How regression, rebounds, and correlation coefficients = Active learning

Students sitting in a classroom interacting with the teacher who is at the front of the class

The question is, “Are your students engaged in learning?” In introductory statistics classes at Middle Tennessee State University, the learning environment is undergoing some major changes. Pedagogical methods are being modified to engage students by building active learning into the classroom experience. These curricular and pedagogical changes are taking place in the ongoing redesign of introductory statistics as part of the university-wide initiative to promote student success.

For this course, ten activities have been developed or modified from existing resources. These activities guide students toward an understanding of the more difficult statistical topics. Instructors have been given “teacher moves” that help them facilitate student-to-student conversations about statistics. The overall goal for these activities is to lead students to create their own understanding.

We save time in the classroom, which allows us to successfully implement the active learning format, by complementing the in-class activities with online, out-of-class, and pre- and/or post-class assignments. Using videos with embedded quizzes, lesson components such as basic terminology, introduction to statistical symbols, and calculator skills are provided using a partially flipped-class format.

As an example, the regression lesson module has a pre-class activity that consists of several short videos that introduce the topics of correlation, regression, and the interpretation of the various numbers that make up a regression model. How to enter data, make a graph, and get a regression equation on the TI 83/84 calculators is also covered. Each of the short videos ends with a few simple questions for the students to answer, then the next video begins with the answers to these questions. Students are graded on whether they have done the assignment.

The in-class activity for the regression module is based on the “Regression on the Rebound” lesson developed by Buskirk and Young. Students bounce balls and record the drop height and rebound height for each bounce. They then use the skills that were presented in the pre-class activity to create a model that predicts the rebound height for a drop height they haven’t used. Class discussion centers around the interpretation of the correlation coefficient, the coefficient of determination, the slope, and the intercept. Teachers have to spend much less time on the calculation aspect of the problem than they would have without the pre-class activity. For other topics, the in-class activity is meant to be an introduction, so the out-of-class activity comes afterward and focuses on reinforcing the topics or filling in details that were not included.

Since the implementation of this teaching method in almost all introductory statistics classes in our departments, instructors not only act as conveyors of statistical concepts and procedures, but also as facilitators in students’ discussions and statistical modeling to guide students toward a conceptual understanding of statistics. Faculty have made changes to their teaching practices which have helped to foster student engagement. For example, students work in teams on activities in which they collect data individually, then combine the data with their team, and then contribute their team’s data to the class data set on the board. Other times the students might use the think-pair-share approach to help them synthesize the material for themselves first before sharing their thoughts with a partner and then the class.

A simple technique for increasing student attention is to use random approaches to call on students after asking a question in class. This increases student alertness, especially in larger classes. Teacher talk moves, like providing appropriate wait time, restating student responses, asking students to restate another student’s response, and prompting students for further elaboration, can also help advance student thinking and understanding of statistics. As we have made a change in our teaching practices, we have observed that both the materials and the pedagogy work together in a manner such that student engagement has increased, and so have final exam scores.

 

About the Authors
Scott McDaniel, Nancy McCormick, Lisa Green, and Ginger Rowell

Scott McDaniel, Nancy McCormick, Lisa Green, and Ginger Rowell

Dr. Lisa Green has been at Middle Tennessee State University (MTSU) since 2001. She has been published in journals such as Technology Innovations in Statistics Education and the Journal of Online Learning and Teaching. She participated in the development of the Resources component of the Consortium of the Advancement of Statistics Education (CAUSE)’s digital library, and also the development of a Probability Course Community digital library. She is the coordinator of the Biostatistics concentration for the Professional Science master’s degree. She recently received an award for Outstanding Achievement in Instructional Technology and the inaugural College of Basic and Applied Sciences Teaching Award.

Dr. Nancy McCormick has been at MTSU since 1989. She has been published in several journals including The Journal of Effective Teaching, Journal of Studies in Education, and Current Issues in Education. She has participated in the early development of online mathematics courses at MTSU and has received an MTSU Foundation Award for Outstanding Achievement in Instructional Technology and the TBR Distance Education Innovations Award. Her research interests include statistics education, as well as strategies for improving student learning outcomes in general education mathematics. She currently serves on the General Education Committee and Faculty Work Group at MTSU, who are in the process of mapping Common Core Standards to the TBR General Education Mathematics learning outcomes.

Dr. Scott McDaniel has been at MTSU since 1995. He has been published in several journals such as Technology Innovations in Statistics Education, Journal of Developmental Education, and the Journal of Online Learning and Teaching. He is currently serving as the learning evaluation consultant on two Department of Justice Grants and principal investigator of the NSF Course Curriculum Laboratory Improvement Education Materials Development grant. His research interests include the flipped classroom and the use of technology in statistics and algebra classes.

Dr. Ginger Holmes Rowell has been at MTSU since 2000. She is nationally known in the statistics education community for her work related to effectively integrating technology to improve learning. She worked with MTSU colleagues (McDaniel and Green) and MTSU students to develop the Resources component of the Consortium of the Advancement of Statistics Education (CAUSE)’s digital library. Dr. Rowell (along with Green, McCormick, McDaniel, and Strayer) are the co-PIs for the National Science Foundation Transforming Undergraduate STEM Education project “Modules for Teaching Statistics using Pedagogies with Active Learning” (MTStatPAL) which develops and tests materials for students and teachers to incorporate student engagement in the introductory statistics course.