WEDNESDAY, JANUARY 11, 2023
Nonwhites said the same things: As our Case Study nears its end, we return to the basic text of the UVa study in question.
We return to the very start of the four researchers' report on the UVa study. Their widely cited research report starts in the manner shown:
UVA STUDY (4/4/16): The present work examines beliefs associated with racial bias in pain management, a critical health care domain with well-documented racial disparities. Specifically, this work reveals that a substantial number of white laypeople and medical students and residents hold false beliefs about biological differences between blacks and whites and demonstrates that these beliefs predict racial bias in pain perception and treatment recommendation accuracy. It also provides the first evidence that racial bias in pain perception is associated with racial bias in pain treatment recommendations. Taken together, this work provides evidence that false beliefs about biological differences between blacks and whites continue to shape the way we perceive and treat black people—they are associated with racial disparities in pain assessment and treatment recommendations.
For the record, we've focused on the alleged false beliefs of those "white medical students and residents." We haven't focused on the alleged correlation between those alleged false beliefs and subsequent treatment recommendations.
In particular, we've focused on the way those alleged false beliefs have been described by journalists and academics over the past few years. Once again, here's the way Professor Sabin described those alleged false beliefs in an essay which was later cited as the source for an account of the UVa study in the Washington Post:
SABIN (1/6/20): “Black people’s nerve endings are less sensitive than white people’s.” “Black people’s skin is thicker than white people’s.” “Black people’s blood coagulates more quickly than white people’s.”
These disturbing beliefs are not long-forgotten 19th-century relics. They are notions harbored by far too many medical students and residents as recently as 2016. In fact, half of trainees surveyed held one or more such false beliefs, according to a study published in the Proceedings of the National Academies of Science. I find it shocking that 40% of first- and second-year medical students endorsed the belief that “black people’s skin is thicker than white people’s.”
The headline to Sabin's essay made it clear that the shocking medical students and residents in question were "white." You can see the whole essay here.
Yikes! According to Professor Sabin, those (white) medical trainees has given voice to shocking, disturbing false beliefs—false beliefs which seemed to be drawn from the days in which this nation was built on a foundation of human enslavement.
According to the UVa researchers, the (white) trainees had expressed an array of false beliefs. According to Professor Sabin, the false beliefs of these (white) trainees weere shocking and disturbing—and were reminiscent of this nation's baldly racist past.
The professor's assessment went well beyond what the researchers had said. But in each of these descriptions, readers were told that it was white trainees who held these false beliefs. Readers were given no reason to think that any other groups of respondents had been involved.
A bunch of white medical trainees had given voice to shocking, disturbing false claims—false claims which seemed to be racist, or at least racist-adjacent! At present, our highly unimpressive blue tribe seems to love making such statements more than we love life itself.
These White Medical Students Today! They hold shocking false beliefs concerning matters of race! Our flailing tribe seems to love such thrilling statements, even where such statements are grossly misleading, or are indeed simply wrong.
A bunch of white trainees had been surveyed, full stop. Based on anything anyone said, no "nonwhites" had been involved!
In fairness, it's easy to see where journalists and academics might have gotten that impression—but hold on! Way, way down at the very end of the UVa researchers' report, this bit of surprising news was offered as a slightly puzzling afterthought:
UVA STUDY: Readers may also be interested in analyses for nonwhite participants (Asian, n = 43; black, n = 21; biracial, n = 28; Hispanic/Latino, n = 11; other, n = 3).
Say what? As it turned out, the UVa researchers had also surveyed a substantial number of "nonwhite" medical students and residents!
As it turns out, the researchers had surveyed 222 white trainees—but they had also surveyed 108 of their nonwhite classmates and colleagues! But in a decision which may seem slightly peculiar, the researchers had decided to restrict their study to the responses of the white trainees alone.
For reasons which went unexplained, responses by the nonwhite trainees were, in effect, disappeared. Why would the researchers decide to do that?
We have no idea.
We've seen no attempt to explain this decision—but as a result of this decision, Professor Sabin was soon denouncing the white trainees in extremely colorful ways, without even mentioning the fact that nonwhite trainees had been surveyed too.
That said, we interrupt this morning's report to bring you some very good news:
In our judgment, very few of the medical trainees—"white" and "nonwhite" trainees alike—actually endorsed (agreed with) any disturbing or shocking beliefs in the course of the UVa study.
We'll offer even better news—it isn't entirely clear to us that any more than a handful of the medical trainees endorsed any false beliefs at all!
We'll postpone our reason for saying these things until tomorrow's report. For today, let's focus on a basic fact, a fact expressed by Kevin Drum in his perspicacious assessment of the UVa study itself.
Who knew! A group of nonwhite trainees had been surveyed, along with a bunch of disturbing whites! And not only that! After he journeyed very deep into the bowels of the UVa study, Drum was able to emerge with some surprising news.
Remember what's at issue here. The trainees were judged on the frequency with which they (allegedly) "endorsed" (expressed agreement with) eleven different (allegedly) false statements.
As we've noted, it seems to us that very few of the trainees endorsed any false statements at all. But this is what Kevin Drum found deep in the bowels of this widely cited UVa study:
Good lord! In the assessment of the UVa researchers, the nonwhite participants had endorsed a bunch of (allegedly) false beliefs too:
DRUM (1/4/23): Beliefs of white and nonwhite respondents are virtually identical. In particular the average score for nerve endings is 1.94 vs. 1.83 (nonwhite [participants] are more likely to believe it) and 1.76 vs. 1.73 for skin thickness. Overall, the belief in false statements is 2.06 vs. 1.98, meaning that nonwhite [participants] are slightly more likely to believe them than white participants].
Belief in false statements is not a problem. The percentages are low and the responses are almost all tentative.
When it comes to belief in false statements, "the responses are almost all tentative?" What the heck did Kevin Drum mean by that?
We'll review that point tomorrow. For today, let's get clear on something Drum clearly says in that passage.
Let's be clear on what Drum is saying! Based upon his review of all the responses by all the participants, he says that the responses of the (disappeared) nonwhite trainees were "virtually identical" to those of the (shockingly racist) white trainees!
In fact, the nonwhite trainees were "slightly more likely" to endorse the false statements than the white trainees were! The differences between the two groups were tiny and inconsequential, but that's what Kevin Drum found.
According to Drum, there was virtually no difference in the way the two groups assessed the 11 (allegedly) false statements. But having said that, so what?
Responses by the white participants have routinely been cited, for years, within the mainstream press. The white participants have been savaged for their shocking, disturbing false beliefs. The similar responses by the nonwhite participants have been disappeared.
The researchers reported the alleged false beliefs on the part of the white respondents. At the same time, they virtually disappeared the fact that the nonwhite respondents had offered assessments which were "virtually identical."
Our professors and journalists took over from there. Why does our blue tribe behave in these ways? Also, is there even the slightest chance that we'll ever stop?
(Disconsolate experts say no.)
We close today by restating the bit of very good news we adumbrated above:
In our assessment, very few of the respondents, white and nonwhite together, "endorsed" any false statements at all! Tomorrow, to explain why we would say such a thing, we'll turn again to the findings reported by Drum.
That said, our deeply flawed, routinely repulsive tribe is, at present, deeply in love with invidious racial claims. We love expressing these claims more than life itself, and we need to make ourselves stop.
Tomorrow: How many participants, "white" and "nonwhite" alike, actually endorsed any false statements at all?
ReplyDeleteYeah, medical science is a tool of WHITE SUPREMACY.
Except that here, in the Democratic Republic of the Congo it isn't!! And it's Democratic -- it's right in the name, you see?!
Come over, dear dembots, wherever you are, whenever you like!!
"For reasons which went unexplained, responses by the nonwhite trainees were, in effect, disappeared. Why would the researchers decide to do that?
ReplyDeleteWe have no idea.
We've seen no attempt to explain this decision—but as a result of this decision, "
As the paper itself clearly states, there was no significant difference on the pain rating and treatment tasks, unlike the white subjects. It is standard practice in research reports NOT to report results when no significant difference is found. You simply state that there was no significant difference. You do not report the results because there is nothing to see there, nothing more to say, when no difference is found. It wastes space and journal space is expensive and limited. This is standard practice across research journals, so the authors did nothing wrong, nothing different than any other authors, and they did NOT disappear anything.
But this complaint illustrates Somerby's ignorance about research reports. He does not know how to read such a report if he thinks this is something to make a big deal about.
A result that shows no significant difference between the groups being compared (white vs black target, high vs low endorsement of false statements) is called a null finding. Null findings are not publishable. They mean that there is no effect, nothing special going, within that group of subjects (nonwhite). It means the nonwhite subjects did not undertreat the black target for pain, the way the white subjects did. And THAT result was reported, not disappeared, the way Somerby has repeatedly claimed.
Why would it be surprising to find that nonwhite medical students did not show the same bias as white ones?
This is not just a failure to read a scientific report -- Somerby has been harping on this for weeks now and he seems motivated to say incorrect things about this study. Even now, when he surely must realize he is wrong, he continues to write another incorrect essay. Surely someone must have pointed out to him that he is making yet another mistake! I think this is lying as surely as Trump lied when he continued to claim his 2020 election was stolen. Somerby is lying about this research report, he has had every opportunity to correct his mistakes. He partially backtracked yesterday, but today he is once again telling a huge lie about this paper.
This needs to stop. It is wrong. It is unfair to the paper's authors. It is unfair to the black people who have historically been undertreated for their pain. It is unfair to the journalists involved. It shows how easily a blog can repeat and repeat wrong information without consequence, if the author doesn't read his comments. In fact, why does Somerby have comments at all? Why doesn't he just turn them off, as numerous other blogs have done? If he isn't going to read them, they serve no purpose except to give a home to beknighted trolls.
Somerby has officially jumped that shark. Whatever his health issues, they don't excuse the racism inherent in Somerby's recent actions concerning this research report. It is time to hang up his pen and become just another fool, instead of a noisy fool spreading right wing and racist garbage to the unsuspecting.
"the paper itself clearly states, there was no significant difference on the pain rating and treatment tasks, unlike the white subjects"
DeleteDumbshit: at discussion is the beliefs.
You triggered, 10:23?
DeleteSome nonwhite subjects expressed false beliefs too, just as in the white group. Unlike the white group, those nonwhite subjects endorsing false beliefs showed no tendency to undertreat black targets for pain. They treated the white and black targets alike.
DeleteTomorrow Somerby is going to expose Kevin Drum's amazing analysis with a flourish. The endorsement of false beliefs was the same for white and nonwhite subjects! Drum will say that each subject in the false belief group endorsement only a few false beliefs. He will say that there was no huge tendency to endorse any particular false belief except the skin thickness one. That is what he said before at his blog.
The mistake Somerby and Drum both make is not understanding that the false beliefs were a composite variable -- which means the 11 questions were combined into a single measurement and that magnified the small differences on each individual question. This is a common practice in psychology. It makes slight differences more measureable. Somerby and Drum are apparently not familiar with this practice, because they insist that only the individual questions matter -- that is not how composite ratings work.
So, they are going to make fools of themselves again tomorrow. They are going to repeat specious criticisms and show their ignorance again. And they are going to waste everyone's time again, while maligning researchers who did nothing wrong.
This comment has been removed by the author.
DeleteHow did whites and non-whites compare on the true statements? The same degree of belief, or significantly different?
DeleteSo white students who got one of the questions "wrong" tended to underestimate (hypothetical) black patients' pain and undersubscribe them pain medication by some degree. On the other hand, black students who gave more "wrong" answers than the white students estimated the pain of and recommended pain medication equally for whites and blacks.
DeleteDoesn't this evidence that there is no correlation between giving a wrong answer and how the hypothetical white and black patients were treated by the wrong answer givers?
Sheesh. What kind of a user name is that?
DeleteAnyway, the researchers said they wanted to focus on the responses of the white med students, which does show the correlation, and that the nonwhite students did not underprescribe pain treatment, as you stated.
It means that the nonwhites did not give lower pain treatment to blacks no matter how they answered the questions. The whites who answered more questions wrong tended to give lower pain treatment to the black target but not the white target patient.
DeleteThe more false statements a white subject endorsed (chose as being true), the more they undertreated the black target for pain. Wrong answers were correlated with and predicted undertreatment of pain for black but not white targets. The true statements were excluded from the composite measure of false belief. Marking a true statement as true is not any kind of false belief. Marking a true statement as false shows a lack of information, not false belief. Only the 11 false statements were included in the false belief measure. You might mix true statements in with false ones in order to prevent subjects from guessing that all of the false statements are false and just going down the list and marking them all false without even reading or thinking about them. Similarly, you might use names to identify the black targets instead of showing pictures, in order to keep subjects from recognizing that the study must be about race, since a black and white patient were being assessed. Someone who recognizes a black name without a photo would perhaps have different kind of racial awareness than someone who didn't pay much attention to the names or didn't recognize the black name as being more common in the black community. The name "primes" race for some subjects and not others. It does this more in the group who endorsed more false questions than in the group who endorsed none of the false questions.
It seems highly likely to me that those subjects with racial bias would believe more stereotypes about black people and thus endorse more false beliefs. But it can also be true, in this study, that someone without racial bias might hold more false beliefs via participation in a culture with odd stereotypes about black people. That's why this isn't just a study of racial bias in med students. The authors are not presupposing that bias is causal in pain undertreatment. Somerby should be happy about that, but unaccountably, he is ignoring it and just assuming the authors consider med students to be bigots (as he has said repeatedly).
So Sabin is “shocked.”
ReplyDeleteIt does seem rather surprising that first year (and beyond) medical students should believe false things (some of them downright weird) about blacks, given that they attended four years of pre-med, where they would have taken biology, physiology, etc. They aren’t exactly naive babes in the woods by the time they reach med school.
Perhaps “surprising” is a valid description of this? Although Sabin has every right to be shocked, with some justification perhaps.
But ultimately, her reaction is irrelevant to the study itself.
That blacks are underprescribed pain medication for acute pain has been validated in numerous studies, most recently by the one helpfully cited by AC/MA from the Boston Globe.
ReplyDeleteIt seems reasonable to try to find out why. It makes sense to attempt to relate this outcome to mistaken beliefs about black physiology.
A possible critique of the study is: what if you were to ask a completely different set of questions, about political ideology, for example, or something random, like which sports teams you root for. The hypothesis that the under prescribing of pain medication for blacks necessarily relates to false beliefs about black physiology might be called into question if a correlation to another set of questions could be shown.
It would not debunk the study, but would suggest further avenues of investigation.
That is all you really need to say as a critique of the study. Instead, Somerby offers his obsessed outrage that frankly makes him seem unhinged about this, going on and on for something like 20 posts now about how the study (falsely) is nothing but a scam that intends to impugn white med students as racist.
I just suggested a way to call the study into question, and this is your response?
Delete
DeleteAd hominem? Meh. Rather, what we observe here is argumentum ab auctoritate: "numerous studies". Appeal to authority.
...pretty much the same shit, though... the other side of the same coin...
The issue is the talking-point about (white) medical students which has emerged in the mainstream press as a result of the study.
DeleteThis comment has been removed by the author.
Delete11:12: Somerby is criticizing the media response, but he is also trying to critique the study itself. Everyone here is aware of that. My comment relates only to his critique of the study, which I find unconvincing. I suggested a critique of the study that is more fruitful. And I still get attacked. Oh well.
DeleteIt isn't clear why endorsement of higher than usual false beliefs would lead to undertreatment. It might be because of the beliefs themselves, but it could also be because of a general lack of familiarity with black people. It could be because endorsing medical false beliefs occurs in people who live where other kinds of false beliefs (racial stereotypes) are commonly held. It could be that those who hold false beliefs also hold negative reactions to black people and other forms of bias against them. We only know that the false beliefs are associated with undertreatmet in white subjects. We do not know why. That the nonwhite subjects who held false beliefs did not show similar undertreatment bias, suggests it is not the beliefs themselves but rather that the beliefs are a correlate of some other factor that leads to undertreatment by white subjects. That is the possibility that Somerby seems to be afraid of, because he is working very hard to absolve white med students of accusations of racial bias.
DeleteIf a different set of questions were to result in a similar undertreatment bias, as mh suggests, I would tend to think that the problem is the white students reacting to a black target, not the questions themselves which might identify simply those white subjects who are likely to be racially biased. The researchers could have administered some other test, such as a racial IAT or screened for geographic area lived in during childhood, to see whether exposure to racial myths might have resulted in endorsement of more false beliefs and also produced biased pain treatment for black targets.
When someone critiques a study, as Somerby and Drum have done, but can only come up with mistaken critiques, the desire to criticize is obvious and it appears they are fishing to find something, anything, wrong with the study. Somerby and Drum have no training in psychology, medicine, research methodology, hypothesis testing statistics, or pain experience, so they make a bunch of mistakes in their attempts to find something wrong. As mh points out, there are other critiques that could be made by someone more knowledgeable. But the problem isn't that the study is wrong. It is that it doesn't actually nail down the reasons why the biased treatment judgements were made, and it doesn't sufficiently demonstrate causality of false information. It is preliminary and suggestive, not conclusive. Such studies are published because they provide direction for future research and are interesting to other researchers. No one in experimental psychology believes that any one study is sufficient to show any major conclusions, such as that white med students are racist (the one Somerby seems most afraid of). Science is incremental. But if Somerby is going to knock down this study, he can only do it by explaining why THIS study got the results it obviously got. He cannot do it by throwing out the data or challenging the methods, unless he repeats the study modifying the methods (perhaps as mh has suggested) and showing that the same results obtain with other questions, or that the modification to improve the methods or remove a flaw produces a different finding.
Somerby isn't in the business of doing research. He is in the business of whining about female journalists and deny that racism is a thing (with no evidence whatsoever). That makes him a propagandist, not a scientist. And Drum is in the same category. He fastens his attention on irrelevant data that has no bearing on the findings, and can neither overturn them nor explain them. That is why it is so easy to see that he has no idea what he is talking about.
In my opinion, the most cogent comment here was from the anonymous person who questioned the graph shown in Figure 2 and asked why the lack of effect in Panel A for pain ratings of the black target was not problematic. I agree that it needs more exploration to see whether (1) subjects could not interpret black pain appropriately, perhaps due to lack of familiarity with black people, (2) there was a cue in the white target that suggested more pain than the black target, (3) there was an artifact of presentation order, failure to counterbalance stimuli, or (4) subjects were aware of the intent of the study and didn't know how to comply with a demand effect when interpreting black pain, perhaps again due to unfamiliarity with black patients. I would want to chase that down with further investigation. Again, it does not invalidate the study -- it is a loose end.
DeletePanel B clearly shows the effect that the researchers were claiming and invalidates whatever Somerby and Drum will say tomorrow -- if they cannot explain why that effect is there (it is not by chance), then their other gripes are empty.
Interesting.
DeleteNote that this comment, to which I responded with my comment at 10:49am is now missing:
“AnonymousJanuary 11, 2023 at 10:47 AM
strawman/ad hominem/tone argument
You're an imbecile.”
It was an anonymous comment, the “anonymous” was black, not green.
Normally, this kind of comment cannot be deleted by anyone…other than the blogger himself.
This has happened to me before, but I speculated it might be because I included a bad word that might be automatically screened by Blogspot. Once I wrote a comment immediately following Somerby's post and he took his entire post down. His was about being depressed when seeing elderly people. I said it wasn't my intention to depress anyone when sitting around in public, and that being old wasn't depressing to me as an old person. I wondered about that. After yesterday's capitulation, I think Somerby does read some of his comments.
DeleteWhat was the astrological sign of those who believed false statements?
ReplyDeleteToo lazy and disinterested to read much of this commentary let alone the study. But curious about one thing. How is a "black" person defined in this study or studies like it?
ReplyDeleteWill they stop doing these sorts of studies once it becomes difficult or impossible to determine what race a person is? Or perhaps have data directly relating darkness of skin with other parameters? What a waste of time... divisive garbage science IMO.
In a medical setting, the patient in front of the physician is clearly visible and audible. If someone is inclined to discriminate based on skin color or accent or class status (assessed by clothing), they will have all the info they need to apply a stereotype.
DeleteIt might be fine to dismiss this as "garbage science" if it were not documented by many studies that there is undertreatment of minorities for pain.
You would care about that, if you were the one in the ER with a broken leg, being given less than the standard of care for your injury.
anon 12:33- probably the student subjects of the study gave their "race" in a questionnaire. The race of the "patients" whose pain was estimated by the students, who then prescribed medical treatment - the patients were hypothetical patients, They were given first names in the study. Black patients got black names lie "Keyshawn"; whites were given "white" names like "Brett." That was the way that the students could tell who was black and who was white.
DeleteHmmm…you mean the mere name of the patient was enough to trigger an assumption, and then a bias in pain treatment? The study just got more interesting.
DeleteYes we already know racism exists. Everywhere, not just the medical industry. There are specific efforts and mechanisms in place to monitor it and curb it that exist in the medical industry.. Does this study help in that effort? Maybe it does. I'll let you explain that though, I have had it with this stuff.
DeleteThe impact I see is that this stirs things up further, gets possibly misrepresented in the media, and doesn't help us remedy anything. But maybe I'm wrong, I always reserve the right to be wrong.
I find the media critique more useful than attacking the study itself, which may be helpful in dispelling certain false beliefs about black patients and ultimately improves treatment. That’s why it’s important (to me anyway) to keep the study and the media response separate.
Delete12:33,
DeleteThey'll stop when the police can't figure out who is black. Win-win.
@12:33,
DeleteAs long as you are white, you have the luxury of ignoring such studies as "garbage science." The minute something affects you personally, you will feel differently about such studies. This is what the woke call "privilege". White patients get discriminated against in medical settings too, such as when they are fat (obese), when they are poor and have no insurance, when female, when a teenager, when undereducated or an immigrant, when the disorder is related to a socially stigmatized behavior such as recreational drug use or deviant sex or being transgender. I wouldn't be too complacent about deciding this doesn't affect you personally.
This is the type of comment that drives me away yet again. Filled with assumptions. Those assumptions are all 100% incorrect. See you in the next life.
DeleteBye, 1:38! Don’t let the door hit you on the way out.
DeleteThe people still arguing about the study with Bob have my admiration. They recall the just tenacity with which Bob once corrected the Press’s constant falsehoods about Clinton/Gore in the days before somebody bought him off.
ReplyDeleteFrom placidodomingo.com:
ReplyDeleteThe Democratic Party presents itself as something enormously positive, indisputable and inaccessible. It says nothing more than “that which appears is good, that which is good appears. The attitude which it demands in principle is passive acceptance which in fact it already obtained by its manner of appearing without reply, by its monopoly of appearance.
Is authoritarian fascism wrong?
DeleteOpinions differ...