Using a sample of 394 student subjects and six teachers, the author sought to prove the hypothesis that highly independent students would achieve a higher degree of pupil growth in mathematics mastery than would highly conforming students when both groups were under the teaching influence of high flexibility teachers. The California Psychological Inventory (CPI) by Harrison G. Gough was the test instrument selected to test student independence (Achievement via Independence Scale -- Ai), and teacher flexibility (Flexibility Scale -- Fx). Pupil growth in mathematics was determined by a pretest- posttest application of the Quantitative Thinking section of the Iowa Test of Educational Development (ITED). Following a 2 x 3 factorial design which provided for two levels of student independence, three levels of flexibility, and many measures of each condition, the author applied an analysis of covariance statistical procedure to the general statistics output, which came from an IBM 360 Computer using a U6604 and U6601 Computer Program respectively in The University of New Mexico Computer Center. An “F” test of significance at the five percent level of confidence was applied to each compared group of independent vs. conforming pupils while under the influence respectively of three selected levels of teacher flexibility. Test results, while reflecting considerable pretest posttest raw score gains in mathematics mastery, failed to reach statistical significance at the five percent levels of reliability and the hypothesis of superior performance by independent students while being instructed by highly flexible teachers had to be rejected as measured by the instruments used in this research. Considerable evidence of scale overlapping was found in the personality inventory instrument which tended to lend support to the objections of several personality test critics. Overlapping of the independence and conformance scales made it difficult to clearly identify students as either more independent or more conforming. Failure of results to achieve significance was attributed in part to the limiting factors of small available (six junior high school mathematics teachers), a rather stratified and selective community (a scientifically oriented environment with above average emphasis on education), and the previously mentioned evidence of scale overlap between the Ai and Ac scales or the personality inventory instruemnt. It was concluded that more sensitive personality measuring instruments were needed for future research in attempting to relate these pupil and teacher personality traits as predictor variables of academic performance.
Stanley James Hipwood (Mon,) studied this question.