Effect of self-evaluation to identify causes of wrong answers
This study was to test empirically 'the effect of self-evaluation to identify causes of wrong answers' in compliance with quasi-experimental design. The 'self-evaluation to identify caus...
Effect of self-evaluation to identify causes of wrong answers
This study was to test empirically 'the effect of self-evaluation to identify causes of wrong answers' in compliance with quasi-experimental design. The 'self-evaluation to identify causes of wrong answers' means a set of activities, which consisted of achievement test, teachers' feed-back of item responses scored wright or wrong, explanation of right answers, and then student's self-report of the causes of wrongness of the incorrectly answered items. Following research questions were raised.
First, did the 'self-evaluation to identify causes of wrong answers' affect achievement
Second, as the self-evaluation went on, was the self-evaluation ability advanced
Third, how hard and helpful was the self-evaluation to students
Research Method
An experiment with a non-equivalent group pre-post test design was done. The experiment was to let students self-evaluate the causes of wrong answers. It was repeated 6 times in a semester. Experimental group(212 person, 4 classes) and control group(106 person, 2classes) were chosen from the 3rd grade Science classes taught by a teacher in a middle school.
First cycle was considered as a pre-test, and the other cycles were considered as post-tests. For the achievement test in each cycle, the teacher made and used a paper and pencil test. And in order to gather data on difficulties, accuracies, and helpfulness of self-evaluation, a questionnaire was used.
For the first question, ANOVA of the value-added score of each cycle against the pre-test was executed. And, to test the relationship between types of cuases and achievement level, correlation coefficients were calculated between ratios of each type of causes of wrong answers and achievement scores.
For the second question, the ratios of each type of causes of wrong answers were compared for each cycle.
For the third question, distribution of frequency ratio of each type of causes of wrong answers was tested by chi-squared test.
Research Results
First, 'self-evaluation to identify causes of wrong answers' had positive effect on the students' achievements. Value-added scores of experimental group were significantly higher than those of control group. And as the higher was the students' achievement, the higher were the ratio of substantial causes of wrong answers.
Second, student's ability of self-evaluation was advanced. As the self-evaluation went on, the ratio of substantial causes was increased, while that of superficial or mechanical causes was decreased. The ratio of mechanical or calculation errors, which might due to carelessness, did not show clear trend.
Third, 'self-evaluation to identify causes of wrong answers' was seen to be difficult but helpful to the students. The majority of the students participated in the experiment wanted to continue self-evaluation further. These results were interpreted as a representation of students' affirmative attitude to 'self-evaluation to identify causes of wrong answers'. As cycle went by, difficulty of reporting the causes of wrong answers was decreased, while accuracies was increased. It meant that 'self-evaluation to identify causes of wrong answers' could be learned. Also, student evaluated positively that a noticeable effect of 'self-evaluation to identify causes of wrong answers' was to correct the misconceptions and reduce the errors.