Underappreciated method-related choices in business research.

Abstract

This paper sounds the bell as to how social sciences are likely to be populated with many under-appreciated method problems and consequent misinterpretations of findings. Our conclusions are derived from a dataset originally collected to evaluate three major likely influences on the profitability and growth of unquoted Small and Medium-Sized Enterprises (SMEs) in Portugal: financial management and strategy (options open to the firms); resources (human capital); and market conditions. We performed a sensitivity analysis of results to methodological aspects at three major levels: (1) the mix of data cleaning procedures ministered to the raw dataset; (2) the method of analysis; and (3) the definition of variables. We conclude results are subject to a certain degree of thereof variability.