Strategy aims at understanding the differential effects of firms’ actions on performance. However, standard regression models estimate only the average effects of these actions across firms. Our paper discusses how random coefficient models (RCMs) may generate new insights about firm heterogeneity and its effects on performance in empirical settings in strategy. RCMs allow testing for and predicting firm-specific coefficients, thereby distinguishing between effects that have a significant mean versus significant variance. RCMs may also be used to explore the sources of firm heterogeneous effects. We develop a simulation testbed using synthetic datasets to show that RCMs are more precise at allocating firm heterogeneous effects to model slopes and intercepts than standard regression models. We also discuss RCMs’ possible limitations due to sample size requirements, nonconvergence problems, and potentially restrictive assumptions. Overall, RCMs allow strategy researchers to test and build new theories at a more granular level.
Keywords: Strategy; Research; Competitive Advantage; Competitive Strategy; Performance;
Citation:
Alcácer, Juan, Wilbur Chung, Ashton Hawk, and Gonçalo Pacheco-de-Almeida. "Applying Random Coefficient Models to Strategy Research: Identifying and Exploring Firm Heterogeneous Effects." Strategy Science 3, no. 3 (September 2018): 481–553. View Details