One-Trial Learning: Predictions about Negative Transfer

Abstract

One-trial learning, though experienced by many has received minimal attention in motor learning research. As previous study in our laboratory provided some support, the present study was designed to follow-up by examining whether there is no expected negative transfer or response competition. 24 subjects were randomly assigned to a Response-Replacement or Stimulus-Replacement group in an ABAB learning study. Prior to testing subjects learned a list of eight S-R pairs to a criterion of two consecutive error-free trials. The stimuli were eight 3-letter nonsense syllables while responses were specific hand signs taken from the Gallaudet Survival Guide to Signing. Once the list had been learned, subjects transferred to a condition in which either three responses or three stimuli were replaced. Subjects were to learn this new list to a criterion of one error-free trial. Once they learned, subjects performed three additional test trials on the new list, without intervening acquisition trials, to assess negative transfer from the old list to the new list. The response-replaced group required significantly more acquisition trials than did the stimulus-replaced group. No significant difference in errors postcriterion between the groups was found. The findings were consistent with Guthrie's (1952) expectations, in that once the new response had been learned, there was no retroactive interference from the initial list.