AN EXPLORATORY STUDY OF THE RELATIONSHIP BETWEEN MATHEMATICS ANXIETY AND THE PROCESSES USED BY DEVELOPMENTAL COMMUNITY COLLEGE FRESHMEN TO SOLVE VERBAL MATHEMATICS PROBLEMS

Statement of the Problem. The primary objective of this research was to investigate the relationship between the level of mathematics anxiety and the problem solving processes and strategies used by developmental community college students to solve verbal mathematics problems.;Procedure. The study involved a sample which was selected from students in the lowest level mathematics course at an urban community college. The instruments used were the Mathematics Anxiety Rating Scale (MARS), the Sixteen Personality Factor Questionnaire (16PF), the Quick Test (QT), the Days Problem Set, and a demographic questionnaire. The 36 students in the sample were chosen from those who scored in the upper and lower quartiles of a distribution (n = 394) of MARS scores and were classified as either high or low mathematics anxious. The sample was stratified according to mathematics anxiety level (high, low), race/ethnicity (Hispanic, white, black), and sex (male, female).;Each student was administered the QT, 16PF, the Days Problem Set, and the demographic questionnaire individually in an interview setting. The think aloud technique was used for the Days Problem Set and it was audio-taped. The protocols were coded according to a slight modification of the Days coding scheme.;Six hypotheses were tested for each case individually by comparing the appropriate scores from the Days Problem Set against two different sets of norms: the median scores for the sample in this study and the mean scores for the sample in the study by Days. Each of these comparisons was done three times--over the simple structure problems, over the complex structure problems, and over the entire set of problems. The data from an individual case were said to have supported an hypothesis if the appropriate score of the student was above (or below) the norm score as predicted by the hypothesis. The G Index of Agreement was used to determine if the support for each hypothesis was statistically significant over all the cases.;Results. The findings indicated at the .05 level of significance that higher levels of mathematics anxiety were associated with smaller quantities and varieties of problem solving processes used by the students, with higher numbers of structural errors, and with lower product scores. The hypotheses, which predicted that higher mathematics anxiety level was associated with higher uses of systematic trial-and-error strategy and with higher time scores, were not supported. However, had these two hypotheses posited that higher mathematics anxiety level was associated with lower use of systematic trial-and-error strategy and with lower time scores, then the G Index of Agreement would have indicated statistically significant support. Mathematics anxiety was unrelated to quantity of executive errors.;Conclusions. Since the high mathematics anxiety group used significantly fewer quantities and varieties of problem solving heuristics, made more structural errors, and had lower product scores, one could assume that high mathematics anxiety is related to poor mathematical problem solving performance. There was some evidence to indicate that the high mathematics anxious are less likely to use systematic trial-and-error and less likely to spend time working to a solution. Computational ability was unrelated to mathematics anxiety level.;The fact that the trends in this study were more evident among the males than the females, might imply that mathematics anxiety, although it may be more prevalent among women, may be more debilitating among men or that mathematical problem solving failure may be more anxiety producing among men. Race and ethnicity were unrelated to support of the hypotheses.