Different stages of empirical research

Eventually, the job market stress comes to an end. So I thought I could start into the blogging year with a bit of humor. During the last couple of weeks I flew out to both economics and more management-oriented departments. That’s were the inspiration for this little comic came from.

state_of_empirical_cropped

Theory Horse Races

I just came back from the 76th Annual Meeting of the Academy of Management in Anaheim, California. The conference was great, not just because of the location. Lots of participants and lots of interesting sessions. Every year they have “professional development workshops”, which are a great opportunity to learn, not only for young scholars.

I attended one of those workshops called “Writing Better Theory”. We all know that the standards for how to develop theory are quite different in management science compared to economics. Therefore I was particularly interested in this session. The panel consisted of very distinguished speakers, e.g., Lorraine Eden from Texas A&M University, who is former Editor-in-Chief of JIBS. They gave good advice on how to make your arguments more concise and develop a real contribution to the theory.

But I lost some sleep over one remark by Alvaro Cuervo-Cazurra, reviewing editor of JIBS and AIB fellow (okay, I was also severely jet lagged). He urged people not to run what he called “horse races” between conflicting theories, but rather to stick to one paradigm and develop your arguments from there. I recall his point like this. If theory 1 predicts A and theory 2 predicts B, you might be inclined to conduct an empirical test, and if the data (clearly) show B you will discard theory 1 in favor of theory 2. According to Cuervo-Cazurra this is a bad idea because of two reasons:

  1. Either you encounter reviewer one who always thought that theory 1 was wrong to begin with and who will think your paper is obvious.
  2. Or, there will be reviewer two who strongly believes in theory 1 and who will find every single flaw in your study in order to reject your paper.

Take a moment to let that sink in… This strategy might actually increase your chances to publish, especially if you’re new in the profession. But running empirical tests in order to falsify theories and to discriminate between competing explanations, isn’t that exactly what science is about? What other way is there to “keep your house clean” and to prevent the accumulation of more and more theories with low predictive power?

I was too shocked to raise a question anymore. But I could have simply gotten something wrong here. After all it was early in the morning. Anyways, I would love to hear your thoughts on this!