Individual Diagnosis in Written Composition

Abstract

A mechanical test-scoring device was evaluated in three ways: (1) experimentally to determine its contribution to classroom learning; (2) by questionnaire; and (3) by assessment of mechanical reliability. Televised learning materials were followed by three randomly assigned follow-up modes: (1) quizzes graded by machine, immediately returned, and discussed; (2) quizzes that were discussed, but with papers returned the next day; and (3) discussion only. Fourth and tenth grade classes were used. The value of such a study for the evaluation of devices for educational use was demonstrated. Machine malfunctions occurred in this target environment which had not appeared in extensive prior testing. The device groups showed no superiority over quiz only groups, possibly due to the reliability problem. Attitudes were favorable toward the machine and toward immediate return of the test results.