Why replication studies are essential: learning from failure and success

Van Witteloostuijn’s (2016) commentary “What happened to Popperian Falsification?” is an excellent summary of the many problems that plague research in the (Social) Sciences in general and (International) Business & Management in particular. As van Witteloostuijn (2016:pp] admits his “[...] diagnosis is anything but new – quite the contrary”, nor is it applicable only to the Social Sciences. When preparing this note, I was reminded of Cargo Cult Science, a 1974 Caltech commencement address by Physicist Richard Feynman (Feynman, 1974), which – more than four decades ago – makes many of the same points, including the pervasive problem of a lack of replication studies, which will be the topic of this short rejoinder.
Conducting replication studies is more difficult in International Business (IB) than it is in many other disciplines. For instance in Psychology – a discipline that favours experimental research – one might be able to replicate a particular study within weeks or, in some cases, even days. However, in IB data collection is typically very time-consuming and fraught with many problems not encountered in purely domestic research (for a summary see Harzing, Reiche & Pudelko, 2013). Moreover, most journals in our field only publish articles with novel research findings and a strong theoretical contribution, and are thus not open to replication studies. To date, most studies in IB are therefore unique and are never replicated. This is regrettable, because even though difficult, replication is even more essential in IB than it is in domestic studies, because differences in cultural and institutional environments might limit generalization from studies conducted in a single home or host country.
Somehow though, pleas for replication studies – however well articulated and however often repeated – seem to be falling on deaf ears. Academics are only human, and many humans learn best from personal stories and examples, especially if they evoke vivid emotions or associations. Hence, in this note, instead of providing yet another essayistic plea for replication, I will attempt to argue “by example”. I present two short case studies from my own research: one in which the lack of replication resulted in the creation of myths, and another in which judicious replication strengthened arguments for a new – less biased – measure of research performance. Finally, I will provide a recommendation on how to move forward that can be implemented immediately without the need for a complete overhaul of our current system of research dissemination.

This is an author accepted manuscript version of an article made available in this repository in accordance with the publisher's self-archiving policy. The final version is published by Emerald in Cross Cultural & Strategic Management and is available at: http://dx.doi.org/10.1108/CCSM-07-2016-0133