Why developing good psychological theories is so hard: Relative lack of robust phenomena that impose constraints on possible theories, problems of validity of psychological constructs, & obstacles to discovering causal relationships

The Theory Crisis in Psychology: How to Move Forward. Markus I. Eronen, Laura F. Bringmann. Perspectives on Psychological Science, January 29, 2021. https://doi.org/10.1177/1745691620970586

Abstract: Meehl argued in 1978 that theories in psychology come and go, with little cumulative progress. We believe that this assessment still holds, as also evidenced by increasingly common claims that psychology is facing a “theory crisis” and that psychologists should invest more in theory building. In this article, we argue that the root cause of the theory crisis is that developing good psychological theories is extremely difficult and that understanding the reasons why it is so difficult is crucial for moving forward in the theory crisis. We discuss three key reasons based on philosophy of science for why developing good psychological theories is so hard: the relative lack of robust phenomena that impose constraints on possible theories, problems of validity of psychological constructs, and obstacles to discovering causal relationships between psychological variables. We conclude with recommendations on how to move past the theory crisis.

Keywords: theory, phenomena, robustness, validity, causation

---

I could add, tentatively, this one: lack of geniuses like Newton or Einstein.

Discussion

In this article, we have discussed three fundamental difficulties in developing good psychological theories: the lack of (sufficient) robust phenomena, the lack of validity and epistemic iteration for psychological constructs, and the problem of establishing psychological causes. These issues should be addressed and discussed to make progress in resolving the theory crisis. We now outline several recommendations for psychological research on the basis of these issues.

First, our discussion supports the recent calls for more “phenomena detection” or “phenomenon-driven research” in psychology (Borsboom et al., 2021; De Houwer, 2011; Haig, 2013; see also Trafimow & Earp, 2016). By discovering new phenomena and gathering more robust evidence for those already discovered, the space of possible theories will be constrained.

Another important reason to support phenomenon-driven research is that phenomena can also be extremely important for science and society as such (Eronen, 2020): Consider, for example, the broad range of cognitive biases that psychologists have discovered, such as confirmation bias, most of which are very robust phenomena (Gilovich et al., 2002). Various theories have been proposed to explain these phenomena, such as the attribute-substitution theory, according to which people substitute difficult computations with simple heuristics, or the more general dual-system theory (Kahneman & Frederick, 2002). However, these theories are far more controversial than the phenomena themselves. Moreover, knowing that these phenomena exist is extremely important for science and society, even if we do not know the theory or mechanism behind them. The same holds for a broad range of other robust phenomena discovered in psychology, for example, the phenomenon that people tend to prefer familiar stimuli to unfamiliar ones (i.e., the mere-exposure effect; Bornstein, 1989). Simply knowing that these phenomena exist and describing them is useful, even in the absence of an accepted theory that would explain them.

In addition to being discovered and being described, phenomena can also be further analyzed by looking for shared abstract structures in different phenomena (Hughes et al., 2016). For example, at an abstract level, phenomena as different as constantly checking your phone and rewarding the good behavior of children with candy can both be seen as instances of (positive) reinforcement (Hughes et al., 2016). For all of these reasons, phenomena detection should be seen as an important goal in itself and as a central part of psychological research (see also Fiedler, 2017; Haig, 2013; Rozin, 2001).

However, we by no means intend to suggest that theorizing in psychology is hopeless or a waste of resources or that we should return to a kind of behaviorism in which theories about mental processes are rejected as unscientific. The issues we have raised should not be seen as insurmountable obstacles but rather as challenges that need to be met before good psychological theories can be developed in a given domain.

This brings us to our next point: It is doubtful whether making psychological theories more mathematical or formal, which is a common theme in the recent literature (e.g., Borsboom et al., 2021; Muthukrishna & Henrich, 2019; Oberauer & Lewandowsky, 2019; van Rooij & Baggio, 2020), will lead to significant advances in psychology as a science.⁵ None of the problems we have discussed is solved by formalizing psychological theories: There will still be no large body of robust phenomena to constrain the theories, the constructs used do not become more valid, and a formal treatment alone does not solve the problem of causality and fat-handed interventions. Moreover, many successful and extremely important theories in the life sciences are not formalized or mathematical theories (e.g., the fermentation theory or the theory of synaptic transmission; Bechtel & Richardson, 1993; Machamer, Darden & Craver, 2000). As pointed out by Rozin (2001; see also Morey et al., 2018), using complex statistical and computational models does not make psychology more scientific and can be even counterproductive if the conceptual and empirical basis (e.g., robust phenomena) is not yet solid.

Finally, it is hard to overemphasize the importance of having clearly and transparently defined concepts as the basis for theories. Note that this is not the same as formalization of theories: Concepts can be well defined in qualitatively formulated theories as well (e.g., Darwin’s theory of evolution), and formal theories can have poorly defined concepts as their elements (e.g., models in memetics that have a clear mathematical structure but for which the central concept “meme” is not well defined; Kronfeldner, 2011). Conceptual clarification and construct validation should be seen as an important and valuable parts of research, and validation should be taken to be an iterative and ongoing process instead of just a hurdle that needs be crossed. In our view, strengthening the conceptual basis of psychological theories is at least as important as improving statistical techniques and practices in psychological research.

In the long run, this will also help with the problem of causal inference, as having clearly defined and clearly measurable constructs makes it easier to perform targeted interventions and to track their effects. With sufficiently well-defined constructs and valid measurements, it may also be possible to eventually infer causal relationships from purely observational data (for more, see, e.g., Eronen, 2020; Rohrer, 2018). Another possible reaction to the problem of finding psychological causes is to develop noncausal theories, for example, in the form of abstract functional principles extracted from phenomena (De Houwer, 2011; Hughes et al., 2016), although whether noncausal theories can be truly explanatory is a matter of ongoing debate (see, e.g., Reutlinger & Saatsi, 2018).

Fortunately, there are ongoing research programs in psychology that exemplify the good practices we have describe above. For example, after the recent disappointments in ego-depletion research, there are now increasing efforts to better define the key constructs, such as self-control and related concepts, and to validate different ways of measuring them (Friese et al., 2019; Inzlicht & Friese, 2019; Lurquin & Miyake, 2017). A broader example is the functional-cognitive paradigm (De Houwer, 2011; Hughes et al., 2016) that aims at first establishing environment-behavior relations (robust phenomena) and then formulating explanations for them in terms of clearly defined mental constructs that act as mediators. Finally, as a more concrete example, Robinaugh et al. (2020) propose a theory for panic disorder that is tailored to this specific disorder and thereby constrained by phenomena (there is robust evidence for many central phenomena related to panic attacks), and the authors also explicitly focus on defining the key concepts.

To conclude, we believe that the most fundamental factor underlying the theory crisis is that the subject matter itself, psychology, makes it very hard to develop good theories (Meehl, 1978). Drawing on contemporary philosophy of science, we have discussed three central challenges to developing psychological theories: There are often not enough robust phenomena to constrain theories, not enough attention is paid to defining and validating constructs, and establishing psychological causes is very hard. We hope that this article brings more attention to these crucial issues and thereby helps to provide more solid building blocks for the theoretical foundations of psychology.