Title: Developing Students' Thought Processes for Choosing Appropriate Statistical Methods
Discipline(s) or Field(s): Research Methods, Statistics
Authors: Elizabeth Knowles and James Murray, Department of Economics, University of Wisconsin - La Crosse
Submission Date: August 23, 2012
Abstract: Introductory statistics classes typically emphasize computation and implementation procedures for a number of statistical tests. While it is essential to build these skills before achieving higher-order critical thinking skills, students often struggle in subsequent research methods courses when expected to select appropriate statistical tests to answer research questions. This requires an understanding of how statistical methods are related to one another; and to achieve this, students must develop a more advanced organization of knowledge. We designed a lesson to help students build a knowledge organization to achieve this outcome, and observed students to better understand their thought processes. We share our thought process map for selecting a statistical test, report on the impact it had for our students, and offer suggestions for improving the lesson. In addition, we describe the thought processes students used, both before and after being exposed to the thought process map, and identify sources of confusion revealed through the lesson study process. These include: when to apply an independent-samples test versus a paired-samples test, how the identification of scale of measurement led students to choose the wrong statistical method, the difficulty students had recognizing or defining what the variables in a problem were, and the lack of understanding of the difference between statistical language and colloquial language.