Paper criticizing CBT for EDs

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.
Get help with your application

Use all the free resources available to you from SDN: articles, guides, expert advising, forums discussions, and school research.

Advertisement - Members don't see this ad
To be clear, there was no intentional collection of those folks or those with that diagnosis for collection with the EX. The result is that no diagnosis was captured for the normative sample, and no inclusion for those folks outside of paid normative sampling based on demographics of the census- which is unlikely to do a whole ton.

It snt in the tech manual. It was personal correspondence with the test author. Ive been part of a lot of mmpi/pai test development and I can tell ya, the tech manual and those reading it regularly miss many of the findings it in. And I can tell you explictly that reviews are NOT rigorous and that a lot of the test is just untested theory and weak science. For instance, how many longitudinal studies on the substantive scales exist (not many)? The validity theory is weak and limited. Recent tests like the iop see it the same way I do- domains dont show up on self report in a reliable and meaningful way
But...with a large enough normative sample, doesn't sampling theory dictate that those people who happen to have those psychotic disorders would be 'represented' in the normative sample (as long as there did not exist some specific bias/reason that they'd be excluded from the process?). Now, obviously, cross-sectionally (currently) floridly-psychotic schizophrenics would be less likely to show up in that normative sample for obvious reasons. But the presence of that dimensional construct (or constructs) relevant to the 'thought disorder' dimensions (and sub-components under a HiTOP framework) would be represented in a large enough normative sample of the population by definition, no?
 
How is sampling affected, with disorders with equivalent cross cultural prevalence?
Are you suggesting that psychosis is seen, described, and experienced/interpreted with the same language norms as those without it, across the world?

Most of the items are just vague descriptors of things which are assumed normative but arent (eg psychosis). This approach is partly why dimensional models work better for some dimensions (int) than others. Its also why psychosis diagnosis favors clinical interview rather than self report as 'gold standard' (i hate that term). Honestly, your disbelief in the idea that people who experience a thing have anything to offer to understanding it and directing its measurement is kinda wild.
 
And (sincere question), are there any alternative broadband tests of personality/psychopathology that you would recommend using with this population (psychotic disorders)? Or any brief, self-report instruments that supposedly serve the function(s) that the various thought disorder scales on the MMPI-2-RF (or 3) attempt to estimate/measure as dimensional elements of psychopathology?
Honestly, none are great. To me the greatest issue is that assessment has abandoned itemageaphic focus and larger awareness of our roots, it leads to a lot of bias in the purpose and scope of testing. The empirical movement largely believes that a mean score tells everything. Im still fighting in every review to make people include cut score comparisons. Recently, mmpi group has shifted to describing ranges of scores as supporting claims, even when those ranges are within normative limits

Lower reading levels are key, and item level follow up. Rc8, for instance, elevates in 50%+ of college students. If it is a scale that measures a singular construct (thought disorder, per RC FA papers- see 2006 JPA special issue) then either it doesnt do that because 50%+ of college students dont have that- or it conflates the term. pai items are better since some mirror rare symptom patterns in psychosis (eg perceptual irregularities, etc). Its also a 4th reading level.
 
Advertisement - Members don't see this ad
To be clear, there was no intentional collection of those folks or those with that diagnosis for collection with the EX. The result is that no diagnosis was captured for the normative sample, and no inclusion for those folks outside of paid normative sampling based on demographics of the census- which is unlikely to do a whole ton.

It snt in the tech manual. It was personal correspondence with the test author. Ive been part of a lot of mmpi/pai test development and I can tell ya, the tech manual and those reading it regularly miss many of the findings it in. And I can tell you explictly that reviews are NOT rigorous and that a lot of the test is just untested theory and weak science. For instance, how many longitudinal studies on the substantive scales exist (not many)? The validity theory is weak and limited. Recent tests like the iop see it the same way I do- domains dont show up on self report in a reliable and meaningful way
I have the tech manual and use it from time to time but can you please elaborate more on what you mean by 'those reading it regularly miss many of the findings in it?'

Do you mean that (a) we're not reading/considering all of the info contained in it (largely, guilty-as-charged); (b) we are reading it but perhaps misinterpreting it (due to not fully understanding the context from whence the data come); or (c) misapplying it clinically? or (d) something else? I'm likely guilty of all/most of these since the technical manual for the -rf is dense and chock full of info.
 
I have the tech manual and use it from time to time but can you please elaborate more on what you mean by 'those reading it regularly miss many of the findings in it?'

Do you mean that (a) we're not reading/considering all of the info contained in it (largely, guilty-as-charged); (b) we are reading it but perhaps misinterpreting it (due to not fully understanding the context from whence the data come); or (c) misapplying it clinically? or (d) something else? I'm likely guilty of all/most of these since the technical manual for the -rf is dense and chock full of info.
Yeh ill give some full in depth answers later when im back home. I have a lot of concerns about modern intepretive approaches, including by those who are the test publishers ive repeatedly published with. Ive just had to explain statistics too many times to them.
 
Are you suggesting that psychosis is seen, described, and experienced/interpreted with the same language norms as those without it, across the world?

Most of the items are just vague descriptors of things which are assumed normative but arent (eg psychosis). This approach is partly why dimensional models work better for some dimensions (int) than others. Its also why psychosis diagnosis favors clinical interview rather than self report as 'gold standard' (i hate that term). Honestly, your disbelief in the idea that people who experience a thing have anything to offer to understanding it and directing its measurement is kinda wild.
Uhh, I was not implying anything and it's kinda wild that you are projecting so much onto an honest question.
 
Yeh ill give some full in depth answers later when im back home. I have a lot of concerns about modern intepretive approaches, including by those who are the test publishers ive repeatedly published with. Ive just had to explain statistics too many times to them.
Thanks! I've found your responses extremely helpful and thought provoking so far and I'm glad that we can have these sorts of discussions among professionals. I think it's how we get better at what we do.
 
You tend to use questions as answers. If you read everything as projection, then thats another issue.

Apologies if I misread. Ill respond later.
Some of your initial criticism is fair, some of that is not. You'll notice that I stick to the question, despite deflection (e.g., msc545 made a declaration as fact, and refused to give backing, and I'm still hounding them for some evidence). However, someone reading a motive into a neutral question is pretty clear projection.

Regardless of my disagreement with the premise of "decolonial", I do think the question is an interesting idea from a psychometric perspective. I'm sure you are better equipped to answer the question than I.
 
Last edited:
But...with a large enough normative sample, doesn't sampling theory dictate that those people who happen to have those psychotic disorders would be 'represented' in the normative sample (as long as there did not exist some specific bias/reason that they'd be excluded from the process?). Now, obviously, cross-sectionally (currently) floridly-psychotic schizophrenics would be less likely to show up in that normative sample for obvious reasons. But the presence of that dimensional construct (or constructs) relevant to the 'thought disorder' dimensions (and sub-components under a HiTOP framework) would be represented in a large enough normative sample of the population by definition, no?

In theory, yes. But psychosis (presuming you mean here delusional thought processes rather than actual hallucinogenic symptoms, which are typically observable) has a low population level base rate so you may need to over sample in order to obtain enough power to establish a factor. Even a four item scale would be tricky if the linked study is to be believed. If my goal is to turn a profit without expending a ton of effort, this would be an easy corner to cut.

It doesn't surprise me one bit to learn that less emphasis is being placed on items and more emphasis is being placed on mean scores. On my research postdoc at the local AMC, I was often in the position of being the lone personality researcher explaining to senior researchers why you can't just shove items together in a factor and run it as a mean score. Sure, the approximate fit indices look fine, but don't you think it's weird that item-total correlations for two items are essentially 0.20 when the other two are 0.90? No? Ok, cool.
 
@Justanothergrad My question was a genuine one. I am a pediatric psychologist. I have not given/interpreted an MMPI since fellowship lol. It sounds like you have much more of an inside track on this than I certainly would, so I was looking to your expertise to just learn more for the occasional FEP case that happens to come across my desk.

Personally, I tend to favor more of a discussion about cognitive deficits when discussing the population affected by psychosis. It oddly rarely gets enough discourse, in my opinion, but I feel like it's an observed deficit that anyone who has treated a psychotic disorder can tend to agree on. There definitely appears to be a lot to categorizing dysfunction based on cognitive impacts within the population at the very least.
 
"ED services rely heavily on Western diagnostics, rigid triage thresholds, and manualised psychosocial packages rooted in CBT; this architecture privileges low body mass index and acute medical risk and tends to overlook binge-spectrum, night-eating or sensory-driven presentations that may be more common in structurally disadvantaged or culturally distinct populations."

Yes, because ED services of this nature are those at HLOCs where there is an emergent medical component. Am I going to admit the night-eating patient with no issues in their lab values to RTC? Yeah, we need rigid triage thresholds and to privilege "acute medical risk" because we need to know where to place an ED patient to best address their medical component, psychiatric component, or both. It's almost like if you're physically ill, you need treatment (and the dx that gets you treatment) more than the person who is not in acute medical distress.

"CBT approaches for EDs assumes the primacy of maladaptive cognitions and avoidance behaviours as maintenance processes, and that modifying these intra-individual processes is sufficient for recovery. Lived experience-led research, especially among neurodivergent people, show that sensory processing differences, gastrointestinal disorders, systemic discrimination, and experiences of iatrogenic harm can be proximal drivers of restrictive and binge behaviours. When the primary illness drivers are somatic, sensory, or socio-politically mediated, cognitive techniques may not only fail but also exacerbate iatrogenesis, shame, and disengagement from care."

Good sir, no. There is significant emphasis in ED treatment, CBT included, that emphasizes interpersonal and social factors in the recovery process. Furthermore, if we're going to talk about actual implementation in practice, every ED company that I have worked for, visited, or referred patients to, does not do CBT in a vacuum. It is always supplemented with other modalities like DBT, ACT, exposure therapy, family therapy, etc. They're not doing cognitive restructuring worksheets all day. Furthermore, if a physical illness is the primary driver of the "eating disorder", the patient likely does not have an eating disorder, and it becomes a feeding concern that we move over to the internal med/OT folks.

I get what the author is trying to say. But the articles he cites are questionable, there is a lot of lived-experience and phenomenological accounts and some of the claims are straight up wrong or unreasonable. Asking us to remove the "privilege" afforded by acute medical risk is to ask us to not triage...
 
"ED services rely heavily on Western diagnostics, rigid triage thresholds, and manualised psychosocial packages rooted in CBT; this architecture privileges low body mass index and acute medical risk and tends to overlook binge-spectrum, night-eating or sensory-driven presentations that may be more common in structurally disadvantaged or culturally distinct populations."

Yes, because ED services of this nature are those at HLOCs where there is an emergent medical component. Am I going to admit the night-eating patient with no issues in their lab values to RTC? Yeah, we need rigid triage thresholds and to privilege "acute medical risk" because we need to know where to place an ED patient to best address their medical component, psychiatric component, or both. It's almost like if you're physically ill, you need treatment (and the dx that gets you treatment) more than the person who is not in acute medical distress.

"CBT approaches for EDs assumes the primacy of maladaptive cognitions and avoidance behaviours as maintenance processes, and that modifying these intra-individual processes is sufficient for recovery. Lived experience-led research, especially among neurodivergent people, show that sensory processing differences, gastrointestinal disorders, systemic discrimination, and experiences of iatrogenic harm can be proximal drivers of restrictive and binge behaviours. When the primary illness drivers are somatic, sensory, or socio-politically mediated, cognitive techniques may not only fail but also exacerbate iatrogenesis, shame, and disengagement from care."

Good sir, no. There is significant emphasis in ED treatment, CBT included, that emphasizes interpersonal and social factors in the recovery process. Furthermore, if we're going to talk about actual implementation in practice, every ED company that I have worked for, visited, or referred patients to, does not do CBT in a vacuum. It is always supplemented with other modalities like DBT, ACT, exposure therapy, family therapy, etc. They're not doing cognitive restructuring worksheets all day. Furthermore, if a physical illness is the primary driver of the "eating disorder", the patient likely does not have an eating disorder, and it becomes a feeding concern that we move over to the internal med/OT folks.

I get what the author is trying to say. But the articles he cites are questionable, there is a lot of lived-experience and phenomenological accounts and some of the claims are straight up wrong or unreasonable. Asking us to remove the "privilege" afforded by acute medical risk is to ask us to not triage...
I have increasingly seen (or perhaps been subjected to) the idea that if stress comes from discrimination then there’s nothing to be done via cbt.

I understand and agree that “resilience” can sometimes be done poorly and be “just shut up about and lump the discrimination you encounter” but the first position is insane.
 
I have increasingly seen (or perhaps been subjected to) the idea that if stress comes from discrimination then there’s nothing to be done via cbt.

I understand and agree that “resilience” can sometimes be done poorly and be “just shut up about and lump the discrimination you encounter” but the first position is insane.

I have encountered several of the "CBT is racist" m orons locally.
 
Advertisement - Members don't see this ad
"ED services rely heavily on Western diagnostics, rigid triage thresholds, and manualised psychosocial packages rooted in CBT; this architecture privileges low body mass index and acute medical risk and tends to overlook binge-spectrum, night-eating or sensory-driven presentations that may be more common in structurally disadvantaged or culturally distinct populations."

Yes, because ED services of this nature are those at HLOCs where there is an emergent medical component. Am I going to admit the night-eating patient with no issues in their lab values to RTC? Yeah, we need rigid triage thresholds and to privilege "acute medical risk" because we need to know where to place an ED patient to best address their medical component, psychiatric component, or both. It's almost like if you're physically ill, you need treatment (and the dx that gets you treatment) more than the person who is not in acute medical distress.

"CBT approaches for EDs assumes the primacy of maladaptive cognitions and avoidance behaviours as maintenance processes, and that modifying these intra-individual processes is sufficient for recovery. Lived experience-led research, especially among neurodivergent people, show that sensory processing differences, gastrointestinal disorders, systemic discrimination, and experiences of iatrogenic harm can be proximal drivers of restrictive and binge behaviours. When the primary illness drivers are somatic, sensory, or socio-politically mediated, cognitive techniques may not only fail but also exacerbate iatrogenesis, shame, and disengagement from care."

Good sir, no. There is significant emphasis in ED treatment, CBT included, that emphasizes interpersonal and social factors in the recovery process. Furthermore, if we're going to talk about actual implementation in practice, every ED company that I have worked for, visited, or referred patients to, does not do CBT in a vacuum. It is always supplemented with other modalities like DBT, ACT, exposure therapy, family therapy, etc. They're not doing cognitive restructuring worksheets all day. Furthermore, if a physical illness is the primary driver of the "eating disorder", the patient likely does not have an eating disorder, and it becomes a feeding concern that we move over to the internal med/OT folks.

I get what the author is trying to say. But the articles he cites are questionable, there is a lot of lived-experience and phenomenological accounts and some of the claims are straight up wrong or unreasonable. Asking us to remove the "privilege" afforded by acute medical risk is to ask us to not triage...
It’s like some people have never even read Harrison Bergeron.
 
It’s like some people have never even read Harrison Bergeron.

Anyone remember this one?

"
  • Xtacle 2: Yeah, so who's read Flowers for Algernon?
  • Nearl: Ken!
  • Xtacle 3: About the kid with all the chains and the goggles and at the end he gets killed with a shotgun?
  • Xtacle 2: No, that's ...
  • Xtacle 4: Boosh!
  • Xtacle 3: Boosh!
  • Xtacle 1: Boosh!
  • Xtacle 2: No ...
  • Xtacle 4: Buh-buh-Boosh!
  • Xtacle 2: That's Harrison Bergeron.
  • [pause]
  • Xtacle 1, Xtacle 3, Xtacle 4: Hollywood Squares!
  • Xtacle 2: That's Tom Bergeron!
  • Xtacle 3: Brother of Menelaus.
  • Xtacle 2: Dammit, that's Agamemnon!
  • Xtacle 3: Boosh!
  • Xtacle 2: Okay.
  • Xtacle 1: Boosh!
  • Xtacle 2: Greek."
 
I didn’t read the paper but really enjoyed the discussion that was had around it. Understanding cultural differences and how that impacts individuals and how that impacts treatment is an essential part of our ethical guidelines. Also, using evidence based practices is a key ethical standard as well. So we are supposed to do both and I don’t see how they would be mutually exclusive, in fact quite the opposite I would think.
Another point would be that if they have some new insights that would actually help treat eating disorders more effectively, that would be greatly appreciated. Extremely difficult to treat and some of the worst outcomes in our field. I suspect that they are just critiquing and adding nothing but noise as opposed to clarity or even anything that could help point research in the right direction.
Final point, apparently since the same large group that are the colonizers also were key in developing modern scientific method, I think o see the convergence of thought there. Colonialism bad, Europeans were colonizers - bad, Europeans also developed the modern scientific method therefore it’s bad too. Perfectly logical, Aristotle would be proud.
 
A gentle nudge:
To be considered legitimate, research went through major hubs like London and Berlin. Because of reasons. That would limit which standards and frameworks would be adopted and what research was disseminated. Even though other regions influenced the rational discourse, a lot of that was still funneled through European interpretation because their work was shared more frequently.

I agree with the sentiment though. I will wholeheartedly evaluate and critique medicine. That doesn't mean I am picking up homeopathic remedies instead.