An investigation of approaches and strategies for resolving students' misconceptions about probability in introductory college statistics

The first purpose of this study was to investigate college professors' perspectives on students' misconceptions of probability. This was accomplished by gathering information from a survey answered by 66 college professors and analyzing interviews conducted with 11 professors who teach probability and statistics at several colleges and universities in the U.S.;The second purpose was to validate a test instrument (PRQ) designed by the investigator with the objective of identifying subjects who hold three highly prevalent misconceptions about probability: representativeness, equiprobability bias, and outcome orientation. This was accomplished by invoking expert validation and known-group validation procedures.;The third purpose was to investigate the approaches and strategies used by college professors to facilitate the resolution of students' misconceptions. To achieve this, 13 college professors teaching 23 intact sections of introductory statistics at 5 different 2 and 4-year colleges in the fall of 2004 were recruited, along with 601 of their students.;The results revealed that, in the opinion of the majority of respondents, instructors, in general, do not address misconceptions in instruction. For all that, college professors believe that misconceptions about probability interfere with students' ability to master inferential statistics, a position supported in this study by finding a significant association between the number of investigated misconceptions and students' achievement in statistics.;The validity of the PRQ was confirmed.;College professors were found to employ a variety of organizational formats and strategies for dealing with students' misconceptions. The most popular format was lecture, followed by whole class and small group discussions.;Two strategies were employed most frequently: (1) confronting students with their misconceptions by having them compare predictions based on inappropriate models with outcomes of hands-on experimentation, then helping students build an appropriate model; (2) using analogies and anchors to reconstruct misconception-laden probability knowledge.;Other popular methods included giving students written assignments designed to address the misconceptions, and having students observe and discuss patterns in computer generated simulations of random processes.;Some evidence was obtain that instructors who targeted the misconceptions directly in instruction achieved better results in facilitating the resolution of misconceptions than those who used formal instruction.