Signal Detection Analysis of Auditory Sequence Discrimination by Children

Abstract

A three-choice auditory oddity-task and two three-choice auditory matching-to-sample tasks in which the sample was presented first or last were presented to 96 normal Ss ranging in age from 7 to 25 yr. and to 36 children with reading problems aged 8 to 16 yr. Signal detection analysis was used to assess differences among positions of the odd or matching sound with respect to proportion correct (hit rate), decision strategy (false alarm rate), and underlying discriminability (d'). On the matching-to-sample tasks there was a very consistent bias at all ages against responding to the sound most remote from the sample, even though the remote sounds were not less discriminable. On the oddity-task a bias against choosing the first sound and toward choosing the third sound occurred for younger children. Positional differences for children with reading problems were almost identical with those of normal Ss.