4. Conclusion
According to Chia (2014), conducting relevant and rigorous European management research requires researchers “to imaginatively seek out ever-newer meanings to both exceptional and ordinary everyday experiences, familiar and unfamiliar happenings, and taken-for-granted conceptual formulations” (Chia, 2014, p. 686). PLS-SEM's exploratory characteristic perfectly fits this point of view and can therefore be considered a valuable data analytical approach for European management researchers. From a statistical explanatory modeling point of view, hypothesis testing is a critical element in developing relevant and rigorous theory (Shmueli & Koppius, 2011). In a PLS-SEM context, hypothesis testing relies on bootstrapping. As evidenced by the overview of PLS-SEM applications in European management research (see also Table 1), the accompanying bootstrap procedures are often suboptimal. In response to this observation, this paper provides a detailed overview of how to construct bias-correct percentile bootstrap confidence intervals and to demonstrate how these bootstrap confidence intervals can be used to test hypotheses related to frequently encountered research situations in management research. Although not directly related to a particular substantive European management research domain, this paper contributes to the European management research domain in the following ways. First, a key recommendation of our work is that in order to make better statistical inferences, which is one cornerstone of (European) management research (Cashen & Geiger, 2004; Thietart, 2001), constructing bias-corrected percentile bootstrap confidence intervals based on a large number of bootstrap samples (i.e., at least 10,000) offers a powerful approach to test a large variety of hypotheses. In terms of practical implementation, Table 3 presents a set of guidelines and minimal requirements for PLS-SEM bootstrap procedures that are relevant to researchers, editors, and reviewers alike.
