Abstract
This study investigated children’s self-report accuracy as a function of task type and also verified generalization of do-say correspondence across tasks. Six children between 6 and 11 years of age participated in the study. “Doing” consisted of reading words, playing a computer game, solving a math problem, and labeling music-related stimuli. “Saying” consisted of reporting on the accuracy of performance following the automated computer feedback. Baseline assessed correspondence for the different tasks. Correspondence training was conducted for the task in which levels of accuracy were the lowest. Generalized do-say correspondence was then assessed in untrained tasks. For four children, correspondence was lowest for the academic tasks. Four of six children exhibited generalized correspondence after the first training, and the remaining two children did so following a second training with a different task. Distinct tasks seemed to control different patterns of self-report accuracy. Results on generalization indicated do-say correspondence as a generalized operant behavior.