Parsimony, Elegance, and Conceptual Coherence in Sport Psychology
In 2008 I was asked to give the Coleman Griffith Lecture at the annual meeting of AASP in St. Louis, Missouri, which had the title, Whatever Happened to Parsimony, Elegance, and Conceptual Coherence in Sport Psychology? I decided to argue that forsaking some fundamental attributes of science was not in the best interests of scholars and practitioners in sport psychology. I believe that there is value in maintaining the attributes of parsimony, elegance, and conceptual coherence. Let me explain.
One of Einstein’s famous quotes is that in science we should make everything as simple as possible, but not too simple! Parsimony in science simply means that we keep constructs as simple as possible. In my Coleman Griffith talk, I argued that there is a trend in sport psychology to make constructs more complex than they need to be. Some of this results from the natural desire to make constructs more applicable to the sport context, such as the efforts of Harwood and colleagues, but scholars often do not realize that by making things more “relevant” to sport they also make the constructs more complex and cause a subsequent loss of parsimony. McFee, a philosopher of science, is harshly critical of scholars in sport psychology because of their penchant for explaining phenomena with “boxes and arrows” and making things more complex than they should be (McFee, 2005). He argues that he is against the “boxes” mentality prevalent in sport psychology, where what is hoped for is a graphic model of boxes connected by arrows, as though sport psychologists knew what the boxes and arrows meant. In my address I gave several examples of complex multilevel theories (e.g., Vallerand, 2001; the hierarchical model of motivation derived from self-determination theory) of this trend toward complexity, in addition to the work of Elliot and colleagues (e.g., 1999) and Harwood and colleagues (e.g., 2008), which I criticized earlier. Is this trend toward complexity one that we should follow?
In her profile of Nobel Prize winners, Zuckerman (1977) gave several attributes of the typical prize winner, but one common attribute is particularly noteworthy: They see simplicity where other people see complexity. As an example, when Crick and Watson (1953) in their quest to discover the structure of the DNA molecule published their model of the double helix, Maurice Wilkins (a fellow scientist who was a rival in the quest) was surprised to see how simple the model was and is quoted to have said, “How simple, how elegant” (Watson, 1996). The quest for expanded frameworks might be valuable because we may be able to provide a better description of the complexity of motivation processes, but a cost is often present, and part of that cost is a loss of parsimony! It is well for sport scientists to remember the famous saying of William of Occam (1285–1347): “Entities are not to be multiplied beyond necessity.” Known as Occam’s Razor, it is a call for parsimony, which is sometimes ignored.
There are many examples of scientists in psychology “seeing simplicity” when others had noticed only complexity. These examples spur research into the phenomenon and allow us to understand it better. The example of the social facilitation paradigm is a case in point. Historically, scientists had noted that the presence of others sometimes facilitated and sometimes inhibited the performance of individuals on a variety of tasks (e.g., Triplett, 1897). But in his review of the findings, Zajonc (1965) saw simplicity when he realized that learning tasks were inhibited by the presence of others, whereas performance tasks or simple tasks were facilitated by the presence of others. Further, he argued, the presence of others creates arousal and arousal facilitates the dominant response. When learning, the dominant response was incorrect; when performing, the dominant response was correct. This simple insight suddenly made sense of the previous 70 years of conflicting research. This ability to see simplicity sometimes gives a conceptual insight into psychological (and other disciplinary) mechanisms.
Again, Crick and Watson (1953) give a classic example of elegance, which is the attribute of being simple but profound in its implications. In their article in Nature, they concluded by stating that it had not escaped their notice that the specific pairing of the double helix they had postulated suggested a possible copying mechanism for the genetic material! As we now know, this was the most profound scientific finding of the 20th century. The double helix solution was elegant in that it gave clear theoretical and practical meaning and changed our understanding of biological systems. Indeed, it was the birth of modern biology.
We can find many examples in psychology, too. One elegant theory is the social facilitation paradigm (Zajonc, 1965), which I discussed earlier. Zajonc demonstrated his elegant conceptual explanation with an elegant experiment. He had two groups of participants learn a complex task over many trials. One group did it in the presence of others, and the control group learned the task alone. Zajonc demonstrated that performing in the presence of others inhibited learning but facilitated later performance when compared to the control group: A simple but elegant and profound experiment demonstrated his theory. This experiment, and the later research that it stimulated (e.g., Martens, 1969), gave us insight into the reasons arousal and evaluative anxiety inhibited learning on motor tasks.
Another example of a parsimonious and elegant construct is self-efficacy (e.g., Bandura, 1977b). Self-efficacy cognitions represent a person’s convictions or beliefs that he or she can successfully execute a course of action to produce a certain outcome. It is likened to a situation-specific self-confidence. But in this context, it is a simple but elegant construct that has stimulated a great deal of research into the determinants of motivation in exercise and physical activity (see Gilson & Feltz, this volume). And, of course, as I have been trying to expound, achievement goal theory in its original form is a parsimonious and elegant theory, too.
Parsimony and elegance are valued attributes for a theory, but conceptual coherence is an essential attribute! As we all know, theory gives meaning to data. Having an empirical paradigm that is weak on coherent constructs is not a desirable paradigm. These empirical paradigms with weak constructs, or constructs that lack conceptual coherence, are what produce the random noise in the literature against which the true signal of advancement is difficult to discern (Roberts, 1989). I must confess that in my experience with doctoral students, I find that they generally become sophisticated in research methods and statistical analyses but often struggle with understanding the psychological constructs and mechanisms about how things work psychologically. My favorite question to students who are preparing a research agenda is to ask them how they know that they are asking the right questions, or even important ones? I am reminded of the adage, If it is not worth doing, it is not worth doing well! The most important task that we can do as mentors of doctoral students is to make certain they understand that theory gives meaning to data and that statistical and research sophistication are secondary and merely tools to demonstrate the meaningfulness of theory. But they must do more than use “just any theory.” Is the theory coherent and meaningful to the question being asked? All of us have to make our own decisions about what constructs give meaning to the research that we are undertaking. My argument here is that the conceptual base should be coherent!
As an example, at the 2009 ISSP Congress in Marrakesh, Ed Deci was asked to react to some research papers, using self-determination theory as the conceptual base. When one researcher used self-determination theory to overlay his own theory to add impetus to his normal theoretical paradigm, Deci commented that it did not make much sense to do that because the theories were quite different in their conceptual base. His point was that it was better to build constructs within the conceptual paradigm used to capture the dynamics of the question being asked rather than bring in another theory and overlay the constructs of that theory over the original constructs. In other words, to give meaning to data, the theory must be conceptually coherent!
Achievement goal theory is a social-cognitive theory that was developed from the learned helplessness (mastery versus helplessness) research of Dweck (e.g., 1975), the cooperation–competition research of Ames (e.g., 1984a), the motivation research of Maehr (e.g., 1983) and the work of Nicholls (e.g., 1978) on the various conceptions of ability that children held. From this research, Nicholls saw simplicity where others had seen complexity and presented the parsimonious constructs that we all know today. Nicholls’ unique contribution was to recognize that two understandings of ability existed and that previous theories had assumed one or the other. For example, self-efficacy (see Gilson & Feltz, this volume) assumes a task-involving conception of ability, whereas attribution theory (e.g., Weiner, 1972) assumes an ego-involving conception of ability. Nicholls, Maehr, Ames, and Dweck developed a theory (with their own interpretations) that recognized that both conceptions of ability need to be inherent in the theory. In addition, they recognized that individual difference variables (goal orientations) and contextual dynamics (motivational climate) were part of the same conceptual paradigm. That is the elegance of achievement goal theory. By recognizing that we each use task- or ego-involving conceptions of ability depending on the context, we have a conceptually coherent theory that encompasses both versions of the definition of success and embraces both individual difference variables (goal orientations) and the effect of the situation (motivational climate) on achievement striving.