Learning to Appreciate (Methodological) Comments From the Scientific Out-Group

How Can Scholars Use Comments Outside Their Discipline to Enhance Their Discipline-Specific Research?

I have come to believe that in general, comments from outside one’s discipline can be of high potential value, because they are less likely to be constrained by the dominant frames of reference within one’s field. They offer an outside perspective on the given research and can thus provide several opportunities for enhancing the research in ways one possibly did not think of. Comments from other disciplines are thus an important source of divergent thinking, which is in turn associated with creativity (McCrae, 1987). In other words, comments from other fields can help us to think outside of the box, whereby the box refers to our own scientific silo.

The potential use of such inputs from other fields is manifold. With regard to theory development, outlining that other fields of research are struggling with the same or similar problems can help to underscore the contribution of one’s own proposal. On the other hand, showing that other fields of research were able to address an issue on a different level of analysis can also highlight the potential contribution of one’s own research. For example, social and organizational psychologists have argued that organization science has demonstrated a positive relationship between diversity on the organizational level and organizational financial performance to underscore the importance of identifying the same association on the team level. When discussing one’s findings, elaborations on their use in other domains can help to illustrate their importance beyond one’s own field.

However, I believe that there are at least two risks associated with comments from other disciplines, which make them a two-sided sword. I will briefly elaborate on these potential issues before discussing possible ways of overcoming them.

First, comments from outside one’s discipline are likely to be perceived as comments from a member of the out-group, and such inputs are less likely to be integrated and acted upon (Gruenfeld, Mannix, Williams, & Neale, 1996). Because we tend to attribute higher status and relevance to our in-group (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987), our own discipline in this case, we are also likely to attribute lower status to members of other disciplines (the out-group). In turn, information that is shared from low-status individuals is remembered poorly and is likely to be devalued (Stewart & Stasser, 1995; Wittenbaum & Bowman, 2005). In sum, social identity theory posits that we tend to downgrade information from out-group members and are less likely to act upon it.

Second, the integration of input from outsiders is conditional on one’s ability to integrate the information into one’s own propositional networks. If there is little cognitive overlap between two individuals, which is likely if they are from different disciplines, they have difficulties with understanding each other (Scholl, 1996) because of their different “interpretive frameworks” (Polanyi, 1958, p. 74). A comment from a person from another discipline carries that person’s prior knowledge, beliefs, and context. If these differ too strongly from those of the receiver, knowledge transfer is difficult.

Thankfully, both issues can be overcome. Demoting out-group members’ input can be avoided by instilling a superordinate and/or a dual identity (González & Brown, 2006). In other words, both the sender and the receiver of a comment must be aware that they are both group researchers who want to advance our understanding of group dynamics, regardless of one’s field. Especially if sender and receiver identify themselves both as members of their discipline and as members of the interdisciplinary group research community (i.e., if a dual identity is active, González & Brown, 2006), cognitive biases can be avoided. Furthermore, studies from the field of team diversity indicate that seeing value in diversity can have a positive effect on the elaboration of information between individuals who are different from each other (e.g., Homan, van Knippenberg, van Kleef, & De Dreu, 2007; Meyer & Schermuly, 2012).

Differences in interpretative frameworks can be overcome by applying findings from cross-cultural psychology and training. To understand the input from researchers from a different domain, we need to understand that domain’s culture: their values, attitudes, and worldview (Fowler, 2006). While I am not suggesting intercultural training to take place at interdisciplinary conferences, I do believe that some of the features of interdisciplinary meetings, such as roundtables and panel discussions, help to facilitate an understanding of other disciplines’ worldviews.

In summary, the activation of a common identity, seeing value in diversity, and developing an understanding for other disciplines’ world view can help to integrate comments from other disciplines into one’s own research in ways outlined above. I believe that the first two potential remedies are already quite prominently embedded in the culture of interdisciplinary (team) research. I do however believe that we could do a better job when it comes to trying to understand the scientific worldview of other disciplines.

How Can Scholars From Different Disciplines Integrate Methodological Differences to Enhance Group Research?

If one views the context of the INGRoup conference with its focus on communication science, social and organizational psychology, and organization science, I assert that these three disciplines are characterized by different research foci, namely by predominantly (although not exclusively) addressing research questions on the micro, meso, and macro level. These foci have, if viewed as a whole, produced a rich set of methodologies for different purposes and levels of analysis. However, these different foci have also made a methodological integration across disciplines somewhat difficult. Nevertheless, we can profit from the rich set of tools available across disciplines by using them whenever our own tools are not appropriate for answering a specific research question.

I believe that this must be the fundamental basis underlying methodological differences or discussions: A method is always a mean to a certain end and should not be employed or advocated in a self-serving way. Thus, methodological criticisms across disciplines are only warranted if an employed methodology is unsuitable for answering a given research question.

That being said, we need to acknowledge that methodologies are not just value-free tools for scientific discovery. The use of certain techniques and methods is engrained in certain disciplines: One is more likely to see a multilevel model in a presentation by an organizational scientist and one is more likely to see behavior coding in a presentation by a communication scholar. Due to their different research foci, different disciplines favor different methods. As such, a certain method can represent the basic underlying assumptions (i.e., the belief structure in Polanyi’s terms) of a discipline, and therefore carry meaning and value for a scientist from that domain. A method that we are unfamiliar with in our own discipline can therefore be perceived as symbolizing the scientific out-group, and the out-group is usually devalued (see above). Accordingly, I have witnessed quite emotional criticisms of other scientists’ methods.

I do however believe that it is possible to take a learning perspective on other disciplines’ methodologies—as Keyton and Waller highlight in this forum. The different disciplines studying group processes and dynamics have developed expert methodologies mirroring the disciplines’ different foci on group dynamics. For example, communication science has produced a rich set of methodologies for observing, coding, and analyzing interaction sequences (e.g., Meyers, Seibold, & Brashers, 1991). In fact, coding schemes have become so useful that other disciplines of group research have begun to develop them (e.g., Schermuly & Scholl, 2012). Social psychology has a lot to offer when it comes to controlled experimental designs in the laboratory and their analysis with variance-analytic techniques. Organization science and organizational behavior oftentimes rely on survey data in the field and have given rise to sophisticated methodologies for analyzing them in an adequate way, such as multilevel or latent growth modeling. Approaches from all three domains can be combined successfully if the research question so requires. I experienced such an attempt as quite fruitful when my colleagues and I employed multilevel modeling on video-coded communication data that we elicited in a controlled laboratory experiment (Meyer, Shemla, & Schermuly, 2011).

Therefore, I believe that the fundamental rule for incorporating methods from other disciplines in one’s own research should be whether they are (better) suited for answering a given research question. It would thus be useful to view methodologies that are more frequently employed in other areas of research not as methods from other disciplines but simply as methods. If one bears in mind that a method is always a tool for answering a question, my outlook on the employment of other disciplines’ methodologies in one’s own research is an optimistic one.

On a closing remark, I believe that the field of group research as a whole is in need of implementing methods from fields that appear to be even further away than other traditional group research silos: There has been growing criticism about a problem-method misfit with respect to the study of group processes (Roe, Gockel, & Meyer, 2012). As Roe et al. argue, a lot of current theory in group research pertains, at least implicitly, to temporal dynamics in teams, such as durations or the sequence of events (i.e., what follows from what). However, such dynamic phenomena are oftentimes analyzed by comparing mean levels of supposedly stable features between groups, instead of investigating the shape of processes within groups. The latter will however require the employment of methods such as event history analysis (e.g., Broström, 2012) or time series analysis, both of which are usually not employed in team research. Therefore, if we are really interested in studying group processes and group dynamics, we must be open to methods from other disciplines. We should even venture out into the jungle of diverse methods from other disciplines and search for ones that are better suited for dealing with temporal dynamics than our own ones. To allude to Beck’s opening piece of this forum: We have to be more interdisciplinary when it comes to research methods—all the time.