Wednesday

191023

Workout of the Day

Rest Day

Post thoughts to comments.

Why Do Scientists Cheat?

9

Science is immersed in a crisis of reproducibility, but how can science go so badly wrong? Here, an anonymous author with intimate knowledge of the academic world and its research and publishing practices addresses this question, drawing examples from his formative years in the university and ensuing professional life. Citing firsthand experiences wherein he witnessed the impact of questionable research practices, conflicts of interest, and perverse career incentives, he explains how “the very enterprise of research can deviate systematically from the paths of truth.”

Read MoreWhy Do Scientists Cheat?

Open Workout 20.2 Analysis

The second workout of the 2020 Open asked athletes to perform as many rounds as possible in 20 minutes of 4 dumbbell thrusters, 6 toes-to-bars, and 24 double-unders. For many, there wasn’t a need to intentionally break up the sets (although many double-under sets were frustratingly broken due to trips and misses). The best performers in this workout breezed through smooth and unbroken rounds, with few if any misses on the double-unders. For those athletes, it was largely a difference of rest and transition time. Here's a look at who participated, which workout versions they chose, and how they fared.

Read MoreOpen Workout 20.2 Analysis

Comments on 191023

9 Comments

Comment thread URL copied!
Bruce Warren
October 23rd, 2019 at 1:46 pm
Commented on: Why Do Scientists Cheat?

One obvious correlation is the rampant fraud in "climate science". In this genre, the fraud is driven by a religiously held belief in ACG - for which there is little to no reproducible research in support. Yet, computer models are constantly published as "science".

Comment URL copied!
Leonardo Nascimento
October 23rd, 2019 at 4:39 pm

First of all, computers models are obviously science. Most advancements in physics, chemistry, engineering, etc are based on computer models. Second, climate science is not only based on climate models, there is a bunch of experimental data involved. Also there is no religiously held belief as criticisms keep being being refuted time after time. While the point of "low standard journals" is valid and there is a lot of shitty research out there (specially in medical sciences and the likes that take correlation as causation) climate science is not one of them. Do you research, buddy.

Comment URL copied!
Richard Feinman
October 23rd, 2019 at 11:53 am
Commented on: Why Do Scientists Cheat?

The problem is real and extensive but this piece is not accurate or appropriate. First, "cheat" means intention and the standard witticism in this business is referred to as Hanlon’s Razor which states that you should not invoke malice until you’ve excluded stupidity. Especially in nutrition and medicine but even in hard, that is, mathematical science, there is honest but limited understanding of scientific and statistical principals. One problem rests substantially with the journals in maintaining very low standards. To take one extreme example, medical journals accept, may even demand, an “intention-tp-treat” analysis of data. The principle says that if a subject is randomly assigned to an intervention, their data must be included in the analysis even if they dropped out of the study; even if you don’t take the pill, your outcome will tell us whether the pill is good or not. The idea is as foolish as it sounds. I and several others, including professional statisticians, have called it out for what it was (Intention-to-treat. What is the question? Nutr Metab (Lond). 2009; 6: 1; doi: 10.1186/1743-7075-6-1) yet the practice persists. Intention-to-treat analysis answers the question “what is the effect of being assigned to an intervention?” but most of us are more interested in the actual outcome if you do take the pill. In exasperation, many simply report both intention-to-treat and. “per protocol” data (you did what the experiment required but many simply report intention-to-treat. The motivation seems to be screwy thinking since, while good for showing that low-carb diets and control are “the same out one year,” almost always makes your data look worse than it is


The article is also somewhat inaccurate. Without in any way underestimating Gary Taubes’s contribution — I described him as the Thomas Paine of the low-carbohydrate revolution — he did not effect a “successful overthrow … of the long-dominant hypothesis that dietary fat causes heart disease, strokes, and obesity” but rather tried — with some success — to bring out to the public what numerous papers in the literature had demonstrated. Nonetheless, we are burdened with what I would call many Mulvaneys: the USDA guidelines says that total fat is no of concern but recommends low-fat products.


There are numerous problems but a single over-riding cause is the poor standards of journal editorial practice. The Journal of Biological Chemistry has begun to set up some guide to practice. A recent article explains the difference between standard deviation (SD) and standard error of the mean (SE) which I will explain in next comment.

Comment URL copied!
Bruce Warren
October 23rd, 2019 at 1:46 pm

Sorry, man. You're dead.

Comment URL copied!
Russ Greene
October 24th, 2019 at 4:33 pm

Dr. Feinman,

Regarding intentionality, I wonder what you think of the source this article cited to support that point:

"How Many Scientists Fabricate and Falsify Research? A Systematic Review and Meta-Analysis of Survey Data"

https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0005738


Yes, the study was based on survey data, but I am not sure how else one could systematically evaluate this issue.


Here are the results:

"A pooled weighted average of 1.97% (N = 7, 95%CI: 0.86–4.45) of scientists admitted to have fabricated, falsified or modified data or results at least once –a serious form of misconduct by any standard– and up to 33.7% admitted other questionable research practices. In surveys asking about the behaviour of colleagues, admission rates were 14.12% (N = 12, 95% CI: 9.91–19.72) for falsification, and up to 72% for other questionable research practices."

Comment URL copied!
Lamiece Hassan
October 23rd, 2019 at 6:45 am
Commented on: Why Do Scientists Cheat?

Good scientists know where to cut corners? There’s no place for anyone but alpha males at the forefront of science? Unless, maybe, other, more modest men. Have I misunderstood something here? As a scientist, CrossFit fan and a woman, I’m disappointed that you’re giving a worldwide platform to this kind of piece.

Comment URL copied!
Russ Greene
October 23rd, 2019 at 11:41 am

Lamiece,

I cannot tell if you misrepresented, or merely misunderstood, this piece.


Not once did "Anonymous" seriously suggest that "good scientists know where to cut corners." Instead the author lamented the fact that this is what he/she learned at his/her "stellar PhD university." The author's point is that the incentives in science do not reward "meticulous and scrupulous" research, and this is unfortunate.


Moreover, the piece does not endorse the presence of "alpha males" at big and successful labs. Quite the opposite. It warns of the problems inherent with the alpha males' aggressive, even reckless approach to science.


The three scientists the author identified as "men" are in fact men. One may not reasonably infer sexist bias from describing men as men any more than one may infer jingoistic tendencies if an author describes Americans as Americans.


Perhaps you object to the absence of female sources, but that concern too is misplaced. The author cites Marion Nestle. Marion is in fact a woman. Perhaps you assumed otherwise, but that betrays your own views, not the author's. (Please do not infer anything about my personal worldview from my accurate identification of Marion as a female).


Now, if you have any sound objections to the substance of this piece I'd love to hear them.

Comment URL copied!
Lamiece Hassan
October 24th, 2019 at 5:36 am

Russ, my point is that this piece presents a bleak, masculine, cynical view of science, and whether the author agrees with the current state of affairs or not or not, the editorial decision to publish this perpetuates a series of unhelpful, narrow stereotypes about science and what it means to be a scientist.


Science is so much more diverse than this and I would have much preferred to see constructive thoughts about how we tackle these challenges instead of using click bait titles tarnishing the image of scientists as ‘cheats’. This may have been the author’s personal experience - and I’m sorry about that because that is not universal - but I don’t think CrossFit.com is the right place for this kind of piece. In my opinion, there are a 1000 more papers and commentaries that could be more relevant to a CrossFit audience.

Comment URL copied!
Russ Greene
October 24th, 2019 at 2:28 pm

The view may be "masculine" and "bleak." I might even grant you "unhelpful," at least in the sense that it's unlikely to be followed by any meaningful reform. I'd not ever accede to "inaccurate" or "irrelevant," though, for the simple reason that neither is true. If you believe the claims to be inaccurate, the burden is on you to not merely describe them, but rebut the arguments made and the sources upon which they rely. This you have not done.


Since CrossFit began to expose bad science, it has encountered many critics like yourself. Rather than meaningfully disputing our allegations, these critics asked why CrossFit is speaking up.


How many instances of scientific misconduct, hidden funding, and other unethical behavior in science must CrossFit expose before it earns the right to speak, in your view?


Would taking down the most prolific researcher in exercise science, William Kraemer, for fraud suffice? Or proving that strength and conditioning field's premiere association had committed scientific misconduct and perjury?


Or proving that the CDC and NIH's foundations had failed to comply with their legal requirement to annually, publicly report the source, amount and restrictions associated with each payment they accepted? How about suing the Department of Health and Human Services over its failure to comply with the Freedom of Information Act, and preventing the department from retroactively redacting emails suggesting its own staff was aware of unethical behavior occurring regarding its corporate partnerships?


These are just a few instances. If you will just delineate what amount of work in the field would grant us the right to speak in your eyes, I am happy to lay out the full scope of our activities.

Comment URL copied!