Every now and then I get an email about R. A. Fisher, someone sending me an article about his scientific contributions, how he reconciled his patriotism with his sympathy for various Nazi ideas, whether he should be honored or canceled or whatever. I posted something last year but no more since then, as I don’t really see there’s much more to be said on the subject.
But I did happen today to be thinking about the general topic that each person is some mixture of good and bad. In a recent thread on bad professors and how they persist in their badness, I contributed this comment:
Feedback is really tough. Years ago I had an operation, and in recovery I was treated by an incompetent nurse. I mean, he was really a weird guy. I told him I had to go to the bathroom but I wasn’t sure if it was safe. He acted like he didn’t want to get too close to me and he kind of stood a few feet away from me, not supporting me as I walked over, and then I collapsed. I almost bled to death! (Not right away; it was a slow leak.) The next day I told the other nurses about what happened and they said, ummm, who was that, was it Tom? I said, yeah, that was the name, and they were like, yeah, he’s a really weird guy. Later, I thought about writing a letter of complaint to the hospital but I didn’t bother.
From the other direction, sometimes I do a bad job teaching a course, and I’m sure that lots of the students want to complain to someone, and I bet some of them do. Actually, my first year as a professor some students complained to the department chair and he came and sat in one of my classes and gave me advice. That was super helpful. Too late for the students in that semester, unfortunately, but I’m pretty sure that this advice made a positive difference going forward. And, yeah, I knew I was teaching badly, I just had no idea of how to right the ship.
Anyway, here’s the point. The department chair who many years ago did me that favor also did two things that I don’t like: he helped with the O. J. Simpson defense team, and he was a sexual harasser. Also he didn’t seem to have much understanding of statistics as applied to social science, but that didn’t stop him from considering himself an expert on the topic. I guess that was a third thing he did that I didn’t like. His problem was that didn’t know what he didn’t know, as the saying goes. But . . . he cared enough about doing the right thing that he listened to the (very reasonable) complaints of some undergraduates and then went to the trouble to sit in on the class of a new professor and give helpful feedback. He didn’t have to do either of these things, but he did. That’s something special he did that I do like.
People are complicated, with a mix of good and bad. We’re not always aware of this because we don’t always see the full story. Some people we only hear good things about, other people we only hear about the bad things. For example, I’m sure that the various practitioners of scientific misconduct we discuss in these pages have done lots of good things too; it’s just that those good things aren’t part of the public record. R. A. Fisher’s life has been well enough documented that we hear about the good and the bad things. I knew the above-mentioned University of California professor for a few years so I got exposed to his good and bad sides. It’s useful to keep such examples in mind when considering people whose lives are less well documented or whom we don’t know personally.
Andrew refers back to his posting in August 1, 2020 regarding R.A. Fisher
https://statmodeling.stat.columbia.edu/2020/08/01/ra-fisher-and-the-science-of-hatred/
In that particular blog, I wrote the following
——————————————————————————
paul alper on August 2, 2020 6:13 PM at 6:13 pm said:
Although Andrew wrote a long exposition and at this moment there are 75 responses, I have yet to see a reference to the magnificent book, “Statistics in Britain–1865-1930; The Social Construction of Scientific Knowledge” by Donald A. Mackenzie. He points out which of the early founders were eugenicists and why/when eugenics no longer became the prime mover for statistical research. If you have not read the book, you should. I interviewed him in the 1960s and he indicated that he had nothing additional to say on this topic and was going on to other topics:
https://www.worldcat.org/identities/lccn-n82008489/
————————————————————————————————–
Eventually, there were 155 responses in all but only one, by Steve Sailor, referred to Mackenzie. Once more, I recommend Mackenzie’s book.
Paul, I may buy it. I’ve long wished there was a book called something like The Death of Eugenics specifically about when and how people gave it up. As far as I know, not one of the leading figures, or for example anyone from the Bloomsbury set or the birth control movement, ever published a renunciation. Apparently they just dropped it after WWII without addressing their mistake (a theme of this blog).
Did R. A. Fisher really give it up?
hmmm, $125 in hardcover only, plus shipping.
Why didn’t the department chair make sure that all new professors get help to improve their teaching?
David:
I dunno, everyone’s busy I guess, and teaching is not a high priority at research universities.
Even if the department doesn’t care about the undergraduates, I would hope they would try to make life easier for their new hires so that they could retain them.
In my experience, the excuse people give for incompetence is often that they are busy. However, competent people know how to manage their responsibilities so that it appears that they are giving full attention to each item even though they are really doing multiple things.
When I was at MIT, the Everett Moore Baker Memorial Award for Excellence in Undergraduate Teaching was given to a math professor (it is given to just one faculty member at MIT each year). A few years later, he didn’t get tenure. The outcry from the student newspaper, student government, etc. was quite loud. The students were saying (essentially), “We don’t care if his research is no good. He’s an excellent teacher. And, you’ve been saying undergraduate education is important, and we pay a lot of money for it.” This put the math department in the amusing position of saying, “No, no, no. His research is good. Just not, um, quite good enough.” The story has a happy ending because he ended up at another fine university where he has had an excellent career, both in research and in education.
“he helped with the O. J. Simpson defense team”
This is complicated too isn’t it?
Part one of the “complicated” is that simpson was never convicted of murder, so what’s the problem with defending him? I don’t know the details of him being found liable for the deaths of his ex and her bo in civil court but I can only imagine what a shitshow that was – the criminal trial was bad enough. From out here in the bleachers, our civil court system seems pretty wacky. It seems ridiculous that he could be held liable in civil court for a crime he wasn’t convicted of in criminal court. I personally have no idea if he was guilty or not.
Part two of the “complicated” is that, because people who are innocent are often charged with crimes, it’s critical that defendants have access to a strong defense – whether we think they’re guilty or not. If no one is willing to work for people who “seem” guilty, there will be no defense for people who aren’t guilty. I’m glad your dept chair took the job on Simpson’s defense team. I congratulate him. I’m glad there are people willing to defend people who “seem” guilty. Maybe someday I’ll be charged with a crime I didn’t commit, but everyone will think I’m guilty. I’ll need someone who’s willing to step up and take the heat.
Overall, the modern habit of tearing everything apart to label what’s “bad” about it is one of the biggest “bads” of society today. Jefferson and the other founders of the US had some great ideas and they did a lot for Americans and for humanity – the US constitution is more or less the world-wide template for freedom today. The fact that some of the founders may have held slaves 250 years ago when slave holding was common around the world is immaterial today. *I DON’T CARE*
And that’s all before we get to the fact that – as our feelings about his former Dept Chair participating in OJs defense show – one man’s bad is often another man’s good. “Good” and “bad” aren’t intrinsic properties. They’re judgements.
Anon:
I get it that you don’t care—even to the extent of you minimizing things by saying that some of the founders of the U.S. “may have” held slaves, as if this is somehow in doubt.
Regarding O.J.: (1) Yeah, it’s just my take. If you want to join Michael Moore and say you think O.J. was innocent, go for it. (2) I understand that everyone, including an accused murderer, has the right to legal representation. I don’t see any principle that says that an accused murderer has the right to the expertise of world-class statisticians. (3) I get it that different people have different values. I’ve accepted $ from the U.S. Department of Defense and various mega-corps (just see the list of sponsors on this blog), and I respect that for many people, that’s a lot worse than working to keep O.J. on the street.
I have no idea whether or not OJ was innocent or guilty of murder. But you didn’t answer my concern that, if we pick and choose who gets what defense based on whether or not we think they’re guilty, then there won’t be adequate defense for innocent people unjustly charged with crimes. You can hardly say that wrongful conviction isn’t not a problem.
As you said, it’s complex
1. I agree that it’s complex.
2. Providing O.J. with a public defender, or him paying a private defender: I agree, that’s a central part of our justice system and I think it would be wrong for him to not have a defense at all. That’s different from a statistics professor taking time out of his busy day to contribute to the defense team.
The standard in a civil trial is different from the standard in a criminal trial: preponderance of the evidence vs. beyond a reasonable doubt. So, you can certainly be acquitted in a criminal trial and lose a civil trial. This being a statistics blog, the former criterion seems more relevant. As for the criminal trial, I believe the prosecution did a terrible job with the jury selection, so the verdict was a forgone conclusion.
“It seems ridiculous that he could be held liable in civil court for a crime he wasn’t convicted of in criminal court.”
a) in criminal court the jury must convict “beyond a reasonable doubt” which is typically considered 95% or greater certainty. In civil court the standard is “based on the preponderance of the evidence” which is typically considered 51% certainty or greater.
b) Some of the criminal trial jurists were interviewed after the trial and revealed that they did not understand, for example, that the recordings of Simpson’s wife calling 911 to report assaults and threats by Simpson had any bearing on the case, and ignored that evidence. The civil court jury might have felt differently.
c) the criminal trial prosecution lawyers did not do a good job, as for example when they asked Simpson to put on the gloves. Perhaps the civil court plaintiff lawyers did a much better job.
The fact that Simpson badly flunked a lie-detector test (assessed to have lied on 23 of the 24 questions) conducted by the defense, made an attempt to flee before the trial, and had threatened his wife previously might have been enough for me to vote guilty in a civil case, lacking any other suspects, although I can’t be certain of that without hearing the defense, and the lie-detector information only came out some time later. Anyway, it seems far from ridiculous to me.
I witnessed Simpson losing his temper on the football field and being ejected from the game for throwing blows, as a result of disrespectful but not harmful treatment by the opposition, during his playing career–against the Giants during his 2000-yard season, if I remember correctly.
Lie detectors (aka polygraphs) are a scam. Even the government has admitted that they work because they tell people they work, so people then confess. I actually saw this happen to a colleague. It was only a minor thing involving handling of possibly-classified information that I didn’t think violated the rules. But, during a routine polygraph, he admitted it. So, it got written up.
However, there was plenty of other evidence that Simpson was guilty.
Lie detectors are far from perfect, and most of their effect is probably psychological (I debate with myself whether perfect lie detectors would be good or bad for society, dictators would love them), but when the defense hires a lie-detector technician they have worked with before, and he tells them, “Your guy is lying his ass off, he did it,” that still counts as something with me. Not proof positive, but another weight on the guilty side of the scale.
> but when the defense hires a lie-detector technician they have worked
> with before, and he tells them, “Your guy is lying his ass off, he did
> it,” that still counts as something with me.
How is that different from the defense hiring a psychic that they have worked with before? Or, a dog handler who tells us what the dog is thinking?
“Lie detectors (aka polygraphs) are a scam.”
It’s amazing that the results from lie detectors are admissible in court. My understanding is that they have failed numerous scientific tests and that some people become very afraid and may actually admit to crimes they *did not* commit.
As I understand it, the case was decided by;
1. Terrible jury selection by the prosecution
2. Excellent character assassination of the cop in charge of the investigation by the defense. They played a recording of the cop using the n-word, talking about sexually harassing female officers until they quit, and fabricating evidence especially against black suspects. Then they asked him if he fabricated evidence in this case and he pled the fifth. So, a pretty good job sowing reasonable doubt.
Now, OJ Simpson was probably guilty, but here’s a complicating moral factor. Does that dirty laundry deserve to be aired? Shouldn’t it have been considered by the jury? So did the defense team do wrong by doing their job well, finding it, and airing it?
never mind, upon further review it looks like the fabrication of evidence never made it to the jury
One advantage stats has over the humanities is I can say, “The guys were racist assholes but they came up with useful tools, so let’s learn them,” and no one has a problem with that. You don’t get any of the controversy you’d find with (say) a D.W. Griffith film.
People may have boycotted Shockley, but nobody boycotts transistors.
For me the main point of article is that great contributions can be made by people who have some erroneous ideas, and by implication we should be alert for the mistakes that we all have in our own thinking. Many times these mistakes arise from accepting common ideas without bothering to examine them since the ideas affect some distant social group. Also, sometimes people want to conform to their society’s norm and not stick out. And sometimes there is a John Nash with real contributions but a serious disorder. Conformity pressure has been my Achilles’ heel.
Oncodoc:
Well put. If Fisher, Turing, etc., can believe foolish things, then so can we!
As the old saying goes: it ain’t what you know you don’t know that gets you in trouble, it’s what you know for certain that just ain’t so.
And sometimes people can be very nasty if they perceive it in their interest to be.
In my career, someone widely thought to be a nice guy actually threatened my family’s well being if I did not adopt their methods. I played along until I no longer needed to but I have only told a few close colleagues many year later. No one else would have a clue.
It’s why I think it’s foolish to name things after people – who ever is chosen will have likely have a few things they hope no one else will ever find out.
I have never understood why nurses act so badly sometimes. I have had many bad experiences since I started dialysis in 2011 in Germany and ended up many times in hospital, sometimes for months. I think the worst ones were when a nurse refused to put on gloves when connecting an antibiotic drip into my arm, and said, smiling at me, that my choice was to either not get the drip or to let her do whatever she wants to; and when a nurse in the hospital dialysis station chided me for being too theatrical when I cried out in pain when he placed a dialysis needle in completely the wrong place (he missed the artery entirely—and shunts are almost impossible to miss because the vessels are enlarged through an operative procedure), and the other nurses were laughing at me for being so dramatic. I made him take out the needle and placed a new needle in the correct location myself (I usually put my own needles in because of this kind of BS, but I had just had an operation and was woozy). I was just blown away by their callousness. There were so many other horrible incidents with nurses over the last 11 years of dialysis, too many to recount.
OTOH I suspect that most of my PhD students probably see me as a not-nice person because I spend a lot of time pointing out things that are wrong, and don’t spend much time on what’s right. It’s just too much negative feedback. I always felt that I’d rather know where I went wrong and others would want that too, but it seems that is completely the wrong thing to do. (I really hate the toxic positivity of American-style discourse though; that’s the other extreme, and it brings nothing to the table and just makes me want to throw up.) I was told I have a reputation among students for being overly tough on them so students are probably terrified of working with me :). For my part, I always thought I am helping them do the best that one can do.
So just as I can’t get what I want from my nurses (thoughtfulness and empathy), my students can’t get what they want from me (positive vibes). Maybe the nurses need to be the way they are to just get through their day and I can’t see their point of view, just as my students can’t see mine.
Many people don’t realize that things can be objectively correct or not correct. So, if you point out their errors, they wonder why you are being mean to them. Don’t you like them? (There was a good Dilbert on this topic, but since Scott Adams has turned out to be one of those people who are mostly bad, I won’t link to it.)
“but since Scott Adams has turned out to be one of those people who are mostly bad, I won’t link to it.”
Ha, that’s hilarious. Shoot the messenger!
Anonymous –
> Shoot the messenger!
A beautifully ironic symmetry.
You equate David not linking to Adams’ cartoon (presumably because he thinks Adams is a malignant idiot) with “shooting the messenger,” in response to a comment on how people equate criticism with “being mean.”
That would be like people with immensely powerful public platforms – where they say whatever the want to millions of people – complaining about “censorship.”
Oh.
Wait.
“Many people don’t realize that things can be objectively correct or not correct. ”
Excellent point. Some people just don’t get that.
But still, quoting Andrew’s famous words, “it’s complicated”, cuz many things that have been touted as “objectively correct” turn out to be incorrect and even badly wrong, like the “slam dunk” of WMD or the once irrefutable Population Bomb.
Last but not least, humans use many different signals to detect lies. Apparently for some people the claim that something is irrefutable or “objectively correct” plays a heavy role in generating suspicion.
There’s a good idea for some reaserach! Surely this has already been done. Would be interesting to hear about: what cues do people use to detect lies, and how often are the different groups of cues “objectively correct”? Is it possible – I assume it is – to play on those cues with reverse psychology? Fun stuff!!!
I have a really hard time associating anything with the term “toxic positivity.” Or rather, I can do it but I have to reach into the realm of satire and fiction to get there. There’s a non-toxic role for positive reinforcement.
Barbara Ehrenreich wrote a whole book on it.
I guess you are referring to Bright-sided: How Positive Thinking Is Undermining America. Looks interesting; will read.
As the thread title says, “people are complicated”, but some complications can be understood. Not arguing from scientific studies (even though I believe they could probably be found) but rather from personal experience, people are often in need of appreciation and positive feedback in order to fell that they’re in the right place doing the right thing. PhD students, who are only criticised and don’t get positive feedback, will often, instead of focusing on the criticised details only, question whether they should be doing a PhD and may have issues with their self-esteem in general to which such behaviour may contribute. Also, if students feel generally appreciated, they may find it easier to accept criticism and to in fact learn from it, rather than taking it just as further proof that either their supervisor is evil or that their work is generally worthless.
Furthermore, criticism can be conveyed in various ways, and in some cases criticism doesn’t only convey what should be improved, but also the (imagined) superiority of the supervisor and more general disdain for the student than just regarding the issue that is criticised. This is of course subjective, the critic may be misinterpreted, and it is sometimes hard to understand for the critic themselves how the criticism comes across.
Personally I criticise a lot but try to be aware of these things, and I make an effort to be positive even when criticising (like showing some appreciation for why the student may have thought XXX is a good idea, or saying how widespread a mistake YYY is). Also, sometimes I ask students of my classes for feedback using an anonymous self-designed questionnaire that has only two questions: “What do you like about the course?” and “What do you think could be improved about the course?” This may obviously seem biased in favour of positivity (I don’t ask “What don’t you like”), but I know that if I ask in this way, reading what the students write will get me in a more positive mood which also makes me appreciate better what is written about how to improve. In principle chances are we all think that constructive criticism is a good thing and we should learn from it, but on a more basic psychological level I think most people (myself included) have difficulties with being criticised and need to get around their initial negative feeling in order to make something good of it. For sure it can’t be wrong to help them with this.
Agree to the points you raise. I do the same things you do, I try to start and end with what I think is good about the draft (it’s usually a draft being prepared for a journal submission). My postdocs tell me I don’t do this enough, though. I have to find a balance between nonsensical positivity and providing some kind of validation of the inherent value of the work done. I do that more in reviews that I write; maybe it’s that phenomenon that you feel you can be more frank with your in-group than with an outsider.
One other trick I started using recently is to make a video recording of my comments; this way, intonation is not lost and the reader looking at my comments in the pdf (hand-written on an ipad) don’t sound as harsh. People tend to impute harshness to the written word. This method really works well; it gets all the criticism across but it sounds much milder.
But I always saw my job as advisor an senior co-author as doing an internal peer review; and I feel that I also need to prepare students for the coming storm. I was talking about toxic positively, an American phenomenon, but there is also a toxic negativity I see in American conferences that generally is missing in European ones. People stand up and aggressively attack the speaker, sometimes even fresh graduate students giving their first talk in front of the entire field, 300+ people. It is usually Americans (one person from Harvard is particularly bad) who love to do humiliating public takedowns, sometimes even shouting down other commentators mid-comment. But it’s not just one person; lots of Americans go to conferences like it’s going to war; the goal is to win in the public domain. A new graduate student of mine once gave a talk at a conference in Stanford (I wasn’t there) and what I heard from my own postdocs who were present was that one group made a coordinated attack on him in the question period, with the sole aim of discrediting the whole line of work (which went against their own theoretical position). These attackers were not inherently mean people, but winning is a thing we are taught to aim for in American academic discourse. You can’t say, yeah, this could be a plausible theoretical position; it’s not what I believe, but it could have some merit.
I see those kinds of reviews a lot in journal submissios too. You did four years of hard work collecting this data, or you did four years of unbelievably hard work developing this single model, which no one else has the technical capability of doing? It has this minor imperfection and by the way why didn’t you implement my own paper-pencil model which I have been waffling on about for 30 years; reject, with no opportunity to rebut. In one of my earliest submissions to the journal Cognition, a senior scientists (now retired I think) shocked the beginner-me by calling me irresponsible. After 20 years of being told that the submitted paper is crap, I feel I need to prepare the student for what’s coming by pushing on every detail that is a potential weakness. I would actually have preferred this kind of feedback at the internal review stage, but as a student I never got it. But not many appreciate this kind of a detailed listing of problems.
I have seen the kinds of thing you’re talking about in Germany and the UK as well, I have seen this kind of reviews from lots of countries, and I’ve also seen a good amount of pointless positivity in various countries. I wouldn’t identify this stuff with America and the Americans.
My online experiences with Americans seem to suggest that they are a relentlessly negative people these days.
Christian wrote:
“I have seen the kinds of thing you’re talking about in Germany and the UK as well, I have seen this kind of reviews from lots of countries, and I’ve also seen a good amount of pointless positivity in various countries. I wouldn’t identify this stuff with America and the Americans.”
Yes of course, I didn’t mean that 0% of non-Americans show some particular kind of behavior. It’s just a tendency I observed after moving from India to Japan and then to the US, the transition from an academic setting in Japan (Osaka University) to the US (Ohio State) was a shock. E.g., I’d be walking down the street in Columbus, and some stranger would call out to me, Hey, how’s your day going? I’d say fine, and ask them how theirs was going. “Great, and it’s going to get better.” That kind of stuff happened a lot (maybe it was just Midwest madness). And over the last 20 years I attended US and European conferences, the tension was palpably higher in the US conferences. But I do agree that there are all kinds of people everywhere. There may also be field specific differences (I mean, look at Fisher, and we do say a quarrel of statisticians for a reason I guess). Linguistics conferences also had that aggressive quality (although of course not always).
I bet someone has done research on this.
Christian and Shravan:
Appreciate both of your efforts to criticize effectively, and at the same time not be quacky positive! From my experience, over-the-top positivity feels good for some people, but for those who can make their own objective assessment, it comes off as insincere and contrived. I like Shravan’s idea of giving a video for review – it lets the reviewer see how their coming across. That was an important component of my public speaking course: I had to review my presentations on video and write my own critique. Shravan, I wonder if your experience with Americans is just a result of the particular people at a particular school in your discipline. I’ve seen many people from other countries behave that way as well.
OTOH, it’s also important for people to not be intimidated by attacks and to be able to handle it and think on their feet. IMO students giving their first few talks should drill the shit out of the talk and prepare for every angle of attack. There’s nothing like rigorous and extensive preparation to build *real* confidence. When you know what the attacks might be and have a plan to respond to them, you don’t feel intimidated. Not to mention the benefits of learning how to prepare that way.
Yes, that’s really good advice re solid preparation. Also, it is good to give some lab-internal practice on getting grilled conference-style, by aggressive and unpleasant people (and yes, not necessarily Americans :). Once I had a colleague who had to go to the US to do a job interview with a major tech company whose name begins with G, and we (two postdocs) trashed his practice talk and he was so angry with us he didn’t talk to us all day. Then he came back from his job talk and thanked us for preparing him for the nightmare that was his question period (he got the job). BTW, the unpleasant person in the audience was Indian-origin, i.e., one of my kind :). LOL.
OTOH I gave three or so talks in the US last year (Stanford, UMass, and Michigan) and it was all extremely friendly and pleasant. So yeah, it’s just a statistical tendency based on anecdotal evidence spanning one person and 20 years. Someone should study this properly (hard to do I guess).
Shravan:
Interesting story about your friend and the G-tech company. It’s great that you trashed his talk! From what I’ve seen, practice talks among students almost never provide the brutal feedback the speaker really needs.
You have to wonder about your friend: did his response to the troublesome questioner play a role in landing him the job? Audiences notice when a speaker handles difficult questions and/or attacks well – answering knowledgably and without getting flustered. It demonstrates strong preparation and a deep knowledge of the topic. Difficult questions are a double-edged sword: a problem if you’re not prepared, an opportunity if you are prepared.