Although as scientists, we tend to eschew truisms, few would challenge the claim that a scientific field can be no better than the tools it uses to derive knowledge. After all, just as a builder cannot rise a 50-story skyscraper without a suitably tall crane, and a French chef cannot bake a flaky croissant without a properly heated oven, so too is the organizational scientist constrained by limitations of his or her tools. Acknowledging this, in our work we strive to use methods and measures that satisfy prevailing standards of reliability and validity. The underlying logic is that these desiderada enhance our confidence in the conclusions our data allow us to draw about human behaviour in organizations.
I am not concerned about this state of affairs because I believe that in our individual studies we currently do a reasonable job of using measures and methods that are reliable and valid. Instead, I am concerned about the degree to which we embrace another truism-namely that because any one research method is limited in what it does, we must rely on multiple methods to build our science (Webb, Campbell, Schwartz, & Sechrest, 1999). Extending my metaphor, a pneumatic hammer cannot be expected to do the job of a crane and the successful builder needs both these tools, and others, to get the building off the ground. Likewise, the chef cannot create by using only an oven, a mixer or any single tool, for that matter. Here too, a variety of implements are required to create the desired final product.
As obvious as this may be, the underlying logic appears to have evaded many of us as we toil to create not buildings nor pastries, but a product that is arguably even more complex-knowledge about human behaviour at work. However, because most of us likely are capable of articulating the benefits of using multiple research methods, it probably is more accurate to claim that any failure to follow through is based not on a lack of awareness, but rather, on something more insidious-momentum. Indeed, I think that many of us conduct research using the tools with which we are most comfortable, knowing full well its limitations, and acknowledging them honestly in the Discussion section. Indeed, it has became formulaic to point out how common method variance may limit what we can learn from studies using same-source data collected at a single point in time, or how studies using college students may not be generalizable to the complexities of 'real workers' in 'real situations ' (Greenberg, 1987b). We see these and other 'mea culpas' in our Discussion sections all the time, but I wonder if anyone is paying attention to them and doing anything to expand the repertoire of tools upon which our field is built.
From my own informal observations, even from atop many lofty perches, I believe that too few of us are doing what it takes to bring things around in this realm. As a member of the senior editorial team of the Journal of Organizational Behavior (JOB), I welcome this opportunity to invite you, my colleagues in the field of OB, to embrace the opportunity to go beyond the comfortably familiar by conducting research that uses a diverse variety of research methods.