Real versus sample-based differences in comparative family business research

This article analyzes the impact of not controlling for “demographic sample” differences on research results in the area of comparative family/nonfamily business research. Using different statistical methods with and without control for “demographic sample” differences, the results show that controlling for these firm demographics in a bivariate as well as a multivariate framework is very important to discover “real” differences between family and nonfamily firms. We found “real” differences for export, budgeting, variable reward systems, profitability and gender, educational degree, and tenure of the CEO. Strategy, networking, long-term planning and control systems, perceived environmental uncertainty, growth, and management training, classified by prior empirical research as different between family and nonfamily firms, do not differ.status: publishe