Information Systems Education Journal

Volume 20

V20 N5 Pages 42-53

Dec 2022

An Approach for Ushering Logistic Regression Early in Introductory Analytics Courses

Niki Kunene
Eastern Connecticut State University
Willimantic, CT USA

Katarzyna Toskin
Southern Connecticut State University
New Haven, CT USA

Abstract: Logistic regression (LR) is a foundational supervised machine learning algorithm and yet, unlike linear regression, appears rarely taught early on where analogy and proximity to linear regression would be an advantage. A random sample of 50 syllabi from undergraduate business statistics courses shows only two percent of the courses included LR. Conceivable reasons for this dearth of LR content is likely related to topic complexity, time constraints, and varying degrees of tool ease of use and support. We propose that these constraints can be countered by: [1] introducing logistic regression early, [2] informed tool selection prioritizing ease of use with comprehensive output, and [3] using/developing innovative, accessible, and easy to understand concept learning aids. This approach would leverage the proximity to linear regression and probability readily embed distributed practice for student understanding of a foundational technique.

Download this article: ISEDJ - V20 N5 Page 42.pdf

Recommended Citation: Kunene, N., Toskin, K., (2022). An Approach for Ushering Logistic Regression Early in Introductory Analytics Courses. Information Systems Education Journal20(5) pp 42-53. ISSN : ISSN: 1545-679X. A preliminary version appears in The Proceedings of EDSIGCON 2021