Diemo Urbig

Nonlinear and interaction effects have become standard in management research and are increasingly common in economics. However, although researchers routinely include linear control variables to address potential confounding, linear controls alone are often insufficient for obtaining unbiased estimates of nonlinear and interactive effects. This presentation introduces the concept of nonlinear and interactive control variables (NICs) and explains why they are critical for the valid estimation of nonlinear and interaction effects. It shows how omitted nonlinear or interactive confounding effects can distort substantive conclusions and why, when suitable observed variables are available, these effects should be addressed through corresponding nonlinear and interactive controls. Based on an in-press article in the Journal of Management, the presentation illustrates the relevance of NICs, introduces a systematic approach for identifying critical NICs, and demonstrates how their inclusion can strengthen causal inference, improve robustness, and support more rigorous theory testing. It also introduces a new related online tool designed to help researchers apply this approach in their own empirical work and to support methods instructors in teaching it.