Connecting HRM scholarship and HR analytics: The benchmark predictivity approach

The aim of this paper is to address the research-practice gap between HRM scholarship and HR analytics in organisations, catalysed by distinct preferences for explanatory and predictive paradigms in either domain. While HRM scholars mostly favour the explanatory paradigm to conduct theory-driven empirical investigation, HR analysts value the predictive paradigm to identify factors of actual predictive relevance for intervention and change. To bridge this gap, we introduce the benchmark predictivity approachto compare the predictive accuracy of models developed in HRM scholarship. Through systematic comparison of predictive accuracies of models grounded in scholarly knowledge, this approach estimates the predictivity of different factors for practical decision-making and provides a critical yet often overlooked evaluation of theoretical propositions in terms of their practical utility. This paper explains the benchmark predictivity approach and offers stepwise instructions along with a practical demonstration using a case study on voluntary employee turnover. The results indicate that some antecedents of voluntary turnover, recognized under the explanatory paradigm, did not predict actual voluntary turnover in the future, suggesting that their practical utility in mitigating voluntary turnover should be approached with caution. The proposed benchmark predictivity approach encourages a closer integration between HR scholarship and evidence-based HR practice.