Empirical research in Human Resource Management (HRM) and the related
fields of industrial and organizational psychology, and organizational behavior
has focused on such issues as recruiting, testing, selection, training, motivation,
compensation, and employee well-being. A review of the literature on these and
other topics suggests that less than optimal methods have often been used in HRM
studies. Among the methods-related problems are using (a) measures or manipulations that have little or no construct validity, (b) samples of units (e.g., participants, organizations) that bear little or no correspondence to target populations,
(c) research designs that have little or no potential for supporting valid causal
inferences, (d) samples that are too small to provide for adequate statistical power,
and (e) data analytic strategies that are inappropriate for the issues addressed by a
study. As a result, our understanding of various HRM phenomena has suffered and
improved methods may serve to enhance both the science and practice of HRM
and allied disciplines.
Valid research results are vital for both science and practice in HRM and allied
fields. With respect to science, the confirmation of a theory hinges on the validity
of empirical studies that are used to support it. For example, research aimed at
testing a theory that X causes Y is of little or no value unless it is based on studies
that use randomized experimental designs. In addition, the results of valid research are essential for the development and implementation of HRM policies and
practices. For example, attempts to reduce employee turnover will not meet with
success unless an organization measures this criterion in a construct valid manner.