The Effects of Maze Training Upon Transference to Standard Cloze Performance

Abstract

The purpose of this study was to ascertain if there would be significant positive transfer of training for students when completing standard cloze exercises after receiving instruction in multiple-choice cloze (maze) exercises. An experimental, matched control group, pretest-posttest design was used for the study. The sample consisted of 44 fifth grade students who were equated in terms of IQ scores. The Degrees of Reading Power Exam was administered in the fall of 1978 as a pretest to all students. The students were randomly assigned Form A or Form B just prior to its administration. The experimental group was administered Form A of the Standard Cloze test which was designed by the experimenter, in January 1979. The treatment group received 15 minutes of instruction three times a week for a four month period. During this time the experimental group worked on activities which focused on the use of contextual clues. All of the exercises presented were constructed in the maze format. The treatment lessons for this program were based upon suggested teaching strategies developed by the State Education Department, Division of Educational testing, at Albany. Upon completion of the four month treatment period, students were posttested. All 44 students were administered Form B of the Standard Cloze test and the alternate form of the Degrees of Reading Power Exam from the one completed in the fall. A series of t tests was used to analyze the data at the .05 level of significance. The results indicated that the treatment group did significantly better than the control group on all posttesting tasks, and that there was significant positive transfer of training for students when completing standard cloze exercises after instruction in multiple-choice cloze (maze) exercises. Recommendations for classroom use of the cloze procedure as an instructional tool as well as recommendations for future research were given.