Examining instructor and instructional effects on students' statistics attitudes
A long-standing hypothesis in statistics education has posited that instructors could have a large impact on students’ attitudes toward statistics. This hypothesis remained untested over the years. Moreover, if instructors do have a large impact on students’ statistics attitudes, then there is a need to provide explanations as to what dimensions of teaching competencies account for this impact. Drawing on a rich data set collected from 1,924 students clustered within 23 instructors across 11 post-secondary institutions in the United States, the hypothesis concerning instructor effects on students’ statistics attitudes was tested in this study. Multilevel covariate adjustment models were employed to quantify the size of instructor effects. The analysis suggested that instructors varied considerably in their abilities to improve students’ statistics attitudes. Furthermore, instructors’ differential contributions to students’ attitudes were found to be positively associated with instructional practices most proximal to tasks involving data collection and analysis in proper contexts as well as with instructors’ attitudes toward teaching statistics classes. Lastly, results showed that instructors who improved students’ statistics attitudes were also effective at improving their expected course grades, a measure that strongly predicts student ratings of teaching. The teaching effectiveness measures explored in this study may be used to orient instructors on the development of new pedagogic skills centered on students’ statistics attitudes. Altogether, these findings necessitate the need of future studies to identify and validate additional instructional dimensions that hold promise for improving students’ statistics attitudes and also herald an exciting opportunity to expose students to a full range of instructional skills in the modern statistics classrooms.