Why Baseline Modelling is Better than Null-Hypothesis Testing: Examples from Research About International Management, Developing Countries, and Emerging Markets

Abstract

Purpose – This chapter reports on a rapidly growing trend in data analysis – analytic comparisons between baseline models and explanatory models. Baseline models estimate values for the dependent variable in the absence of hypothesized causal effects. Thus, the baseline models discussed in this chapter differ from the baseline models commonly used in sequential regression analyses.

Baseline modeling entails iteration: (1) Researchers develop baseline models to capture key patterns in the empirical data that are independent of the hypothesized effects. (2) They compare these patterns with the patterns implied by their explanatory models. (3) They use the derived insights to improve their explanatory models. (4) They iterate by comparing their improved explanatory models with modified baseline models.

Methodology/approach – The chapter draws on methodological literature in economics, applied psychology, and the philosophy of science to point out fundamental features of baseline modeling. Examples come from research in international business and management, emerging market economies, and developing countries.

Findings – Baseline modeling offers substantial advantages for theory development. Although analytic comparisons with baseline models originated in some research fields as early as the 1960s, they have not been widely discussed or applied in international management.

Practical implications – Baseline modeling takes a more inductive and iterative approach to modeling and theory development.

Originality/value of paper – Because baseline modeling holds substantial potential, international-management scholars should explore its opportunities for advancing scientific progress.

Keywords: baseline model, model comparison, theory development, hypothesis testing