| This study examined misconceptions of probability among students (n = 66) at a graduate school of education, each of whom intends to teach mathematics at the elementary, secondary or college levels. Most participants had teaching experience. The purpose of the study was to evaluate teacher understanding of certain pivotal probability topics, given that teacher knowledge is a key factor in educational quality.; Participants filled out a questionnaire concerning their educational background, their views of the importance of the study of probability and their understanding of the concept of probability (including the effects of their beliefs on their view of probability). All reported some level of formal instruction in probability. Participants also completed a supervised nineteen question multiple-choice test of probabilistic concepts and misconceptions and provided explanations of their answers.; Overall, the average rate of correct responses on the test problems was 56%. Average success rates on the individual test items ranged from 98% to 3%. All participants demonstrated a basic understanding of the concept of probability and could carry out simple probability calculations. Participants of all levels showed evidence of the equiprobability bias (miscounting of outcomes in a question concerning two dice), exhibited ignorance of the effect of sample size and were seldom successful on counterintuitive conditional probability problems.; Probability preparation, especially at the college level, was found to be strongly related to performance on the probability test. Participants who intended to teach at the secondary and college levels generally performed better than those who intended to teach at the elementary level, which may be related to the level and timing of their probability training. The greatest differences between the levels seemed to result from differences in comprehension of compound probability and in knowledge of basic combinatorics.; Gender differences were observed. The overall correct response rate of males (64%) was significantly greater than that of females (52%). Males and females also tended to answer differently, based on the type of question; many of these differences were statistically significant. |
