A scale to measure attitude toward one's mathematics teacher

Among factors which influence the teaching-learning environment in a mathematics class is the mathematics teacher. Although teacher behaviors in the classroom are difficult to catalog and measure, the influence that a mathematics teacher exerts upon the learning environment as a result of these behaviors is reflected in students' attitudes toward that teacher. The purpose of this study was to develop a reliable scale to measure attitude toward one's mathematics teacher. The resulting Attitude Toward Mathematics Teacher (ATMT) Scale was designed specifically for use in college undergraduate mathematics classes in which 15 to 40 students are enrolled and a traditional lecture format is employed.

Student comments collected over a period of eight years at two universities were analyzed to provide 90 potential items which were administered in tryout to 150 students in six classes. On the basis of item content and statistics, five experienced teachers selected 30 items for inclusion on the Likert-type ATMT Scale. The final form of the ATMT Scale was administered to 894 students in 40 classes at the University of Tennessee, Knoxville. Students in this sample were enrolled in general mathematics (N = 355), calculus for engineers (N = 373), and mathematics for elementary education majors (N = 166). Scores on the Math Attitude Scale (N = 93), Mathematics Anxiety Rating Scale (N = 1l3), Enjoyment and Value of Mathematics Scales (N = 45), and ACT achievement test in mathematics (N = l42) and ATMT retest scores (N = l4l) were obtained from separate subsamples of the total sample.

iv

For the group of 30 items on the ATMT Scale, item-total score correlations ranged from .84 to .44; 15 of these correlations were at or above .67, 22 of them were at or above .60, and 28 of them were above .52. Item means for the high and low 25% of ATMT scores were obtained and differences between these means ranged from 2.7l to .99 out of a possible difference of 4.Oo. Eighteen of these item discrimination differences were at or above l.50, and 27 of them were above 1.25.

Test-retest reliability for the ATMT Scale was computed to be .93 (N = l4l) after an interval of one week during the last two weeks of a ten-week quarter. Coefficient alpha was found to be .92 (N = 894). Distributions of ATMT scores were given for the total sample and for students in each of the three types of mathematics classes. . Distributions of ATMT scores were given for each of 40 classes, and the distribution of teachers' average ATMT scores (N = 40) was also given. Correlations obtained between ATMr scores and scores on other scales related to mathematics indicated that attitude toward mathematics teacher is distinct from the factors measured by the other scales.

On the basis of results obtained in this study, it appears that college students do respond in good faith to the ATMT Scale and that there are significant differences in teachers' average ATMT ratings. Several suggestions were made for research based upon use of the ATMr Scale.