Slashdot Mirror

Inception! This is an actual problem that is fairly well researched but ignored by many for the sake of profit and others for the sake of more comfortable ignorance.

- Reverse-discrimination against men? Rejected, per the observation that there is evidence of discrimination against women when gender is identified.

Study CLAIMS no evidence of gender bias (while concentrating on bias towards or against women only) when gender is identifiable (from name, photo or profile on GitHub).
Study shows bias in interpretation of data though.
It ignores error bars in graphs (no real numbers are shown, just percentage graphs) AND paints a simplistic "more-pulls-4-women-except-when-identified" image.
Study's graphs on the other hand show NO BIAS towards women or men.
https://peerj.com/preprints/17...

First of all, when COMPLETE IDENTITY (and thus quality of work, not just gender) is known - there is NO SIGNIFICANT DIFFERENCE between male and female "pulls".

In "insider" cases where identity is known, supposedly gender-positive (you can tell who's male or female) IDs have a slight pro-female bias (~87.5% female vs. ~86% male), with barely existent error bars for males and tiny bars for females.
I keep using tildes, cause all that is presented are graphs - no percentages or real numbers are shown for this case, you have to eyeball them.
Where supposedly gender-neutral IDs are used, there is a tiny pro-male bias (~88% +/- ~1% female vs. ~88.5% +/- ~0.7% male).

I.e. There is NO pro- or contra-, male or female, bias when COMPLETE IDENTITY is known.

ON THE OTHER HAND...
In "outsider" cases where gender and identity is supposedly unknown (though clearly identifiable through email vs social network profiles comparison - which is where they got their data and what they are basing their study on)...
There is a slight pro-female bias in supposed gender-neutral IDs (~72% +/- ~1.5% female vs. ~69% +/- ~1% male) and a tiny pro-male bias in supposed gender-positive IDs (~62.3% +/- ~0.7% female vs. ~63% +/- ~0.0something% male).

Again, these are eyeballed values.
Study lists female gender neutral percentage in "outsider" cases as 71.8% and female gender identifiable percentage as 62.5% - claiming it as proof of bias against women who identify as such.
These are the only values presented in numerical form in this hypothesis.

I.e. There is NO pro- or contra-, male or female, bias when full identity is (supposedly) unknown either, but study tries to claim the opposite by ignoring own findings.

In other words, while looking for a hypothesis to explain their findings of bias, they accidentally took the gender-bias hypothesis they found behind the shed, and controlled it by putting a bullet in its head.
Then, not noticing that said finding of bias is dead, they kept on beating it, claiming it's alive and highly agile.
Kinda like in the dead parrot sketch, only here the salesmen really do believe that the parrot is just stunned and pining for fjords.

Thin women are considered more attractive generally, even in countries with low food security.

https://peerj.com/articles/115...

"Participants from three Caucasian populations (Austria, Lithuania and the UK), three Asian populations (China, Iran and Mauritius) and four African populations (Kenya, Morocco, Nigeria and Senegal) rated attractiveness of a series of female images varying in fatness (BMI) and waist to hip ratio (WHR). There was an inverse linear relationship between physical attractiveness and body fatness or BMI in all populations. Lower body fat was more attractive, down to at least BMI = 19. There was no peak in the relationship over the range we studied in any population"

Also:

"For example, the BMIs of Playboy centerfolds and glamour models over the last 50 years are almost all in the range 17 to 20 (Katzmarzyk & Davis, 2001; Tovee et al., 1999; Voracek & Fisher, 2002). Women and men asked to manipulate female 3D computer models to make them maximally attractive make them have BMIs of 18.9 and 18.8 respectively (Crossley, Cornelissen & Tovee, 2012). The biggest outlier in previous studies of attractiveness at low BMI was the observation that in Poland the highest rated attractiveness was at a BMI of 15 (Koscinski, 2013), and potentially lower as this was the smallest stimulus in the set presented."

.

The problem with Microsoft's approach in TFA, is akin to "when all you have is a hammer, everything looks like a nail".

It is actually a lot better to solve the actual root problem, than trying to find and treat symptoms.

Yes, but how? Peer review is not cutting it. Suppose you were not all that well educated on what has already been done, and what has not. You get an idea, try to find if anyone has done any work on it, but find none. You carry out the study and try to publish. A peer reviewer sees that it is essentially a rehash of the old idea, carried out with modern methods, and perhaps says it is novel enough for the journal. The editor agrees and your paper is rejected. What do you do? Well, I can tell you that nobody would abandon the paper if it is already in a written manuscript form. Instead, most scientists would not even go and generate more data for the paper, but rather, if they are even honest, cite the original work, but just mention it in passing and emphasize that what was done in their study was, in fact, slightly different. And they'd be right: The original papers probably used slightly different techniques. They now send the paper to a different journal and hope for a different referee (which in all likelyhood is going to happen, especially as you do get to recommend referees; the editor needs not abide by these recommendations, though). The odds are ever in your favor for if you were not aware of the previous publications before, it is a near-certainty that the average peer is not aware of them either.

Let's start deconstructing the story from the beginning. Why was the researcher not aware of the previous, perhaps well-known, work? Just wrong Google keywords? That, too, can be an issue, because in 30 years the terminology does evolve with the field. If one has not been there when it began to change, as fields become more interdisciplinary, the change might be difficult to see. And who was there since the beginning? If the current field is essentially a mash-up of people from several fields, medicine and physics, say, how should a medical researcher be aware of the stuff that has been done in physics? Should it be expected? I'd say yes, it should: Either you learn the stuff, or get a co-author who knows the stuff that you don't want to learn. However, publish or perish. You have to start contributing and publishing even before you have a good grasp on the field, so when you venture off to new paths, it is easy to get to hasty conclusions. This pressure to publish is why you keep trying different journals even after being initially rejected. If you don't get this piece out, you've wasted months of lab resources.

If the manuscript is rejected from one journal, surely, surely in this age of internet and open access to all kinds of information, the editor of the next journal will have the reports by the peer reviewrs at hand. Nope. What? Nope. With very few exceptions (notably PeerJ) peer review reports are not made public in any for or shared across journals (sometimes they are, when the publisher of the journals is the same). There are efforts to change this, but academia is a slowly moving beast, and is run by governments and government funding, something that is even slower to change.

Finally, after a successful shopping for peer reviewers, the paper is published. What about post-publication peer review? Surely someone will publicly point out that the results have already been known for a long time. Maybe the peer reviewer of the original submission should come out and denounce the work? No such luck. Happens very rarely, for people don't want to stick their necks out and contradict their peers when they know that these peers are going to be sitting on the other side of the peer review or a grant application at some point or another. See for example this piece of news on a recent writeup by a Harvard professor: He's essentially arguing that research showing previous research false is wrong and should not be published. My phrasing was perhaps too extreme, but read the article and judge for yourself; I think you'll g

"A lot of people ignore the collateral functions of the so-called 'peer review' system administered by the publisher."

I don't see this as a stumbling block, though. There are already public-access peer-reviewed journals. They may have a way to go yet but I expect them to get better and their number to expand in the near future.

It it surprising to me that the article makes no mention of language workbenches, when they are a good solution to the problem the article describes. See our preprint https://peerj.com/preprints/112v2/ for an example of language interoperability, and references therein for descriptions of language workbenches.

Peer-review. PeerJ is particularly good on this, in that it allows the whole peer-review history of papers to be published alongside the final version: the original submission, the reviews, the handling editor's decision, the authors' rebuttal letter and revision, subsequent editorial comments, etc. As an example, you can see this audit trail for our own PeerJ paper on sauropod necks.

Peer-review. PeerJ is particularly good on this, in that it allows the whole peer-review history of papers to be published alongside the final version: the original submission, the reviews, the handling editor's decision, the authors' rebuttal letter and revision, subsequent editorial comments, etc. As an example, you can see this audit trail for our own PeerJ paper on sauropod necks.

From the FAQ :

What subject areas do you cover:

PeerJ considers submissions of Research Articles in the Biological and Medical Sciences (this scope includes, for example, disciplines such as the life & biological sciences; biotechnology; basic medical sciences; medical specialties; health sciences and other similar fields).