On the nature and role of theory in psychological science

[EnPsychlopedia]

Hello, friends and colleagues!

This discussion is a topic about which I have thought a lot over the past several years, but a recent narrative overview by Flis (2022) published in APA Div. 1's Review of General Psychology has reignited my overall interest in the topic. As anyone who has studied psychology to at least the bachelor's level should know, psychology is different from other sciences in that it has yet to unify itself through the creation of a robust explanatory theory or set of theories. For example, at the macro- level, all of biology is unified via the theory of evolution by natural selection; all of chemistry is unified by robust atomic theory; and physics, while still seeking a grand unifying theory, is unified by three sets of theories which each work at different "levels" (Newtonian mechanics at the local level, general relativity at the more zoomed-out level, and quantum mechanics at the very zoomed-in level). Psychology, on the other hand, is left bereft of any unifying theoretical background around which to build strong inferences and against which to check any published findings for congruency with an agreed-upon base of "known" facts connected into theoretical models. Most psychologists give lip service to working from a biopsychosocial framework, which is a start, but as yet no widely-applicable theoretical models for how these three dimensions work together to manifest human behavior, cognition, and emotion have arisen. In clinical psychology, the closest we have to a clinical theory is the CBT triangle, in which emotions, behaviors, and cognitions bidirectionally interact to manifest certain conditions. How these interact has never been explained beyond basic principles of conditioning, and how a person comes to develop his/her own unique triangle is also begging to be theoretically modeled (though the stress-diathesis model is at least one attempt to come to some sort of conclusion).

This is all well-known and attested throughout the field of psychology. But what struck me about the Flis (2022) paper is its discussion of the role of theory in the psychological science reform movement which was born in the wake of the replication crisis. Essentially, he argues that there exist two camps of reformers who are at odds with one another: (1) those who insist that correcting the research methods and journalistic practices of psychology will help create a healthier, more robust literature which will then eventually organize itself into theory; and (2) those who insist that a lack of emphasis on the creation of theory in psychology will mean that efforts to correct these practices, while needed, will ultimately fail to correct the field since there will be no background theory against which to test if the reformed practices have accurately moved us closer to publishing findings which represent the truth (as opposed to findings which are "false positives").

Any way, with all that said, the topics I want to bring forth for discussion are:

1. What is the nature of theory? What constitutes the difference between theory building and simply reporting effects?
2. How explanatorily powerful does any given theory need to be for us to unify psychology? Do we need one unifying theory like biology or chemistry, or can we exist as a unified field with a larger, but contained, set of theories working in tandem together?
3. Is theory necessary for true reform? Is it enough, at this moment in time, to simply reform our research and publication practices and continue to largely focus on publication of effects, and the replications (or failed attempts at replication) of those effects? Can pre-registration of hypotheses, meta-science, and open science fix "the problem" sans any theoretical development?
4. If effects are sufficient to build a unifying knowledge base, will those effects, given enough time, organize themselves into theory, or should more emphasis be placed on theory now, despite the knowledge that such emphasis may slow down the rate at which unreplicable findings are pruned from the literature?
5. Is theory even possible in psychology, given that the topics of psychology (i.e., humans) are partially the products of culturally-distinct and ever-changing societal structures?

Perhaps this discussion won't strike anyone here as particularly important or interesting, but the part of me which pushed me to do a philosophy minor in my undergraduate years was really intrigued by Flis' outlining of these discussions, and I was just hoping to see what everyone thought.

Wishing you all happy times and good health!

Reference:

Flis, I. (2022). The function of literature in psychological science. Review of General Psychology, 26(2), 146-156. doi: 10.1177/10892680211066466

---

Members don't see this ad.

 

[EnPsychlopedia]

P.S. I am not sure this will particularly interesting to folks here, given that most of us are in/interested in the more applied areas of psychology, where effects tend to replicate well relative to other areas of the field (and given that it is debatable to what degree a unifying theory would improve clinical care and outcomes), but I thought it may be a neat change of pace for everyone!

 

[ccool]

Hi. Thanks for posing. I will say that my first thought was a memory of the time I accidentally said something (theoretical and) critical to mainstream thought on this board and the mess that was. That was the On Violence convo. Linking a decent starting point in the case you’re interested in seeing some dynamics Flis refers to play out in the real time microcosm of this internet board.

I’m slightly wary to engage bc I don’t have space to manage any subsequent weirdness, but I’m interested in engaging in this discussion with curious, open people.

My initial thoughts are related to one of my (many) critiques of the field. To me the history and present of psychology is marred by its fight for “legitimacy” as a “worthy” and valid field of scientific inquiry and practice.. particularly as compared to the practice of medicine or the scientific inquiry of fields like physics or chemistry, for example. It’s always seemed to me that the development of the field of psychology was influenced in large part by its efforts to convince others that it is real and meaningful, which I have empathy for. However, it also seems like it was trying to convince itself of this too. And in doing either/both, there are ways that it falls victim to a weird stereotype threat.

An example of this related to this discussion are the mainstream biases against qualitative research in psychology. In my experience, it is largely seen as a weak, non-rigorous approach that (psychology) people associate with so-called “soft sciences” like anthropology or sociology. It’s funny though because fields like biology, chemistry, physics and their foundational theories are all based on qualitative research. Before we can feign to act on a thing and explore things like mechanisms of change, we have to be able to say what something is. These other sciences have no compunction about doing this precise thing! And so, theories, testable products of observations and rational thinking.

But because psychology is a real science and not froufrou like sociology or whatever, we can and must go directly to a thing and act upon it. /s We can make strong hypotheses about what might happen when we act upon the thing based on the results of others who have done similarly. We can also make inferences about why we found (or not) the relationships that we did based on related things we already know. We can also empirically examine the veracity of those inferences. Doing so means we are objective and we have something immediately worthwhile and actionable to offer to inform policy and practice. Not like that qual, theory stuff. /s

And while those things are all important and have led to important knowledge and interventions for the specific thing that was studied, we are still limited because in many cases. We still are not articulating the essence of the thing, just it’s manifestations. And while I don’t think more qualitative research is the only answer to beginning to articulate and test theories, I think it would be a big help.

I think it would also help for the field as a whole to stand in its power and remember that it is valid and worthy in itself and in comparison to others’.

And until the more traditional, white, western notions/practices/areas of the field (ie the status quo) catches up with the researchers and areas related to feminism, queer, disability, black, indigenous, and other traditionally marginalized communities who are openly welcoming to qual research and exploring what is in addition to examining its implications and manifestations, I don’t see much changing in the way of production and exploration of unifying theories of psychology.

 

[Ollie123]

Interesting topic. I'm drowning in work so won't have time for a detailed response but some general thoughts.
- I'd push back on the idea that evolution is a fully unifying theory. That is perhaps true for a small subset of the work that constitutes biology and touches on evolution, but there are whole subfields of biology where evolution has little to do with the day-to-day. To say its a unifying theory for the field seems grandiose. In this regard, we could eventually have a unifying theory for all of psychology but I'd frankly be happy if we could come up with a unifying theory of a sub-sub-sub-subfield.
- I would argue psychologists as a whole suck at developing true scientific theories. We tend to either come up with grand, abstract explanations for things that aren't necessarily falsifiable in any way or we come up with something incredibly locked in to our measure of choice and call the SEM model we built that ignored a dozen or more critical variables and used a college student sample a "theory."
- We are equally bad at testing theories. Its a problem in all of science, but psychology in particular that we often seek out confirming evidence. When a study disproves our pet theory, clearly its because we did the experiment wrong. Too often, we don't even do the experiment and build elaborate causal models out of an mturk survey which is honest-to-goodness embarrassing in this day and age. Fine is a starting point, embarrassing when we leave it at that and pretend the work is done.
- Computational neuroscience/cognition is gradually starting to worm its way into the mainstream. The field absolutely has its issues, but one of the major upsides is it forces people to really articulate their theories in a specific way. I think this (among other things) will help push is in a better direction over time. I think we're at least 1-2 decades out from seeing the fruits of this though.
- 100% agree with ccool we need to embrace all methods. I've never been able to understand the disdain qual and quant researchers have for one another. I've seen this play out on study section. The quant people believe the qual research isn't rigorous and the qual people think the quant work misses the point and I think both camps are narcissistic idiots.


That said, I think all of this misses the forest for the trees RE: the replication crisis and broader issues in the field. Misaligned incentives are the problem. Neither theory nor pre-registration fixes "If this grant doesn't get funded I'm going to get laid off" or "If this paper doesn't get into a high-impact journal I won't get tenure." Everything else is barely going to move the needle.

 

[Fan_of_Meehl]

For me, one of the things that drew me to psychology as an undergrad was that it appeared to be 'applied philosophy.' I was really into philosophy (as well as the natural sciences, being initially pre-med) but, when I took intro to stats/experimental methods in undergrad along with a course in social psych (which covered all the interesting stuff done at the inception of social psychology such as the Robber's Cave experiments, the Milgram experiments, Solomon Asch's conformity stuff, etc.) the notion that here is a field that actually strives to try to ANSWER the eternal and deep philosophical questions about human nature just blew me away and really got my interest.

We can pose initial questions based on prevailing theory (drawn from any number of sources such as current status neurobiology, personality theory, Freudian theory, CBT-based theories, etc.) at various 'levels-of-analysis' and create a situation in which we can *meaningfully test* our theorizing by 'running an experiment' and give our educated guesses about how it will turn out or how people would respond a chance to actually fail (in which case, we'd need to revise our theory or our thinking about how we operationalized our theory in that instance).

Over my career as a student of clinical psychology as well as a practitioner, the most pernicious enemy of the quest to successfully apply the philosophy of science to subjects in our field that I've encountered is the tendency of the vast majority of academics/clinicians to fail to enter into their activities with a true spirit of honestly: (a) realizing that they don't know everything and may have something to actually learn from the results of their experiments or critical questions that others may have (including their own students); and (b) the fact that their thoughts and actions are massively influenced by personal contingencies of reinforcement/punishment (put simply, they act in their own self-interest at every stage of their 'scientific' endeavors). I think this has only worsened over the years with the phenomenon of needing to continually chase grant money as well as the fact that you're rewarded for churning out publications (regardless of merit or importance) that basically 'agree with' the current paradigm (trivial 'puzzle-solving' in the Kuhnian sense) rather than actually striving to challenge (and potentially falsify) prevailing models which are--at the very least--incomplete accounts of the phenomena they purport to cover (I'm looking at you, CBT) but which current academics/practitioners within the field are loathe to challenge/criticize because of the 'gravy train' reality that it's the model they use to justify their salaries and their practices.

I used to teach a seminar for interns on the application of the principles of the philosophy of science (proper) to clinical psychological practice and one of the most important things I highlighted in the first section was a quote by Charles Sanders Pierce about which I said, 'If you learn or remember anything about this course and keep it with you as you go through your career as a clinical psychologist, let it be this quote (below):'

"There is one thing even more vital to science than intelligent methods; and that is, the sincere desire to find out the truth, whatever it may be."

Look back on your career. Ask yourself, how many of your mentors TRULY engaged in their work with this spirit (especially the parts about the SINCERE desire to find out THE TRUTH, ***whatever it may be?***

THAT is the issue at the heart of what caused the replication crisis as well as all of the ways that we fall short in our field.

 

[ccool]

"There is one thing even more vital to science than intelligent methods; and that is, the sincere desire to find out thwhatever it may be."
Look back on your career. Ask yourself, how many of your mentors TRULY engaged in their work with this spirit (especially the parts about the SINCERE desire to find out THE TRUTH, ***whatever it may be?***

THAT is the issue at the heart of what caused the replication crisis as well as all of the ways that we fall short in our field.

Full agree. But I think to argue that people’s choice and/or ability to do this is mostly a function of personal weaknesses or barriers, which for some people it is, is to obscure the systems in place that reward this behavior and swiftly punish those that do not fall in line.

That said, I think all of this misses the forest for the trees RE: the replication crisis and broader issues in the field. Misaligned incentives are the problem. Neither theory nor pre-registration fixes "If this grant doesn't get funded I'm going to get laid off" or "If this paper doesn't get into a high-impact journal I won't get tenure." Everything else is barely going to move the needle.

You mention this and ollie makes this clear in their point above, there often are major professional, social, and/or psychological costs to pursuing and articulating potential truths that challenge the status quo. This occurs in small interactions like a lab or clinician or department meeting or activities like developing a manuscript and other professional activities. And even if you haven’t personally been at the receiving end of the costs (e.g. public degradation following an honest expression of feedback, a superior calling a meeting with you to subtly tell you what you said or did would cause professional and personal trouble for you, paper can’t get published for grounds that have nothing to do with the science), you have almost certainly witnessed it happen to someone else and most certainly have heard about it. These are strong and pernicious messages we receive across our career, and it is understandable that self-preservation might be a stronger value for some than authenticity or exploration.

That said, I am consistently shocked when I reflect on the fact that this is the state of things in a field of science, whose express purpose is to grow and learn and build and question and challenge. The thing that brings it over the top for me is how much and that it happens in the field of psychology, where our expertise is human behavior and mental processes. It’s like wow, damn, if we (mainstream psychology field at large) whose clinical and/or scientific expertise is optimizing human behaviors, experiences, and relationships are engaging in behaviors that actively or passively stymie growth, discovery, freedom then what hope is there…

We will continue to rely on the individuals who choose to persist regardless of the costs. Whose alignments towards authenticity or freedom or exploration for all are stronger than that of self-preservation in itself. Who continue to work to navigate the tensions between survival and integrity, particularly in spaces where there are a limited number of acceptable ways of being and doing…and risks to questioning what is acceptable or doing or being seemingly something outside the bounds are numerous and extremely undesirable. In doing so, these people create space for more of us to do so now and in the future. But it is up to us to support these change makers and to utilize the space they are creating, and that we all have the power to create, to keep moving forward as well.

 

[futureapppsy2]

One really important thing to note is that issues like the replication crises, harking, p-hacking, etc, show up everywhere, especially in medicine. I remember reading one study early on in the pandemic that purported to be about pediatric COVID-19 but defined COVID as having literally any gastrointestinal or upper respiratory symptoms, despite the fact that those are highly non-specific, especially in children. Still, the study was published in a great journal and shared uncritically in the media and on social media. I think sometimes psychology acts like these issues are not a thing in other, “harder science” fields, and they 100% are.

 

[summerbabe]

Interesting topic, thanks for posting. I would definitely support a push for both more theory and for it to be more effectively incorporated into understandable education for patients.

Working in the VA, PTSD treatment seems to go something like assessment (sometimes quite poor) —> diagnosis —> offer CPT and PE as evidence based treatment approaches. But my hunch is a lot less (or no) time is spent discussing/providing education on topics such as why PTSD may have developed in the first place since the vast majority of Criterion A eligible events never turn into clinical PTSD.

I try to provide some of this theory/info when I’ve had the chance but even when you read things like Foa’s original work, the theory portion seems to get glossed over so I feel like I’m kinda working piecemeal since I’m not a ‘PTSD person’.

One current theory I do find quite helpful is Thomas Joiner’s interpersonal theory of suicide. With patients, I’ll hear things that sound a whole lot like perceived burdensomeness and I can push to inquire about acquired capability, which I think helps with both more accurate risk assessment as well as targets for possible intervention.

 

[EnPsychlopedia]

I 100% agree that academic culture is a major driver of why QRPs, the file drawer problem, and journal practices exist as they currently do. I’m also aware of the underlying issues of what caused the replication crisis and how it affects other fields at large. What I’m really interested in here is your perspectives on if fixing the underlying problems leading to the replication crisis (and thereby hopefully producing a more replicable body of literature) is enough to say we’ve finally found our voice as a somewhat young science. I guess I’m hitting at the more fundamental question of whether a literature of replicable effects is enough to truly a make science reflective of reality if there is no theoretical backing to ground and unify that science. I have no official perspective on this issue and simply find the discussions of it fascinating. I’m learning through more reading that the “methods and practices only” reformers and the “we need theory along with research praxis reform” are often contentiously at odds…just a disclaimer that I don’t want to push either side, just learn how others in the field think about the future of our beloved science. Thanks for engaging so far!

 

[EnPsychlopedia]

Interesting topic. I'm drowning in work so won't have time for a detailed response but some general thoughts.
- I'd push back on the idea that evolution is a fully unifying theory. That is perhaps true for a small subset of the work that constitutes biology and touches on evolution, but there are whole subfields of biology where evolution has little to do with the day-to-day. To say its a unifying theory for the field seems grandiose. In this regard, we could eventually have a unifying theory for all of psychology but I'd frankly be happy if we could come up with a unifying theory of a sub-sub-sub-subfield.

I agree with most of what you’ve said, but I think calling evolutionary theory a unifying theory for biology is fair. While many biologists and biological subfields don’t necessarily interact with evolution on a daily basis, it’s engrained in every professional biologist from Day 1 to view all of life and all of biological functions through the lens of common descent, natural selection, and so forth. Maybe a microbiologist studying human inmune cells doesn’t interact with evolutionary predictions every time he/she looks at the nucleus of a T-cell, but he/she *knows,* based on deeply held and unifying principles of biology, that everything under that microscope is the product of billions of years of evolution and reproductive change. Evolution is so deeply engrained in the genome of biology that it cannot be removed and the science as a whole remain intact. Psychology has nothing which even approaches that level of being engrained field-wide. Principles of learning and conditioning are probably the closest, and those are far from being “in the DNA” of what being a psychologist means. Just my $0.02…maybe wishing for something akin to evolutionary theory in psychology is a pipe dream.

 

[Fan_of_Meehl]

I 100% agree that academic culture is a major driver of why QRPs, the file drawer problem, and journal practices exist as they currently do. I’m also aware of the underlying issues of what caused the replication crisis and how it affects other fields at large. What I’m really interested in here is your perspectives on if fixing the underlying problems leading to the replication crisis (and thereby hopefully producing a more replicable body of literature) is enough to say we’ve finally found our voice as a somewhat young science. I guess I’m hitting at the more fundamental question of whether a literature of replicable effects is enough to truly a make science reflective of reality if there is no theoretical backing to ground and unify that science. I have no official perspective on this issue and simply find the discussions of it fascinating. I’m learning through more reading that the “methods and practices only” reformers and the “we need theory along with research praxis reform” are often contentiously at odds…just a disclaimer that I don’t want to push either side, just learn how others in the field think about the future of our beloved science. Thanks for engaging so far!

These are hardly new ideas but some may include:

1) a much more thorough, rigorous and formalized training component in the philosophy of science proper during graduate school training (e.g., concepts such as roots of positivism/ logical positivism within the natural sciences, operationism (a la Percy Bridgeman), Feyerabend's work, Philip Kitcher ('Believing Where We Cannot Prove'; 'hypotheses are tested in bundles...'; 'fallibility is the hallmark of science') etc.); I think the 'clinical science' folks are arguing for this (and Stephen Hayes likes this idea as well)
2) more emphasis on exploratory work (hypothesis / theory generation) rather than sole emphasis on 'confirmation' of existing theories
3) valuing the scientific impact/worth of studies rather than just a 'tally count' of how many trivial 'studies' (which are 'low risk') confirming the current paradigm an academic can produce
4) lesser emphasis on p values and 'statistical significance' and more emphasis on interpreting effect sizes; relatedly, encouraging research designs that make 'risky predictions' or counterintuitive predictions rather than connecting obvious empirical dots; maybe more emphasis on multivariate methods and even machine learning (or its subset, deep learning) algorithms
5) finding ways to ensure that most (if not all) studies are available to the public and not 'paywalled' so that anyone wanting to read/critique them who doesn't (or their institution doesn't) have a subscription has to pay ludicrous fees just to read the damned thing

A couple of my favorite authors that I read extensively (even though much of it was on my own (it wasn't all required reading)) were Paul Meehl (obviously), Jacob Cohen (stats/methodology/ philosophy of science), and Richard McFall. It's been around 25+ years since I was in grad school but I wonder how much the work of these folks is emphasized these days.

As a clinical psychologist, it is most disconcerting when students, interns, or new psychologists express a certainty that we sort of 'have it all figured out' and that XYZ cognitive-behavioral protocol for disorder/syndrome X (plug and chug) represents the epitome of the assessment --> case formulation --> treatment pathway. The actual definition of 'evidence-based psychotherapy' is the intersection of: (a) empirical literature; (b) clinician competence/training; and (c) client preferences. At least this is the APA's definition. I have NEVER seen--anywhere--in print a definition of 'evidence-based' psychotherapy that equated this concept solely with the use of fixed-length protocols for which the agenda for every single session had already been pre-written. Many folks new to the field (and many experienced in the field) would do better with LESS certainty in the utter scientific infallibility of their methods.

 

[EnPsychlopedia]

These are hardly new ideas but some may include:

1) a much more thorough, rigorous and formalized training component in the philosophy of science proper during graduate school trainings.

I agree with this, for sure! My philosophy undergraduate courses were among the most formational for me with regard to developing critical thinking and logic skills, as well as developing the mindset of never believing wholesale (or without a proof) that we’ve fully reduced a problem down into all its parts.

 

[Fan_of_Meehl]

I agree with this, for sure! My philosophy undergraduate courses were among the most formational for me with regard to developing critical thinking and logic skills, as well as developing the mindset of never believing wholesale (or without a proof) that we’ve fully reduced a problem down into all its parts.

I took an elective philosophy of science course (in the philosophy department for the philosophy PhD students) near the very end of my grad school experience and it was SO useful and eye-opening. They really focused on how the philosophers of science (and physicists, chemists, biologists) around and just after the turn of the 20th century utilized the basic principles of the philosophy of science in the various areas of the natural sciences leading, of course, to exponential advances in all of those fields during the 20th century. It was really cool to see the origins of concepts within psychology (such as 'operationism/ operational definitions' and 'logical empiricism') and realize that the psychologists who emphasized these things weren't the originators of the concepts and that they borrowed them from the natural sciences. It was also incredibly useful to see how contentious the debates were among philosophers of science on various topics/issues which served as an antidote to the unfortunate tendency for a lot of people in professional psychology to think that 'the science is settled' and the choices that have been made in terms of how to approach scientific psychology (underlying philosophical presumptions/assumptions) have already been definitively chosen and settled for us and that there is merely 'one way' do approach our field from a scientific perspective.

 

[EnPsychlopedia]

Jumping back into this to say that I agree with those of you who are pointing out the silly territorialism present in the quantitative vs. qualitative debates. All methods, as long as they are rigorous, systematic, and empirical, should be welcome.

 

[Ollie123]

What I’m really interested in here is your perspectives on if fixing the underlying problems leading to the replication crisis (and thereby hopefully producing a more replicable body of literature) is enough to say we’ve finally found our voice as a somewhat young science. I guess I’m hitting at the more fundamental question of whether a literature of replicable effects is enough to truly a make science reflective of reality if there is no theoretical backing to ground and unify that science.

I guess I'm not really understanding the issue. I'd certainly posit we have a voice as a young science. Perhaps the best comparison is epidemiology. I don't think they have anything approaching on a unifying theory. They're in a similar boat to us in many ways. They are (generally) less theory-driven than we are, but also have much less of an identity crisis than we do about not being a "real" science. This is despite the fact that their field relies heavily (in some cases entirely) on purely observational work and is thus arguably sits further from "traditional science" than most of psychology.

We may not have that agreement on a particular issue, but I don't in any way believe a unifying theory is fundamentally necessary. At the same time, I also don't believe fixing the underlying problems leading to the replication crisis is: A) Going to happen by current efforts (see earlier comment - efforts are dealing with 1% of the problem and ignoring 99% of the problem); B) Really has that much of an impact one way or another. Its generally good. I support their efforts for the most part, though I do think their approach has been misguided and (in some cases) selfish. Should have started with academic administrators/leadership and funders rather than throwing all their junior colleagues under the bus.

 

[EnPsychlopedia]

I guess I'm not really understanding the issue. I'd certainly posit we have a voice as a young science. Perhaps the best comparison is epidemiology. I don't think they have anything approaching on a unifying theory. They're in a similar boat to us in many ways. They are (generally) less theory-driven than we are, but also have much less of an identity crisis than we do about not being a "real" science. This is despite the fact that their field relies heavily (in some cases entirely) on purely observational work and is thus arguably sits further from "traditional science" than most of psychology.

We may not have that agreement on a particular issue, but I don't in any way believe a unifying theory is fundamentally necessary. At the same time, I also don't believe fixing the underlying problems leading to the replication crisis is: A) Going to happen by current efforts (see earlier comment - efforts are dealing with 1% of the problem and ignoring 99% of the problem); B) Really has that much of an impact one way or another. Its generally good. I support their efforts for the most part, though I do think their approach has been misguided and (in some cases) selfish. Should have started with academic administrators/leadership and funders rather than throwing all their junior colleagues under the bus.

Thanks for your perspective. I’m not sure having a unifying theory matters, either. Like I said, I’m genuinely curious to read multiple perspectives. I’m still mulling over the readings I’ve been doing. I’ll certainly think about what you’ve said here, though, as much of what you’ve said seems intuitively correct to me. I certainly agree that we have a voice...what I should have said is "have we truly defined what unites a social psychologist and a personality psychologist, or an I/O psychologist and an evolutionary psychologist?" Just an interest in human behavior? Behavior geneticists also study human behavior, as do sociologists, economists, etc., but none of them call themselves "psychologists." What defines what differentiates psychology when, thus far, the only real definition is "the scientific study of behavior and mental processes...[?]"

I hadn't really thought about epidemiology lacking an underlying theoretical base, and haven't seen it mentioned in my readings on this topic. It's an interesting line of thought, for sure, although I'm wondering if the analogy fully works since epidemiology--as you've astutely described--is almost entirely based on observational trends and inference and that is all it needs to be, since its main goal is to investigate the prevalence and incidence of disease, which is a largely mathematical enterprise. I acknowledge that epidemiology is more than just this, I'm just trying to point out that maybe (if those who push for more theory are correct) a larger portion of epidemiology is reducible sans theory relative to psychology as a broad discipline. I'm not convinced that is true, and I'm not attempting to convince anyone else that it's true, either...I just think that this is a fascinating subject and it interests me how diverse the viewpoints are.

Finally, I again agree that the mainstream proposals for "fixing the problem" of journalistic practices and QRPs won't solve the larger academic culture issue. That's something we agree on, 100%. I'm asking if, an imaginary world, a magic wand was waved and all the methodological, journalistic, and administrative/academic culture issues went away, would theory still be needed for psychology to stay healthy? Again, maybe...and maybe not. I certainly don't think we will find the answer on this forum, but I do very much enjoy the intellectual discussion...it helps me exercise my good ole thinking machine.

 

[Ollie123]

I'm asking if, an imaginary world, a magic wand was waved and all the methodological, journalistic, and administrative/academic culture issues went away, would theory still be needed for psychology to stay healthy? Again, maybe...and maybe not.

Ahh. This is a very different question than it sounded like you were asking in the original post. Of course it will. Are there really people arguing against that? I've met a fair number of the folks pushing the open science, replication project, etc. line of work and I've never heard anyone say anything remotely like that. Ultimately, theories are needed to guide and shape the questions we pursue and the data we collect. Data across studies needs to be integrated in a parsimonious way for there to be any meaningful lasting impact. There are certain exceedingly narrow questions where that may not be true (e.g., "Does this treatment work better in group or individual format"), but globally it does and good luck developing new and better interventions without some kind of guiding theory. We could perhaps argue the relative focus. Certainly, theory alone would get us practically nowhere just like a giant correlation matrix of 10,000 variables collected from everyone on earth wouldn't get us very far.

 

[EnPsychlopedia]

Are there really people arguing against that? I've met a fair number of the folks pushing the open science, replication project, etc. line of work and I've never heard anyone say anything remotely like that.

Evidently there are! I'm with you in that I had not personally really encountered this train of thought either, though the paper in the original post certainly seemed to think it was a debate worth publishing in the Review of General Psychology. One of the pieces he cites in favor of the claim that some reformers don't value theory is the fact that this paper created a pretty large kerfuffle on science social media (like Twitter, etc...though of course we need to not fool ourselves into thinking that Twitter is representative). The paper certainly at least generated controversy, prompting responses from other reformers. Flis argues that some reformist psychologists focused on methodological factors implicated in the replication crisis have to some degree adopted the reformer persona as a way to build a career on social media, blogs, and in peer-reviewed publishing of critical psychology works which miss the point or at least miss the forest (theory building) for the trees (procedural reform). It's reassuring to hear that others here don't seem to have the same perception, so maybe sampling bias is afoot in the claim that theory-negligent reformers represent a significant portion of research psychologists.

 

[ccool]

Evidently there are! I'm with you in that I had not personally really encountered this train of thought either, though the paper in the original post certainly seemed to think it was a debate worth publishing in the Review of General Psychology. One of the pieces he cites in favor of the claim that some reformers don't value theory is the fact that this paper created a pretty large kerfuffle on science social media (like Twitter, etc...though of course we need to not fool ourselves into thinking that Twitter is representative). The paper certainly at least generated controversy, prompting responses from other reformers. Flis argues that some reformist psychologists focused on methodological factors implicated in the replication crisis have to some degree adopted the reformer persona as a way to build a career on social media, blogs, and in peer-reviewed publishing of critical psychology works which miss the point or at least miss the forest (theory building) for the trees (procedural reform). It's reassuring to hear that others here don't seem to have the same perception, so maybe sampling bias is afoot in the claim that theory-negligent reformers represent a significant portion of research psychologists.

it is extremely hard to imagine that anyone is theoretically against theory development.

However, theories are only useful to the extent that they have bearing on the real world.

I'm asking if, an imaginary world, a magic wand was waved and all the methodological, journalistic, and administrative/academic culture issues went away, would theory still be needed for psychology to stay healthy?

So while this question may be intellectually interesting, it has low practical value because of how extremely far away we are from resolving the methodological, administrative, professional and academic culture barriers that stand in the way of doing this. These things can’t be abstracted out. Who is going to do this work? At what costs? How long will it take to get published? Who is going to utilize their work? At what costs? Will enough people examine and utilize the theory for it to reach some kind of viability? When will that be? And so on.

And I can imagine that the *replication only people might have positioned themselves in a we can’t talk abt theory until we’ve tackled methodological and journalistic integrity issues argument. But idk. And the other camp more like regardless of and maybe despite that, w/o theories these problems the “replication crisis” will continue bc we’re not really operating from strong theories and are relying on post-hoc explanations argument.

But also, what does it mean for psychology the study? practice? to be or “stay healthy”?

 

[smalltownpsych]

This discussion is great and reminds me of why I got into this field. I am a little more practically based in my thinking these days, but these types of questions are still at the heart of what I do every single day. How to apply these grand theories to help the individual keeps me enagaged and challenged and I always question “accepted wisdom“ and especially my own because it is so often wrong!

 

[borne_before]

First and foremost, we need clarify what we mean by psychology. I think it's useful view the field as being defined by the goals of psychology, which are the goals of any natural science but applied to human behavior: to describe, explain, predict, and influence outcomes.

This is a fun one for a couple reasons. But, I think we have something to learn from out physics friends, who are quite good at using multiple models, laws, and theories concurrently: As Richard Feynman notes (who wasn't a fan of philosophy of science), theories, laws, etc., are approximations. I think physistist are probably some of the most mentally flexible people for their ability to interpret things from various vantage points.

No models/theories are perfect, and models/theories can be extremely helpful. But the usefulness and accuracy of the theory/model depends on the goal of using it or it's function/purpose.

Human behavior is complex, and our goals determine the utility of a theory. If we're building a house, do we need more tools than a hammer?

There will never been a unifying theory - but theory is important for the goals of psychology. However, more important is the ability to choose the correct theory/model depending on the situation.

For example, I'm pretty behavioral. Personally, I think it does the best job at modifying behaviors in the kiddos I work with (lots of neurodevelopmental disorders). However, when I'm doing therapy with a 15 year old girl who is depressed, I often reach for other tools.

 

[Fan_of_Meehl]

First and foremost, we need clarify what we mean by psychology. I think it's useful view the field as being defined by the goals of psychology, which are the goals of any natural science but applied to human behavior: to describe, explain, predict, and influence outcomes.

This is a fun one for a couple reasons. But, I think we have something to learn from out physics friends, who are quite good at using multiple models, laws, and theories concurrently: As Richard Feynman notes (who wasn't a fan of philosophy of science), theories, laws, etc., are approximations. I think physistist are probably some of the most mentally flexible people for their ability to interpret things from various vantage points.
View attachment 355928

No models/theories are perfect, and models/theories can be extremely helpful. But the usefulness and accuracy of the theory/model depends on the goal of using it or it's function/purpose.

Human behavior is complex, and our goals determine the utility of a theory. If we're building a house, do we need more tools than a hammer?

There will never been a unifying theory - but theory is important for the goals of psychology. However, more important is the ability to choose the correct theory/model depending on the situation.

For example, I'm pretty behavioral. Personally, I think it does the best job at modifying behaviors in the kiddos I work with (lots of neurodevelopmental disorders). However, when I'm doing therapy with a 15 year old girl who is depressed, I often reach for other tools.

To me, one of the coolest aspects of clinical psychology is its place at the intersection of a multitude of disciplines (scientific and otherwise) as well as the utility of drawing upon these myriad disciplines at multiple different levels-of-analysis. Subjects as broad-ranging as history (and the personalities, mass hysteria, sociology involved), philosophy (very apparent if you've ever taken or taught a history and systems course), biology/neurology/medicine, mathematics/statistics, computer science (AI), even art/literature, religion and mythology (and what they have to tell us about the organization of the human psyches that produced them over the years). Studying operant behavior at the individual level is relevant but so is neurobiology and the effects of psychotropic medications and illicit drugs. And then there's the interesting areas of intersection in such sub-disciplines as 'interpersonal neurobiology,' for example. The study of buddhism (and its history) is relevant to understanding of the origins of, say, mindfulness meditation and its psychotherapeutic applications.

It's one of the few remaining disciplines that still allows one to be a polymath, of sorts, and to continue to justifiably delve deeply into a number of varied interests. I suppose that is the gift that we get at the price of lack of precision and prediction/control of the phenomena under study.

 

[EnPsychlopedia]

However, theories are only useful to the extent that they have bearing on the real world.

Indeed. However, I think many would argue the same about replicable effects, at least in terms of how well we can explain them without some kind of theoretical grounding. It's pretty easy, on balance, to observe that, e.g., "Exposure to x increases the risk of y to the nth degree." It's much harder to explain or generalize this finding without theory.

So while this question may be intellectually interesting, it has low practical value because of how extremely far away we are from resolving the methodological, administrative, professional and academic culture barriers that stand in the way of doing this. These things can’t be abstracted out.

I totally agree that we have a looooooong way to go as yet to fix these issues. I guess the question is: Does theory play a role in the reform movement now, or is it enough to focus on fixing these methodological issues and hope that our replicable effects can be organized into theory down the road? It seems most folks here seem to think that theory development is of at least some importance at the current stage, which I think is correct.

First and foremost, we need clarify what we mean by psychology. I think it's useful view the field as being defined by the goals of psychology, which are the goals of any natural science but applied to human behavior: to describe, explain, predict, and influence outcomes.

Yes, defining psychology is often much harder than people give it credit for being. I think your definition definitely exists as the most standard definition, but we still suffer from a bit of a problem of ambiguity--after all, sociology, anthropology, behavioral biology, etc. are all interested in applying scientific methods to study human behavior, but they are typically differentiated from "psychology," even if the methods and goals are quite similar. Maybe the difference, sans unifying theory, is the lens? A sociologist looks at behavior like a sociologist does, a biologist like a biologist does, etc.

There will never been a unifying theory

I'm quite sure this is correct, but it's nice to dream! I probably have a little bit of Freudian envy toward the natural sciences, particularly chemistry and biology, in that they do have unifying theories. *sigh*

Thanks to all for engaging this topic! It really is an interesting exercise, and unlike many posts here is simply an exercise in discussion (no advice to provide, likely no immediately practical information being meted out). I'm happy to have stimulated some brain activity.

 

[Fan_of_Meehl]

Indeed. However, I think many would argue the same about replicable effects, at least in terms of how well we can explain them without some kind of theoretic grounding. It's pretty easy, on balance to observe that, e.g., "Exposure to x increase the risk of y to the nth degree." It's much harder to explain or generalize this finding without theory.

I totally agree that we have a looooooong way to go as yet to fix these issues. I guess the question is: Does theory play a role in the reform movement now, or is it enough to focus on fixing these methodological issues and hope that our replicable effects can be organized into theory down the road? It seems most folks here seem to think that theory development is of at least some importance at the current stage, which I think is correct.

Yes, defining psychology is often much harder than people give it credit for being. I think your definition definitely exists as the most standard definition, but we still suffer from a bit of a problem of ambiguity--after all, sociology, anthropology, behavioral biology, etc. are all interested in applying scientific methods to study human behavior, but they are typically differentiated from "psychology," even if the methods and goals are quite similar. Maybe the difference, sans unifying theory, is the lens? A sociologist looks at behavior like a sociologist does, a biologist like a biologist does, etc.

I'm quite sure this is correct, but it's nice to dream! I probably have a little bit of Freudian envy toward the natural sciences, particularly chemistry and biology, in that they do have unifying theories. *sigh*

Thanks to all for engaging this topic! It really is an interesting exercise, and unlike many posts here is simply an exercise in discussion (no advice to provide, likely no immediately practical information being meted out). I'm happy to have stimulated some brain activity.

Philip Kitcher's manuscript 'Believing Where We Cannot Prove' is a brilliant essay examining the crucial role of theory in science.

Also, the revolutionary progress in applied behavioral analysis relies heavily on theory (as any approach does) as well as epistemological decisions regarding how to properly understand its subject matter and define its methods (e.g., 'operational definitions' are absolutely key to behaviorism, including the operational definition of 'reinforcement' vs. 'punishment,' which differs from the lay conception of those terms). Skinner relied heavily on concepts from the philosophy of science proper (such as 'operationism' [known these days in psychology circles as 'operational definitions') which goes all the way back to near the turn of the century when Percy Bridgman wrote, 'The Operational Character of Scientific Concepts' which was written to address the application of the philosophy of science to the natural sciences such as chemistry, physics, and biology and which touched off the revolutionary amount of progress we saw in the 20th century in those scientific disciplines. There were publications in the early days of behaviorism (70's?) that explicitly highlighted the link between Bridgman's work in the philosophy of science and the roots of behaviorism.