This system too often rewards cronyism rather than hard work or creativity — and perpetuates the gross inequalities in representation …

This post is by Lizzie. I started this a while ago, but Andrew’s Doll House post pushed me to finally get it up on the blog.

The above quote comes from a recent article on the revelation that the person Philip Roth decided should write his authorized biography has a history of sexual harassment accusations (I mean, the irony…). It reads more fully, “this system too often rewards cronyism rather than hard work or creativity — and perpetuates the gross inequalities in representation that disfigure the American literary landscape”, but I think it certainly applies to lots of other `landscapes’ including the one within which I exist much of the time: ecology and evolutionary biology (or EEB).

It relates to a quote that has been rolling around in my head for many months now, from a student in my lab: “It feels like the message is that the careers of these [double-digit] people were worth less than the career of this one [purportedly] brilliant man.”

And I really did not have a good response to this student other than ‘ummm.’ I have asked around for a better response from my colleagues around me and I am not sure any of us have one.

It’s a relevant question, that could be posed for many publicized and less publicized events in EEB. Two recent publicized examples include the just-risen star of Jonathan Pruitt, a spider behavioral biologist, who has been accused of fabricating most of his high-profile data, passing out said data for junior folk to write up with him as senior author, with multiple papers now retracted or expressions of concern issued. On the rising-star front is Denon Start. He’s had recent accusations of suspicious data, but before those he had two awards abruptly (to me at least) rescinded from the Canadian Society for Ecology and Evolutionary Biology (CSEE). You can fall down the Twittersphere to try to figure out why, but CSEE never said.

There are similarities and differences in these cases galore, but the one similarity that has grabbed me is how much those around me, especially my faculty colleagues, seem to pin as much blame as possible on each man — I get the feeling 99.9% would be a good amount to many, leaving just enough room to acknowledge ‘we could all and should all do better.’ I agree a lot of blame falls on the perpetrators, but I also think that, by not leaving enough room to blame ourselves and the community we create, we open the gates for the behavior to continue.

Both of my examples had remarkable publication records (potentially falling under The Armstrong Principle), but they also were promoted by many people to get where they were (and are perhaps, Pruitt at least seems to still be employed by McMaster). Academia seems good at passing along and promoting people where we should perhaps have cause for concern. This to me means that either we are too out of it to know what is happening, or we look the other way because it seems easier, and it protects the careers of the rising star and all those connected to their star. There’s always some talk about the former — how can we build a better community where senior people know what is going on and can intervene? But what I want to talk about is: How do we hold a community of researchers accountable when they may have known there were concerns?

I suggest a step forward to is to hold the letter writers more accountable. Academia does function on reputation and a big part of your reputation is formalized in your letters of reference. American letters are often so positive as to feel almost useless. But they’re not. When they are short, brief or otherwise feel perfunctory, they can say a lot. Here’s a letter that might raise eyebrows:

Dear committee of special award or position:

I have known [X] in [this way] for [this long]. S/he/they have published [Y] papers, taught (or TA-ed) Z classes.

If you have more questions about this applicant, I encourage you to call me at the number below.

Sincerely,
Dr. Especially-eminent

So, that’s something we could start to do if you ask me. I also suggest the following after-the-fact potential actions:

(1) If you read a glowing letter for someone whom shortly after you hear had major concerns, go look at that letter. Did you miss something? If it seems like you didn’t, why not call up the letter writers and ask about the disconnect? Letter writers feel more okay writing these letters because there is generally no consequence for their careers.

(2) We could formalize some of this. Department chairs who receive glowing letters and there are issues later could be expected to contact the department who employs the letter writer and express some concern. We effectively write reference letters as part of our service, so it’s part of our job; if we’re doing that part of our job poorly shouldn’t it count against us somehow?

(3) The other thing I suggest we all do, after the fact, is be very cautious in how much we push back on whether the system needs to change by focusing on the perpetrator or fears that someone or some organization will be sued. I don’t think it feels like you’re saying you don’t want the system to change when you focus so strongly on the individual perpetrator or when you say, as many did to me, ‘what if [insert some society or senior person] is sued?’ These are both important things to do, but when they are most of what you do — then you just did trade in these action items of fear and blame for an actual closer look at the system.

Which brings me back to my student. Was my student reading the message correctly? ‘That the careers of these [double-digit] people were worth less* than the career of this one brilliant man’ in the particular case we were discussing.

It’s a good question to ask, if you ask me, as there are some implicit weighting and numerical assumptions here. Every time we don’t question the letter writers, or worry about what will happen to some established person or society, I think we effectively do send the message my student felt they had received. I even heard a colleague recently say we should not scrutinize too strongly the high fliers with the many, many publications, because ‘what if we discourage them? What if we lost [insert name of new NAS member] to that extra scrutiny?’ I have two replies to this. One is that if they are that great they will stand up to a little extra scrutiny.

And the other is to think more on how implicitly we undervalue those we lose when we say this. Ask people to explicitly count up those we lost along the way — who drop out, leave or otherwise are valued less, and how we value the creativity and exciting science we never got to see from them. We either think we aren’t losing many, or that they’re not worth the one great man we saved.

*In earlier version of this post I mistakenly wrote “worth more than the career of this one brilliant man,” which led to much understandable confusion.

56 thoughts on “This system too often rewards cronyism rather than hard work or creativity — and perpetuates the gross inequalities in representation …

  1. This concern about letter-writers seems out of place in the Pruitt case. How were they supposed to know that he was faking data? We know that Dan Bolnick was one Pruitt’s recommenders, and also that he played a big role in supporting the first people who found problems in Pruitt’s data, and was threatened by Pruitt’s lawyer because of it. And he’s been very open about the whole thing. Is there really that much room for major improvement here? I like the idea of focusing on systemic issues, but I don’t see the letter-writers as a productive avenue, at least if the Pruitt case is typical.

  2. I’m sorry, but I can’t make much sense of this post.

    The “letter that might raise eyebrows” is such an obvious “no” that I would conclude that whoever wrote it hated the candidate and didn’t have the decency to refuse to write a letter.

    “Letter writers feel more okay writing these letters because there is generally no consequence for their careers.” The “consequence” is that the people you recommend reflect on your judgement. I can’t see what other consequence you want. Is there some rubric for scoring the accuracy of a letter? Not everything can be reduced to a number or a true/false statement — certainly not everything of importance.

    “if we’re doing that part of our job poorly shouldn’t it count against us somehow?” Lizzie writes that she has a high opinion of Dr. Smith’s work, which is true. Three years later, Dr. Smith commits fraud. Should that count against you? Why? What if Dr. Jones says you should have had deeper insights into Smith’s character?

    Recommendation letters seem severely peripheral to the important topic of academic integrity.

    • “Letter writers feel more okay writing these letters because there is generally no consequence for their careers.” The “consequence” is that the people you recommend reflect on your judgement. I can’t see what other consequence you want.

      I think that’s the consequence being asked for, and the argument is that it’s not true in the status quo. I don’t really think recommendation letters really reflect on people’s judgement these days. Anecdotally, my peers give the impression that it’s become sort of a fire and forget arms race where the whole distribution of recommendations letters becomes progressively more exaggerated, and that professors will claim perfectly ordinary candidates are the greatest they’ve ever seen pretty much every year without issue.

      “if we’re doing that part of our job poorly shouldn’t it count against us somehow?” Lizzie writes that she has a high opinion of Dr. Smith’s work, which is true. Three years later, Dr. Smith commits fraud. Should that count against you? Why? What if Dr. Jones says you should have had deeper insights into Smith’s character?

      Well if the people you recommend reflect on your judgement, then yes, it should count against you?

      • About reflecting on the writer’s judgement: For graduate school letters of recommendation, I’ve gotten feedback that faculty found the letters I wrote useful, both during the admissions process and later; this is a statement (positively, here) about my judgement. I’ve written far fewer faculty job letters, but the principle is the same.

        About the arms race: yes, I agree with this. It’s like grade inflation…

        About “counting against you” in my example: as a reflection of judgement, yes I agree it should count. But in this paragraph, I think the implicit claim was that it should have some concrete consequence as “part of our job” about which one’s Department Chair is going to be contacted.

  3. 1. Totally agree letters should be more transparent. But I would turn this around on the committees or employers requesting letters: those folks should be more clear whether they want an evaluation of the applicant (warts and all) or a recommendation (emphasizing positives, abilities, achievements, potential). Maybe some ask for warts and all, but in my experience it’s uncommon.

    2. “How do we hold a community of researchers accountable when they may have known there were concerns?” This question is easy to ask in the abstract but gets much more pointed in the particulars wrt an actual person. Should we pillory Susan Reichert (Pruitt’s PhD advisor)? Should we accuse and blame Jay Stachowicz (postdoc advisor)? Are they really to blame for the fall-out from the Pruitt affair? Who would decide what constitutes their culpability for Pruitt’s falling upward? How could anyone know what Reichert or Stachowicz or others suspected but could not verify? Yes, this affair generated bad collateral damage for others who applied for the CPB postdoc at UC Davis, or interviewed for those faculty jobs at UC Santa Barbara and McMaster University. But trying to measure the culpability of Pruitt’s former mentors for that seems like a mistaken measure of the causes of that collateral damage.

    3. This blog has sometimes featured the idea of perverse incentives that might drive bad behavior among researchers. Trying to hold mentors to account for the unknown failings of their trainees would create a wicked incentive to hedge, speculate, and otherwise cover one’s ass in recommendations (or evaluations, whichever). “My student worked hard, had good ideas, and was a positive influence on others in my group. However, it is possible that she faked all her data and stole her ideas from her labmates. I don’t know that this happened, but I can’t rule it out.” A system like that does not seem like an improvement over the (admittedly flawed) system we have now. The system of mentoring and promoting the careers of trainees can surely be improved but not like this.

    • In some cases I think we are talking about known failings, not just unknown. That said, I wouldn’t be surprised if most knew nothing of what was perhaps really going on with Pruitt and his shooting-star career, but I don’t totally see the harm in it being a question you ask. The CPB postdoc went to both Pruitt and Start. Would it be the end of the world (or just an extreme transgression against our academic norms) to ask Jay Stachowicz and others in that group what is up? It seems that it might be better than never discussing it.

      • Yes I know both guys were CPB postdocs. No, asking “What’s up?” is not the end of the world or even a transgression. But we’re moving the goalposts here. The OP focused on a different question: “How do we hold a community of researchers accountable when they may have known there were concerns?” Accountability and sanctions for mentors seem like misplaced responses to misbehavior by their trainees. The OP assumes that Gilbert, Stachowicz, and others must have suspected or should have known something was “up”, and assumes that those mentors said nothing about what was “up” in their recommendations. That’s assuming too much imho. But maybe Jay will drop by the comments here with his two cents.

  4. Given the post’s examples are of research misconduct, I’m not sure that recommendation letters are the right focus. The people in the best position to prevent such work from getting into print are co-authors. I don’t know the examples cited here, but the summary of Pruitt sounds a lot like the data fabricators I’m most familiar with–Dederick Stapel and James Hunton.
    In those cases, and apparently in Pruitt’s, one person handled the data from original data generation all the way to a structured file for analysis.

    In Stapel and Hunton’s cases (both running human subjects experiments), their co-authors were well aware that one person was claiming to have solicited subjects, delivered the instruments to them, receive them after completed, and transcribed them into a spreadsheet. In fact, it appears that they didn’t do any of the tasks but transcribing fabricated data into a spreadsheet. If another co-author had been involved in any one of those steps, the fraud would have been quickly uncovered, or the purported data gatherer would have had to work much harder (e.g., actually fabricating written responses that didn’t look fabricated).

    This is no excuse for the fabricator–they shouldn’t have done it even if their co-authors made it easy. But co-authors also have a responsibility to implement the most basic controls, like separation of duties (as described above) and redundancy of duties (having two sets of eyes on every susceptible task). I understand how tempting it is when a well-regarded senior colleague says “here’s data I gathered–let’s write together.” But it’s also a recipe for getting conned by your own co-author!

    You can read more detailed advice in my free eBook, How to be a Good Professor–or Professional, in the chapter on Honesty.

    On the larger points of community responsibility, it’s worth noting that in both the Stapel and Hunton cases, many years passed before there was any basis for suspicion outside the co-author groups. Both worked closely with highly qualified people to develop very strong theory for their predictions, crafted excellent instruments, and other than the fact that their experiments always turned out so well, no reason to doubt that they were just very good at research. Stapel’s reputation collapsed because finally (!) a junior co-author risked her career to tell a Dean that Stapel (also her Associated Dean at the time, I believe) refused to share the handwritten subject responses with her. Hunton’s case collapsed when went a little outside his domain of expertise–instead of just gathering data from individual subjects, he claimed to have worked with offices of the Big 4 accounting firms. Unfortunately for him, someone pointed out that he reported working with more offices than actually existed. In both cases, once one paper was found to be fraudulent, their other work came under scrutiny.

    Until that happened, I don’t think anyone would have had a basis for writing anything but the most glowing recommendation letters for these highly-published, highly-cited authors of apparently highly-polished papers in the most high-prestige outlets.

  5. You would be interested in a very old paper on this topic. (I say very old but they don’t seem very old to me. :) )

    Revusky, S. (1977). Interference with progress by the scientific establishment: Examples from flavor aversion learning. In N. W. Milgram, L. Krames, & T. M. Alloway (Eds.), Food aversion learning (p. 53-71). Plenum Publishing Corporation.

    It’s about trying to make headway when your science is sound but it butts up against the establishment. He coined the phrase Type II incompetence similar to Type II stats errors. It’s been well established that new researchers tend to make Type I errors and more senior are biased to the Type II end. One my argue though, that the whole system is Type II incompetent.

    Now I’m reminded of the Deborah Rhodes imaging story (TED talk).

  6. Along with others above, I think the reference letter issue is misplaced – irrelevant at best, perhaps even counterproductive. These high-profile perpetrators are part of the superstar phenomenon. As the world becomes more complex, and our attention becomes scarcer, the value of superstars increases. This is a setup for enormous values for those that become superstars – and enormous mistakes in at least some of those cases. I do agree that we all bear some responsibility and there is much we should do to make reputations matter more and be based on real value. We could begin with more genuine assessments of research and teaching, preferably not placed in the hands of the select few (or in some cases, one administrator). That is why the reference letter issue is so misplaced. The reputation of the letter writer is amplified if we put even more weight on those letters. The reality is that a reference letter written by a nobody (reputationally speaking) counts for little, and one written by a superstar is given far too much credibility. It is part of the dysfunctional academic game that has graduate students competing for those treasured letters from superstars.

  7. I don’t understand what the quote “It feels like the message is that the careers of these [double-digit] people were worth more than the career of this one [purportedly] brilliant man.” is taken to mean, maybe I’m missing some context?

    • I had the same problem. It appears to relate to some context other than the Philip Roth case that preceded it and the Pruitt and other cases that followed it, but the quote is impenetrable without the context.

    • “maybe I’m missing some context?”

      I agree. I guess I got the general drift, but couldn’t discern the relevance of the student quote either. Maybe the author can tell us who the [double-digit]s were and how they fit in to the argument and then it would make sense.

      • I had perhaps too much going on in this post (perhaps part of why I wrote the draft, then didn’t post it right away), but I am a little surprised at the confusion over this.

        I have seen more than a few cases where a person was accused multiple times of bullying, fraud and/or sexual harassment and had glowing letters written for them. In some cases the letter writer(s) knew and debated what to do, but went ahead and wrote a glowing letter anyway — sometimes under the auspices of ‘well, despite complaints by six women none ever formally reported anything’ or ‘the investigation is not finished yet.’ That’s what I am talking about (in other cases where people say they did not know I think they could do more to figure out why they were so out of the loop).

        • Then shouldn’t the quote be:

          ““It feels like the message is that the careers of these [double-digit] people were worth ***LESS*** than the career of this one [purportedly] brilliant man.”

        • The folks at Language Log (and I guess linguists more generally) have a term for this (“misnegation”: https://languagelog.ldc.upenn.edu/nll/?p=1925) although it usually occurs in more complex settings (“no head injury is too trivial to ignore”).

          Also: while I am not (thank goodness) personally embroiled in the Pruitt situation, I’m closer than most – I know him, work at the same university, and am only a couple of professional connections away – and, as far as I know, the trouble really did come out of the blue. I have heard no whispers of missed warning signs/stuff that people knew or should have known. The blog posts by junior authors like Kate Laskowski ( https://laskowskilab.faculty.ucdavis.edu/2020/01/29/retractions/ ) are a model of transparency and due diligence. RJB’s suggestion about “separation of duties” (making sure that every task is performed by more than one person) is a good one, but I’m not sure it’s always feasible.

        • Hi Ben,

          Thanks for the word — but I am not sure that’s what was going on here. I think I may struggle more than some with making the comparison where one great man is worth more than the people we lose in order to promote that ‘star.’

          And unfortunately I think we lose more of them when we put so much effort into saying, ‘we never could have known! Let me tell you …’. But that seems to be what we do.

        • Thanks for the link to the Laskowski post—I didn’t know anything about the Pruitt thing and I agree that her post is a model of clarity. I noted two things:

          1) I thought her post made it clear that she believes she *should* have caught these issues years ago, and that they were fairly obvious once she began to look. So I don’t understand the claim that there were no red flags. The red flags were there, just no one was willing to see them.

          2) Although I dislike the term “fake data” when used to refer to legit simulations, fake data was literally what Pruitt was doing and it is funny to me how bad he was at it. Sure, he got maybe the first and second moments right, but overall he made up data patterns that could never have plausibly been produced in the experiments as designed. This made me realize that if he wanted to do a good job faking data, he would have needed to have an actual theory and implemented it in a formal model. And even if he never collected the data to test that theory, if he had been able to use prior knowledge of the mechanisms involved and the results of similar studies to come up with a theory that was so good no one would have suspected the data were fake, that’s actually an accomplishment! If only he had faked his data better, he might actually have been doing science.

        • Regarding this particular episode: I agree that retraction was the right step here. I checked Pruitt’s wikipedia page and it says, “He is currently writing a fantasy novel,” which sounds about right. I don’t know if he’s taken the step of apologizing to all the people he misled, but I hope he does so. So far, of all the fakers and manipulators we’ve heard about over the years, it seems that Stapel is the only one to have said he was sorry. I wish all these people, from the disgraced primatologist to the pizzagate guy to the humility columnist and the wikipedia plagiarist and everyone in between, not to mention all the harassers, would apologize not just for spreading untruths about the world but also for wasting the time of all the people who went to the trouble of tracking down what went wrong and, perhaps even more so, to their friends and supporters who ended up looking like idiots for sticking by them as the evidence piled up. I’ve seen enough not to expect such apologies but I remain disappointed. More and more I’m feeling like Stapel was a special person and admirable in his ability to come to terms with his actions.

        • P.S. The wikipedia page also says Pruitt was the “Canada 150 Research Chair in Biological Dystopias.” Indeed.

          I wonder if this is a partial solution to problems of research misconduct. Just give people appropriate-sounding titles. Weggy could be the Virginia Chair in Copypasting; Pizzagate dude could be the Cornell Chair in Imaginative Extrapolation; beauty-and-sex-ratio dude could be the Galton Chair in Sexism; that UCLA student who faked the survey data could receive a Fellowship in Creative Graphics; the critical positivity team could receive the Numerology Prize from the Association for Psychological Science, the disgraced primatologist could be given the Harvard Chair in Monkey Business—nah, that’s too easy, let’s say that William James Chair in Evanescent Phenomena, the Berkeley statistics professor who chased his postdoc around could receive the Australian Award for Persistence, the ESP researcher could receive the prestigious James Randi Endowed Chair in Sleight of Hand . . . Just make the title fit the crime. Don’t fire them, that’s too much trouble. Just gradually reduce their salaries to zero (Ok, legally I guess you can’t go below the minimum wage) while giving them endowed chairs with more and more ridiculous titles. That way, they don’t lose their jobs and they get the prestige that they crave.

        • With regard specifically to sexual misconduct, the atmosphere in academics is so stupidly permissive that it’s hard for me to call anything but the most egregious behavior “sexual misconduct”. I’ve seen dozens of instances where undergrad / grad students hook up with faculty / grad students while the two are in a working relationship.

          I’ve always found the universities’ laissez-faire attitude about relationships between students grad students and faculty appalling. In addition to encouraging both faculty and students to hit on one another and all the problems that entails, it frequently causes problems outside the two parties, like some smitten prof/TA hovering around one student (sometimes highly encouraging the attention) while other students are blatantly ignored. I mean seriously I’ve seen this go on for entire semesters with the rest of the faculty just ho-humming away.

          Faculty/grad/undergrad relationships should be strictly forbidden under any circumstances where the two parties have a working relationship. If as a faculty member you hook up with your PhD student, great – you can resign from your working relationship in every capacity with that student immediately or resign from your job.

        • I left more than 15 yrs ago but from many discussions I’ve seen in blogs, contact with people in academia and even from this post sounds like not much has changed. A few years ago on another academic forum the denizens were outraged when I claimed every department I’ve been associated with had at least one ass-grabbing (male) prof. Sadly numerous cases of harassment within that discipline have since surfaced yet the same people claim to have been utterly unaware! *very interesting* Meanwhile Andrew notes below “I’ve known lots of professors who have done these things”…. For my money it would be hard not to know a faculty member who has done these things.

          I stand by that claim and we wouldn’t be discussing it here if it weren’t still true today.

        • I am personally aware of a woman who is now a professor whose postdoc mentor told her that if she didn’t have sex with him he wouldn’t allow her to publish her postdoctoral work, and she wouldn’t be able to apply for professorships.

          she is a professor now. Years later the same professor was accused of similar things by someone else. There were ultimately few serious consequences. It makes me sick. Stories like this are yet another part of why I opted very definitely out of academia.

        • “I was going to write asking what century jim is writing from, but I had too many other things to do. Now this has been answered.”

          Faculty apparently have a different perspective than students and grad students, but the apparent unawareness strikes me as surprising given that items like this surface in the news relatively commonly – indicating that they occur on the ground with much greater frequency. In this case the system worked fairly well – at least the perp was fired. But OTOH not faculty either – just a journeyman athletics official. What if buddy was a high-powered tenured researcher with a strong record of brining in cash?

        • When I was at Berkeley some ~3 years ago, there were some pretty well corroborated accusations against a wide swathe of the faculty. One amusing case involved a statistics professor emailing a student with an invitation to Hawaii for a “dirty smoke-filled weekend of unadulterated guilty pleasure and sins”. I personally knew someone who had received some inappropriate advances from said professor, but simply rejected his proposal and decided to move on with their life and not deal with the messiness. I had also heard of quite a few incidents that were not tied to any later public reckoning. What was clear in many of these cases was that the offenders were in the habit of behaving in that manner with impunity until one person decided it was worth the trouble.

          I’ve always found the universities’ laissez-faire attitude about relationships between students grad students and faculty appalling.

          My general impression that faculty-student relationships are now technically prohibited, but inappropriate behavior from faculty is still tolerated in practice, at least much more than in other white collar workplaces. Unprofessional abuse from faculty members directed at grad students or other faculty members also seems to be pretty much common place. Tenured professor, if you can get it, gets you a kind of formal immunity well beyond what’s appropriate.

          https://en.wikipedia.org/wiki/Geoffrey_Marcy#Sexual_harassment

          Notably, he was not fired. The university wanted to fire him for a long while, but formally could not by its own convoluted set of rules. After it successfully forced him, out, it could not strip him of his honorifics, so he can continue to identify himself as emeritus. I really struggle to understand the potential benefit of this red tape nonsense.

        • Somebody:

          Wha??? That’s a Berkeley statistics professor you’re talking about??? OK, I guess there’s been a lot of turnover in the faculty since I taught there. Indeed, all the profs I knew from those days are either retired, deceased, or are old enough that “dirty smoke-filled weekend of unadulterated guilty pleasure and sins” is probably, ummm, I dunno, insert some geriatric joke here. In all seriousness, jeez, yeah, it sounds like someone would only send an email like that if for some reason he felt immune from consequences.

        • @Andrew

          This guy was an adjunct professor, so maybe of a different kind than the old guard of frequentist asymptotic theorem provers than you’re thinking of. But yeah, he was wild. A couple semesters before this news broke, I had a friend in one of his classes. Apparently, he held his office hours at 6:30 in the morning to dissuade anyone from coming. He insisted that his class rate him on ratemyprofessors. Then, when he got the ratings, he didn’t like what they said, so he sent an angry email to the class wherein he declared his intention to not lecture for the rest of the semester and to make the final extremely difficult. Considering that he taught for two more semesters after that, I guess he *was* pretty immune to consequences.

        • For kicks, I checked the ratings

          “There were 400 students and he only printed 300 finals. Can’t bear such irresponsibility.”

          “Do yourself a favor and take it with someone else. Howard doesn’t know how to teach and he will ruin your college experience. He didn’t have enough finals for the whole class last semester. This is outrageous and simply not fair to the students who had to wait half an hour before he brought more copies. You have to get the book and reader.”

          “Exam fiasco: He didn’t bring enough final exams for part of the class and made us wait for an hour. This then lead to part of the class getting more time. Some people tried to email him and he said not to waste his time in all caps. He seriously had the nerve to say this after he made us wait for an hour!! Rumors that he writes fake reviews here.”

        • Somebody:

          OK, here it is: “OPHD found Howard D’abrera’s repeatedly emailed a student with invitations to Hawaii and other destinations, mentioning orgies and threatening to lower the student’s grade if they did not accept the invitation. Though D’abrera denies any sexual intent to his emails and communications, OPHD found D’abrera had more likely than not violated UC sexual harassment policy. He was placed on administrative leave the day the complaint was filed and later resigned in January.”

          I love the bit about how he “denies any sexual intent to his emails and communications.” Orgies, that’s all. Nothing sexual. Also, it seems that he was “the old guard of frequentist asymptotic theorem provers.” A web search finds that his Ph.D. was from 1973 on the topic of Rank Tests for Ordered Alternatives. Kind of a crime to have someone like that teaching intro statistics to undergraduates. There must be a few thousand UC grads in the area who knew more applied statistics than this guy, who would’ve been more qualified to teach the class, would’ve treated the students better, and, hey, no sexual harassment either. This is the sort of case that makes it really really clear what’s the old boy’s network: a prestigious university that could get all sorts of amazing people to teach their classes but instead hires someone unqualified and who doesn’t even care about the job. Obviously these jobs are not being filled on the basis of “merit,” under any definition of the term.

          Just to be clear: I’m not saying all the faculty should be top researchers. You can be a good teacher without doing good research. But to hire someone who doesn’t do research and is a terrible teacher . . . there’s no excuse for that, even without considering the sexual harassment issue.

        • Oh hey I was in that class! The one where he forgot to print the final exams. First statistics class I ever took. Also the worse class of any subject I’ve ever taken. I don’t really remember the content of what he taught (it was intro to stats/probability) but I do remember him being antagonistic to students during lecture and complaining about the RateMyProfessors reviews. It was weird, I don’t think his teaching of stats was particularly objectionable (in the sense that it was more or less the standard way intro stats is typically taught, but I agree with Andrew that this typical way is not the best way) but he would mix in hostility in his lectures e.g. ranting about the students who didn’t attend lecture.

          But he also brought Tim Tams (an Australian cookie) to class and raffled off a free international plane ticket to the students who got the top grades or could solve his riddles or something like that. Maybe that was his way to try and boost his RateMyProfessor score. The cookies were fine I guess but all the rest added up to what I think was unprofessional behavior, and this is not counting the sexual harassment.

          @somebody: I don’t remember the exact time of his office hours but I do remember he held them before an 8 or 9am lecture. So it might have been 6:30. He seemed to take a lot of pride in waking up early and seemed like he wanted to flex that.

        • I did some googling and it seems that Howard D’Abrera was scheduled to retire from the University of Sydney (Australia) in 2005. I wonder why he retired? Possibly sexual harassment, I suppose, or maybe just being a really bad teacher.

          You can also look up D’Abrera’s salary at the University of California. Here it is:

          2010 30,099.96
          2011 60,199.92
          2012 30,099.96
          2013 55,183.00
          2014 60,200.00
          2015 80,121.00
          2016 13,716.00

          The database doesn’t go back before 2010.

          2016 was the year he was fired resigned so that’s why he didn’t pull in another 80K that year.

          To pay $80,000 for an adjunct to teach introductory statistics . . . it’s not hard to find someone who can teach introductory statistics. On the downside, dude had no research record and was a terrible teacher. But, on the upside, he was a well-connected member of the old boy’s network. I’m still trying to picture in my mind this 70-year-old dude trying to get his 19-year-old students to party with him in Hawaii. None of it makes sense to me—except the bit about the University of California paying hundreds of thousands of dollars to an unqualified creep who happens to have the right friends. That part I can believe. It’s not like the administrators were spending their own money, after all!

        • “I’m still trying to picture in my mind this 70-year-old dude trying to get his 19-year-old students to party with him in Hawaii.”

          Ha, by email no less! Class and tact. Only from Berkley

        • His 2019 death notice in the Sydney Morning Herald referred to him as a “gifted educator”. This is just 3 years after he resigned from UC Berkeley.

  8. When searching for senior leadership positions (Dean, Provost etc.) it is common practice for the search committee to call “off list” references who weren’t suggested by the candidate.
    As someone who has sat on award committees for national awards in a couple of professional societies I can say that a very large “discount rate” is applied to what gets written in nomination letters for these things – the committees tend to focus on demonstrated accomplishments (although personal reputation and a lack of having made serious enemies also matters a lot).

  9. To all:

    For my opinion on letters of recommendation, see here (“One more reason to remove letters of recommendation when evaluating candidates for jobs or scholarships.”) and here (“One more reason I hate letters of recommendation”).

    • I think I’d be fine with this solution.

      But what should a PhD/MSc/postdoc supervisor do when they know of multiple complaints of fraud/bullying/sexual harassment? Work with the accusers to find a way for them to come forward in a way that is feasible? Hope the system ‘works’?

      • “Hope the system ‘works’?”

        What system? There is no system. Each university or research org acts on it’s own. This is appropriate for employee relations issues but not for fraudulent research. Sexual harassment is conduct that impacts two individuals. Fraudulent research has impacts across all of science.

      • Lizzie:

        I’m not sure. In my experience, when graduate students have behaved unethically, it’s often the other graduate students who are aware of the problems first, but they don’t always have a way of telling the faculty about the problem. As a faculty member, I don’t usually think very hard about the social interactions among the students I’m working with.

        And then there’s fraud, bullying, sexual harassment, and other unethical behavior by faculty. I’ve known lots of professors who have done these things, and it seems that information about these problems is spread informally. In more than one case, I was told about bad behavior of colleagues and I was like, OK, and I didn’t act on the information. When I say, I didn’t act, I mean that I didn’t do anything. I didn’t go the university administration or the department chair or whatever; also I didn’t talk with the perps directly; also I didn’t alter my own behavior toward them. In retrospect I regret not talking directly with any of these people when I first heard about these things. It could’ve reduced the problems that arose later.

        In the cases I’m familiar with, I don’t think “the system” worked well. In many cases, the perps just keep doing what they’re doing; in other cases, the person was fired or forced to leave, but by then lots of damage has already been done. I think it would be better to talk with the people first, and then, sure, if they keep doing it, go through the disciplinary processes. But it’s my impression that lots of bad stuff gets done for years without anyone sitting the perp down and saying, Hey, you better clean up your act or you’ll be in big trouble.

        • This is my sense of most of the cases I know of also. And it’s why I don’t know what to say to graduate students. Part of what I’d like to see, and was getting at in my post, is more senior people speaking up to one another and expressing contempt at bad behavior.

  10. I see this as something of a call for academics to take greater personal responsibility, in the sense that when we endorse someone, we personally should be held accountable for that endorsement. This can be good if the endorsee turns out brilliant and bad if they turn out to be a huckster. In particular, if things turn out badly, “being held accountable” means that we should have to defend our prior endorsement. In the end, others will decide if we were justified based on current knowledge or if we should have known better, but we don’t get to just walk away. (The Sandusky case is perhaps a better example of “recommenders” who *did* know better and still did jack shit.)

    This is in contrast to what usually happens when we say, “the field also bears some responsibility for the actions of a fraud.” Instead of taking “the field” to mean the actual people that comprise it, we outsource the responsibility to some mechanical bureaucratic procedure. E.g., if only Stapel had pre-registered his studies (or used Bayes factors, etc.), we would have known! No, we *already* should have known if his collaborators would have taken any personal responsibility for their conduct. In the end, we found out about Stapel because of the actions of a person who took great personal risk to herself. She accepted the responsibility, and we should all do the same.

  11. I wrote a comment applying the post’s own logic to itself but the comment did not make it through moderation. Let me try again.

    Lizzie, why should we think that you owe your position to hard work and creativity rather than cronyism?

    As gec points out, the post calls “for academics to take greater personal responsibility, when we endorse someone, we personally should be held accountable for that endorsement.” I find this post cause for “concern” (for example, “the person Philip Roth decided should write his authorized biography has a history of sexual harassment accusations.” One, accusations should not be taken at face value. Two, so what, what does such accusations have to do with anything?). Andrew, have you written letters recommendation for Lizzie? If so, what do you have to say for yourself? Because as, again, gec points out, if so, that, given the post’s own logic, “means that [you] should have to defend [y]our prior endorsement.”

    • Mark:

      I’ve never written a letter of recommendation for Lizzie—we collaborate but we are in entirely different fields. But if someone were to ask me to write a letter for Lizzie, perhaps for some joint appointment or because someone has a goal of demonstrating her interdisciplinary breadth, I’d gladly write one. As noted above, I hate the institution of letters of recommendation, but that’s the way the system goes, and I don’t hate letters so much that I refuse to write them. I write lots of letters, actually!

    • “accusations should not be taken at face value.”

      Absolutely. Strongly agree. Nor should they be judged in the Kangaroo court of Twitter or the blogosphere. Not the place to render formal verdicts on misconduct. Such things demand a systematic investigation.

  12. I agree with many others here that the emphasis on letters is misplaced.

    The problem is that there is no formal mechanism for dealing with misconduct in science. If someone notices some problems or irregularities, who do they report it to? Twitter? Twitter is a shitty place to address these problems. The various agencies in different countries (NSF, NSERC, etc) need an office that handles claims of research misconduct. The problem with Pruitt is interesting – in a way he’s right to sick his lawyer on people to get them to shut up until investigations are complete. He has rights too. But OTOH, from the community perspective, who’s doing the investigation, what exactly are they investigating (what laws/rules were broken?), and how far does their authority reach? People need assurance that claims won’t be shoved under the rug.

    That’s why we need an agency or office to handle claims at the national level. The office probably wouldn’t have the authority to fire people or retract papers itself, but it should issue reports with recommendations that carry enough heft that it’s recommendations are universally implemented. Those recommendations should include fingering people and institutions who are complicit in promoting misconduct. And it should also be tasked with determining the funding response from Federal agencies of whatever country it’s in.

    I also don’t get the [double digit] thing.

    • jim (assuming you are the same jim – if not, I apologize)

      You need to make up your mind about regulation. Here you are calling for “an agency or office to handle claims at the national level,” while you only recently complained that
      “In the case of radiology, if there were a functional AI alternative, we can be sure that the regulatory state would be use by doctors to protect their turf, as we already see with so many other licensing requirements.”
      Regulatory capture is a real thing, but it is not always and automatically the situation. The details matter – how is such an institution set up, who staffs it, who do they answer to, how are they compensated, etc. Setting up an office or agency is easy – making it effective is hard. So, how would you propose we make such an agency not fall victim to the same regulatory state you lament regarding radiology?

      • Same Jim!

        No question about that. However I don’t see the conflict. I’ll suggest their are two distinct categories of “laws.”

        Administrative regulations that control *market access* and rules that set a *standard of conduct* and enforce it are two different things. Presumably both a radiologist and an AI system automating radiology would be held to the same rules regarding standard conduct. That is, fraudulent claims, negligence etc on the part of an individual radiologist or fraudulent claims negligence etc on the part of a company offering AI technology would be treated the same.

        Administrative regulations could be used both ways – either by companies offering an AI product or by human radiologists – to protect their share of a market.

        With regard to scientific misconduct, the rules are already there at the University level. They just aren’t very effective. I blithely presume they are incomplete, convoluted and inconsistent, because this is something Universities don’t do very much, thankfully. Because they don’t do it very much, they don’t have the necessary expertise or capacity. And because they reap the benefits of researchers drawing in cash, they have an obvious conflict of interest.

        Having a national office that investigates these claims would remove the burden from the universities and create a national standard. Uni’s could then streamline local regs.

        We’re also discussing two completely different activities: business and science. Businesses are privately funded and ultimately regulated by success of the product or service in the market, as well as public laws of conduct. Science is publicly funded and has no independent market. I don’t like Apple products so I don’t buy them. But I pay for publicly funded science like it or not.

        People seem to be suggesting that, because every regulation doesn’t result in “regulatory capture,” those regs that aren’t “captured” are by definition good. This is misses the bulk of the damage caused by regulations. Rent control laws are generally not subject to regulatory capture yet are widely agreed to be highly detrimental. Regulations can force the hand of the market in ways that are destructive, capture or no.

      • I think there’s a difference between regulatory red tape and enforcement. Stating clear rules and then investigating violations after the fact is an order of magnitude less onerous than requiring every action be approved. Notably, it preserves permissionless innovation
        http://www.econtalk.org/michael-munger-on-permissionless-innovation/
        I’m thinking SEC vs FDA here. It’s true that even the rules of the game can be rigged, but no system works perfectly without some public attention

  13. The topic of the rising stars’ supervisors was raised in a previous comment and I want to build on it. The supervisors have the responsibility of assessing the accuracy and validity of their students’ and postdocs’ data before publication. They also bear responsibility when their rising star’s data come under suspicion of misconduct or their publications are retracted.

    In the case of Denon Start, his supervisor (Ben Gilbert) failed to perform due diligence in assessing the accuracy and validity of Start’s data prior to publication. Prof. Gilbert has never faced public scrutiny or censure to the same degree as Start. An example of the inequality in blame for the retraction of a paper comes from the following:

    Start’s co-authors (Gilbert and Art Weis, both tenured professors at U of Toronto) were allowed by the editor in chief of American Naturalist, David Bolnick, to issue a retraction notice and then publish a new version of the same article. In this notice, they say they needed to correct accidents and errors in data analysis but they don’t say that they found any research misconduct.
    https://www.journals.uchicago.edu/doi/10.1086/715870

    The case of Denon Start is peculiar. In addition to the imbalance in punishment between Start and his supervisors, The Canadian Society for Ecology and Evolution rescinded two awards previously given to Start:
    https://www.csee-scee.ca/public-statement-from-csee/

    CSEE’s decision is concerning because of a detail that was noted by Retraction Watch. CSEE did not notify Start about their process for rescinding his awards. Instead, Start learned of this decision only after CSEE announced it publicly and on Twitter. Start was never allowed to speak in his defense:
    https://retractionwatch.com/2021/05/03/ecologist-who-lost-thesis-awards-earns-expressions-of-concern-after-laptop-stolen/

    It is unethical for any scientist to be stripped of awards without any prior notification of the process, and they have the right to speak in their defense.

    This unethical process is similar to the way that The American Naturalist investigated issues raised about Start’s papers. As reported by Retraction Watch reported, Am Nat editor in chief, Daniel Bolnick, admitted that no official complaints were made against Start, that the issues raised were made anonymously by someone he knew, and that he took accusations made on Twitter into account when he wrote his expression of concern regarding Start’s work. Because the accusations made by Bolnick’s anonymous acquaintance regarded work conducted by Start during his phd in at U of Toronto in the department of ecology and evolutionary biology and prior to his UC Davis work, this acquaintance is a member of this department. I understand why a a person making this kind of allegation might not want to be named publicly; I am posting this comment anonymously. What I don’t understand is why Bolnick would issue statements like these against Start without providing solid evidence in data to back up his statements – remember that Start’s co-authors, Gilbert and Weis, said that mistakes were made but not misconduct.

    This also raises the question of what process was activated in Start’s department to address the allegations made against his work. Given the secrecy of the CSEE’s process, the question of whether or not Start’s department informed him of the allegations needs to be answered. If Start’s department conducted a real investigation, did they pass this information to Bolnick so that he could make his decisions and public statements?

    There is a major imbalance in the punishment given to junior unestablished scientists and their senior established supervisors. The issue of institutional permission of this imbalance – in journals and universities – needs more discussion.

Leave a Reply to Andrew Cancel reply

Your email address will not be published. Required fields are marked *