Time Dependence in the Cox Proportional Hazard Model as a Theory Development Opportunity: A Step-by-Step Guide

Abstract

The Cox proportional hazard model has often been used for survival analysis in organizational research. The Cox model needs to satisfy one critical assumption—time independence—that the effects of independent variables are constant over survival time (also known as the proportional hazard assumption). However, organizational research often encounters time dependence in the Cox model. Organizational studies have traditionally seemed to view time dependence as an empirical nuisance, but we highlight that it is also a theory-development opportunity. Indeed, from our review of AMJ and SMJ papers published in a recent 10-year period, we found that researchers rarely considered time dependence as a theory-development opportunity, and worse, many of them did not test for (or report tests for) time dependence. The purpose of our study is to change this pattern. To this end, we provide a step-by-step guide to facilitate testing for time dependence and using time dependence as a theory development opportunity. We also demonstrate our step-by-step guide with an empirical example.