Logits and tigers and bears, oh my! A brief look at the simple math of logistic regression and how it can improve dissemination of results

Logistic regression is slowly gaining acceptance in the social sciences, and fills an important niche in the researcher’s toolkit: being able to predict important outcomes that are not continuous in nature. While OLS regression is a valuable tool, it cannot routinely be used to predict outcomes that are binary or categorical in nature. These outcomes represent important social science lines of research: retention in, or dropout from school, using illicit drugs, underage alcohol consumption, antisocial behavior, purchasing decisions, voting patterns, risky behavior, and so on. The goal of this paper is to briefly lead the reader through the surprisingly simple mathematics that underpins logistic regression: probabilities, odds, odds ratios, and logits. Anyone with spreadsheet software or a scientific calculator can follow along, and in turn, this knowledge can be used to make much more interesting, clear, and accurate presentations of results (especially to non-technical audiences). In particular, I will share an example of an interaction in logistic regression, how it was originally graphed, and how the graph was made substantially more user-friendly by converting the original metric (logits) to a more readily interpretable metric (probability) through three simple steps. Accessed 7,862 times on https://pareonline.net from June 06, 2012 to December 31, 2019. For downloads from January 1, 2020 forward, please click on the PlumX Metrics link to the right