This paper is from Session 1E: Assessment: its lessons and effects
which comes under Topic 1: Statistics education: Looking back, looking forward
Paper 1E1 (Friday 13th, 14:00-15:30)
Improving student learning and instructional effectiveness through the innovative use of automated analysis of formative assessments
Presenter
- Alex Lyford (Middlebury College, United States)
Co-author
- Jennifer Kaplan (University of Georgia, United States)
Abstract
Twenty years ago, Gal and Garfield (1997) argued that complex curricular goals associated with statistics learning could not be addressed adequately using solely multiple choice or short answer questions. In addition, the GAISE College Reports of 2005 and 2016 suggest that instructors use formative assessments, such as constructed-response questions, to improve student learning. In this paper, we demonstrate an innovative, forward-looking method for meeting the historical goals of improving student learning and instructional effectiveness through the use of algorithmic, automated categorization of student constructed responses to formative assessment items. We then demonstrate how instructors can use reports generated from their students’ responses by the automated algorithms in an online portal. This portal allows instructors to integrate formative assessments into instruction quickly, especially in large-lecture classes.