Mental retardation and inspection time: a two-stage model for sensory registration and central processing

Mildly mentally retarded and nonretarded young adults were compared in simple visual discrimination tasks from which can be derived an estimate of inspection time (lambda), assumed to reflect the time taken by some basic component in the decision process. In the first experiment retarded and nonretarded subjects' performance showed improvement in a condition employing a backward mask to limit the duration of sensory information available to subsequent central processes, but substantial group differences in lambda remain unchanged. When target stimuli were presented for short durations, but without interruption by masking, retarded subjects performed less effectively than did nonretarded subjects and, despite extensive practice, did not achieve nonretarded subjects' levels. These results suggested a two-stage serial model of discrimination that was tested in the second experiment. Concurrent manipulation of two factors confirmed that (a) lambda involves initial sensory registration followed by central processing, (b) for both groups these stages are independent, and (c) poorer performance among retarded subjects is the consequence of deficiencies at both stages.