The construction and comparison of scales for predicting academic success in grade six

The purpose of this study was two fold: (1) to construct scales for predicting academic success in Grade 6 of an "intermediate" school from school marks in Grade 5 and from scores made on The Science Research Associates Achievement Test, and (2) to compare these scales. The sample group consisted of 185 students who completed Grade 5 in one of three elementary schools, and who also completed the first semester of Grade 6 in 1966 in the same intermediate school in the area, and for whom Science Research Associates Achievement Test scores were available. The predictive variables consisted of grades made in basic subject matter areas in Grade 5 and selected scores on The Science Research Associates Achievement Test, a total of twelve factors, or variables. "Academic Success" was defined as attaining a grade-point average of C+, or above, in basic subjects in the first semester of Grade 6. Quantitative tabulations were made of the measures of twelve variables indicating the per cent success for each interval of each variable. The "Score Sheet" provided the means for computing the "Total Predictive Scores" for the 185 individuals of the sample group. These 185 "Total Predictive Scores" were then tabulated into a "Predictive Scale." For the purposes of comparison, three sub-scales were constructed from among the twelve predictive variables. Thus, three separate "Score Sheets" and "Predictive Scales" were produced, one based on "Verbal", and one on "Quantitative" factors. The third was based on using school marks in Grade 5 and the composite total score on The Science Research Associates Achievement Test. Comparisons were then made among and between the four predictive scales constructed from the same twelve variables Conclusions On the basis of the results of this study, the following conclusions seem justified: 1. The method of prediction shown here provides a simple but accurate method with no complicated statistics involved. 2. From a knowledge of fifth grade averages and scores made on the regularly administered test, The Science Research Associates Achievement Test, the counselor or others responsible for classroom placement can predict the chances for a student to make a C+ average or better in the sixth grade. 3. This type of predictive scale can be used successfully on the intermediate level of educational progress. 4. In the comparison of the predictive scales constructed, the two scales based on the twelve separate variables and the one based on seven "verbal" factors proved to be the most effective. 5. The predictive scale based on the "Quantitative" factors did not have as great a distinct "break" in the middle range -- 45 per cent to 76 per cent. 6. The predictive scale based on two variables, school marks and composite achievement scores was the least effective. 7. Predictive scales with several factors seem to be more effective.