RESPONSE TIME AS A FUNCTION OF RESPONSE MODE AND ITEM DIFFICULTY.
AMERICAN INSTITUTES FOR RESEARCH IN THE BEHAVIORAL SCIENCES PALO ALTO CALIF
Pagination or Media Count:
Response modes typically were investigated as possible influences on posttest errors. Such investigation occurred in the present study but, in addition, response-mode influences on performance during learning were also considered. One result of this latter consideration was that promising procedures were developed for predicting learning outcomes posttest errors from the way Ss respond to the learning task. These may aid the study of variables other than response modes. College-student Ss studied a psychology program individually under laboratory conditions 7 Ss were randomly assigned to a think-answers condition and 7 Ss to a write-answers condition. Each of the 14 Ss completed 670 items in the condition to which he was initially assigned. After an immediate posttest, and one lesson, response-mode conditions were reversed and remained reversed while 13 of the Ss completed the next 474 program frames. The times each question and each answer were in view were recorded separately and automatically for each S. While writing their answers Ss persisted longer on difficult, but relatively unambiguous, instructional items U 1, p 0.001. Such persistence was greater for Ss whose posttest errors were fewer than the median U 9, p 0.027. Persistence on difficult items decreased when Ss who had been writing were changed to the think-answers condition and became greater for those Ss who had been thinking answers when they began to write U 1, p 0.001. Author