Abstract
1- Introduction
2- Method
3- Findings
4- Recommendations and conclusion
References
Abstract
This study explores the selection, use, and reporting of control variables in studies published in the leading international business (IB) research journals. We review a sample of 246 empirical studies published in the top five IB journals over the period 2012–2015 with particular emphasis on selection, use, and reporting of controls. Approximately 83% of studies included only half of what we consider Minimum Standard of Practice with regards to controls, whereas only 38% of the studies met the 75% threshold. We provide recommendations on how to effectively identify, use and report controls in IB studies.
Introduction
Control variables (CVs) constitute a central element of the research design of any empirical study. Confounding variables are likely to covary with the hypothesized focal independent variables thus limiting both the elucidation of causal inference as well as the explanatory power of the model (Stone-Romero, 2009; Pehazur & Schmelkin, 1991). Therefore, researchers must seek to rule out threats to valid inferences in order to determine to what extent the focal independent variables behave as hypothesized. This is typically done by including (controlling for) extraneous variables that are deemed theoretically (or empirically) important but are not focal variables of the study (Kish, 1959). The literature sometimes refers to such variables as covariates, confounding variables, nuisance variables, control variables or simply controls (Atinc, Simmering, & Kroll, 2012; Breaugh, 2008). Researchers need to account for these variables either through experimental design (before the data gathering) or through statistical analysis (after the data gathering process). In this way the researchers are said to account for their effects to avoid a false positive (Type I) error (i.e. falsely concluding that the dependent variables are in a causal relationship with the independent variable). Inadequate attention to controls is a major threat to the validity of inferences made about cause and effect (internal validity). One way of controlling by inclusion is to use a matched-group design where particular entities (e.g., state-owned and privately owned firms) that vary in terms of independent and dependent variables are matched on specific criteria (Estrin, Meyer, Nielsen, & Nielsen, 2016). An alternative way of controlling is exclusion by holding particular variables constant, such as limiting a study to emerging market firms only (Buckley, Elia, & Kafouros, 2014). Yet the most common way to control for extraneous influences is via statistical controls. Statistical controls aim at identifying potential sources of influence during study design and including CVs representing these sources of influence during data collection. During data analysis, researchers then control for these extraneous effects by mathematically partialling out variance associated with CVs in calculating relationships between other variables, thereby reducing the risk of Type II errors (Carlson & Wu, 2012; Spector, Zapf, Chen, & Frese, 2000). In this study we focus on IB research that includes statistical controls as non-hypothesized variables in regression type studies.