretical framework for how optimality affects A-CC, Wells and Murray's (1984) statistical explanation of this relationship, and other research from a broad range of related areas. Three forms of the optimality hypothesis, based on each of the three measures of A-CC, were developed. The forensic relevance of each form of the optimality hypothesis was discussed.
It was proposed that confidence is affected by trace effects (which influence the association between subjective feelings of trace strength and memory accuracy), and by inferential effects (which involve witnesses' assessments of factors that they believe ought to have influenced their levels of accuracy). Further, it was proposed that accuracy is influenced by trace effects (which affect a witness'potential to have accurate recollections), and context effects (which influence the realization of a witness' potential to have accurate recollections). A case is made in the thesis for the separation of situational factors into: (a) the physical conditions of witnessing (which, it is argued, primarily exert trace effects on both accuracy and confidence); and (b) social influences (which, it is argued, primarily exert either context effects on accuracy or inferential effects on confidence).
The empirical work included five major experiments in which subjects were exposed, in a laboratory setting, to slides of a staged motor-vehicle accident. Written responses to questions, and ratings of confidence in each response were elicited in each study.
Independent measures in the program of work included: manipulations of inspection time and retention interval (aspects of the physical conditions of witnessing); manipulation of "information" regarding the possible consequences of subjects' testimony, and "information" regarding the optimality of witnessing conditions (social influence variables); and seven individual difference variables, these being imagery ability, extraversion, neuroticism, attention, arousal, sex and age. ....................................