| Abstract: | Our earlier article in Personnel Psychology demonstrated how general-izability theory could be used to obtain improved reliability estimates in the human resource (HR) and firm performance literature and that correcting for unreliability using these estimates had important implications for the magnitude of the HR and firm performance relationship. In their comment, Huselid and Becker both raise criticisms specific to our study and broad issues for the field to consider. In our present article, we argue, using empirical evidence whenever possible, that the issues and criticisms raised by Huselid and Becker do not change our original conclusions. We also provide new evidence on how the reliability of HR-related measures may differ at different levels of analysis. Finally, we build on Huselid and Becker's helpful discussion of broad research design and strategy issues in the HR and firm performance literature in an effort to help researchers make better informed choices regarding their own research designs and strategies in the area. |
