A Primer for Social Worker Researchers on How to Conduct a Multinomial Logistic Regression

ABSTRACT The purpose of this article is to explain how to conduct a multinomial logistic regression (MLR) to increase its usage among social work researchers. A challenge for social work researchers carrying out practice-based research is to incorporate statistical analyses that are comprehensible to clinicians. Another challenge is that categorical variables, as opposed to continuous, commonly occur in clinical settings, yet their usage is seldom taught in social work education. This article will discuss MLR, a categorical data analysis used when there are three or more unordered categories in the outcome variable. Multinomial logistic regression determines differences across client groups, and can be useful in assessment, case planning, and examination of outcomes.