Base Rates of Invalid Test Performance Across Clinical Non-forensic Contexts and Settings, 2020, Martin and Schroeder

[Andy]

Objective
Base rates of invalidity in forensic neuropsychological contexts are well explored and believed to approximate 40%, whereas base rates of invalidity across clinical non-forensic contexts are relatively less known.

Methods
Adult-focused neuropsychologists (n = 178) were surveyed regarding base rates of invalidity across various clinical non-forensic contexts and practice settings. Median values were calculated and compared across contexts and settings.

Results
The median estimated base rate of invalidity across clinical non-forensic evaluations was 15%. When examining specific clinical contexts and settings, base rate estimates varied from 5% to 50%. Patients with medically unexplained symptoms (50%), external incentives (25%–40%), and oppositional attitudes toward testing (37.5%) were reported to have the highest base rates of invalidity. Patients with psychiatric illness, patients evaluated for attention deficit hyperactivity disorder, and patients with a history of mild traumatic brain injury were also reported to invalidate testing at relatively high base rates (approximately 20%). Conversely, patients presenting for dementia evaluation and patients with none of the previously mentioned histories and for whom invalid testing was unanticipated were estimated to produce invalid testing in only 5% of cases. Regarding practice setting, Veterans Affairs providers reported base rates of invalidity to be nearly twice that of any other clinical settings.

Conclusions
Non-forensic clinical patients presenting with medically unexplained symptoms, external incentives, or oppositional attitudes are reported to invalidate testing at base rates similar to that of forensic examinees. The impact of context-specific base rates on the clinical evaluation of invalidity is discussed.

Paywall, https://academic.oup.com/acn/article-abstract/doi/10.1093/arclin/acaa017/5822934
Sci hub, https://sci-hub.tw/10.1093/arclin/acaa017

Base rates of invalidity have been well explored in forensic neuropsychological settings over the past two decades. Mittenberg, Patton, et al., (2002) initially surveyed board-certified neuropsychologists conducting forensic work. From their study, it was found that estimated base rates of malingering and symptom exaggeration were roughly 30% for examinees pursuing personal injury, disability, or workers’ compensation claims and roughly 20% for examinees with criminal charges. Additionally, neuropsychologists estimated that greater than 30% of forensic examinees with mild head injuries, fibromyalgia/chronic fatigue, or chronic pain met the criteria for probable malingering. Larrabee (2003) examined the empirical literature by reviewing 11 studies containing 1,363 consecutively evaluated litigants with mild traumatic brain injuries, finding that 40% of the examinees failed validity testing. Larrabee, et al., (2009) reviewed additional base rate data and found that rates of invalidity in settings where external incentives are present typically ranged between 30% and 50%, leading to the conclusion that the data “are sufficiently consistent to propose a new ‘magical number’ in neuropsychology: 40 ± 10” (p. 843)

 

[Mithriel]

So despite the continual emphasis that "the symptoms are real to the patient" we are back to 50% of MUS patients are lying. There goes the PIP

 

[rvallee]

I'm trying to make sense of what is being communicated here but the writing is so bad I can't find it. It's just a word salad with a few sentences that come close to making sense but never quite make it. I think it's trying to make a case for confirmation bias and the god of the gaps by cherry-picking things and arguing stuff to support the prior conclusion that malingering is possibly something something. Best I can do, sorry.

Even great writers need editors. Bad writers especially need them. Stringing words together isn't good writing anymore than stringing code together makes for a functional program.

Though this is a great example for the problem of how STEM skills are overvalued above "soft" skills like communication. You need to be able to communicate if you want to communicate technical stuff.

 

[Amw66]

How coherent are the Larabee papers quoted ?

 

[alktipping]

what I get from this is their happy to get it wrong 50% percent of the time . me I know they get it wrong 100% of the time it is all based on ridiculous notions of history that other people are sick because of some emotional weakness I truly wish there was no financial incentive to carry on with this grossly stupid idea .

 

[rvallee]

How coherent are the Larabee papers quoted ?

Better literary skills than this, but the substance is just as bad. It takes suspected cases of malingering then uses tests and questionnaires to confirm they are malingering. Which of course no test or questionnaire can do.

Just your typical put stuff in a box, label that box then confirm you labeled that box correctly since the conditions to be put in that box were met to personal satisfaction. It seems to be a different context but the same mentality for arguing that imprisoned people must be guilty since otherwise they wouldn't have been imprisoned in the first place, validating the assumption that no over has ever been wrongfully imprisoned and confirming there is no need to check individual cases.

People are weird. Smart people can be a whole special kind of weird.