Document Type


Publication Date





In signal detection theory, an optimal observer exploits all available information to achieve the desired goal of a particular decision strategy (Green & Swets, 1966). Detectionexperiments often provide the observer with complete knowledge of results (CKR) in order to ensure best possible performance for the task. If optimal behavior is indeed dependent upon CKR, then a degradation of that information should also reduce the likelihood of achieving optimal response bias. A single-interval auditory detection experiment was conducted to measure changes in response bias in the presence of incomplete knowledge of results(IKR) (i.e. feedback for some combination of true/false detections and true/false rejections)(Davis, 2015). The results were compared with the theoretical “optimal” bias level for the task. Statistical tests revealed significant differences between complete and incomplete feedback conditions. These results are consistent with the hypothesis that IKRcan significantly degrade an observer’s ability to achieve optimal response bias.