Empirical Methods in Strategy Research: Regression Analysis and the Use of Cross-Section Versus Pooled Time-Series, Cross-Section Data

Empirical research in strategic management has relied extensively on the use of cross-section data analysis. This paper undertakes a formal examination of methodological issues raised by a cross-section analysis. Specifically we address the limited ability of a cross-section analysis to account for coefficient variation across firms and over time. These pose as important limitations in the context of strategic management research. This paper discusses alternative analytical techniques that can control for many of the limitations of a cross-section analysis. Regression methods based on pooled time series, cross-section data can help overcome biases introduced because model parameters can vary across firms or over time.