Hacker News Reader

Another interesting bias along those lines: https://en.m.wikipedia.org/wiki/Women-are-wonderful_effect

I imagine these biases swing in all sorts of directions depending on the context. Some are intuitive, many are not.

good point.

How do you reconcile the https://en.m.wikipedia.org/wiki/Women-are-wonderful_effect with https://en.wikipedia.org/wiki/Gender_pay_gap

Not needed because "women are wonderful" is to some degree mostly orthogonal to "men are more professional/deserve more/"... wonderful good attributes are not needed or even to some degree counter to "high performance" (but depends a lot where we look at, sure!)

Also pay-gap has likely also to do with the current men-bias also? Wonder if that would be the same if we had 90% woman throughout in these decising positions..

In the end all mine assumptions, what I actually want to say is: I don't think those two contradict or relate at all.. what GP already said well with "biases swing in all sorts of directions depending on the context"

Not GP, but:

From your wikipedia link

> There are two distinct numbers regarding the pay gap: non-adjusted versus adjusted pay gap. The latter typically takes into account differences in hours worked, occupations chosen, education and job experience.[1] In the United States, for example, the non-adjusted average woman's annual salary is 79% of the average man's salary, compared to 95% for the adjusted average salary.[2][3][4][5]

The remaining 5% could be from the "women are wonderful" positive attributes not being the narrow selection of ones that are highly sought after in well-renumerated jobs, or (spitballing here) psychosocial (Expectations of a pay gap driving negotiation behavior or something).

They're two related but ultimately very different phenomena, they don't necessarily need reconciliation.

I'm not a psychologist, I have nothing but that BA in the field, but if I had to make a guess, my guess is that "wonderful" is not the same as "competent".

Even at the time, there were studies that showed traits associated with women were rated more positively by men and women than traits associated with men.

The way you study this is you'd give a list of words:

gentle caring assertive stubborn aggressive loving

(ideally you'd randomize the word order too)

And then you'd have a group of subjects rate them as more associated with men or women on a scale.

Ideally you'd get a big sample and replicate this study.

Then either in the same study at a different time, or another study, you'd take those same words and you'd ask your subjects to rate them as positive or negative.

That's where you see effects like the one mentioned in the wikipedia article.

That kind of thing's been done a bunch of times.

But studies on gender and competency have been done too, and at the time they showed the same pattern as we found.

I say "At the time" because this is >20 years ago.

>my guess is that "wonderful" is not the same as "competent".

Right. There are two different sets of benefits that help or hurt someone in different ways. Being competent when standing trial can work against you while being wonderful will reduce the risk of a conviction and reduce the sentence if you are convicted.

We have identified the "competent" bias and are taking steps to correct it, but we need to do the same with the "wonderful" bias in other systems. For starters we need to recognize how strong that bias is in certain fields. For one example, there are specific crimes that people would bet are extremely gendered in nature, and the crime statistics show they would be right, but interviewing the population at large and querying victims, including those who never went to the police or who were turned away by the police (or even worse, who couldn't legally be victims because of how biased even the laws are), we see the gender component goes away. The rate of men victimized by women and women victimized by men are at near a 50/50 ration (I think 49.8 to 50.2).

Even the extent of studies measuring the impact of the wonderful effect is lacking compared to studies measuring the competent effect (which itself is likely a bias of the wonderful effect).

That reminds me of a bias study where the researchers were investigating the effects of stereotypes.

They told one group of Asian women that “women are worse at math than men” and another group of Asian women that “Asians are better at math than non-Asians”. Both are common stereotypes.

They then measured how well the two groups did on subsequent math exercises.

Interestingly, they found that the latter group (positive stereotype) performed better than those in the former group (negative stereotype).

As I recall, other research found similar results with Black males and golf scores when told “white people are better at golf” as opposed to “black people are better athletes”.

Not only do stereotypes influence our perceptions of “the other”, but they influence the performance of the other.

In other words, when society attaches a label to people, those people start behaving accordingly.

”You call me an addict and refuse prescription? Yeah sure whatever doc, I’ll just buy that fentanyl off the street since I’m already an addict.”

We need to end prohibition.

The experiment you discuss is called "stereotype threat" and it belongs into the huge set of psychological results that cannot be replicated. [1] It seems that there is a bias against publishing studies with null results, which skews the overall picture both among the researchers and in the media.

[1] https://www.tandfonline.com/doi/full/10.1080/23743603.2018.1...

Interesting, good to know, thanks.

That’s called “stereotype threat”, and it’s been caught up in the replication crisis. In short, the effect is hard to reproduce and tends to be small. It’s been known for some time now, here’s an article from 2015: https://www.psychologytoday.com/us/blog/rabble-rouser/201512...

I wish people would stop bringing up studies that don’t reproduce, they’re no better than anecdotes.

The problem with that is, at least in my experience, that you read such things once from a reputable source and never check it again.

When I was studying undergrad psych, Psychology Today was falling into disrepute.

Are they still a reputable source, today?

Isn't that one of those clickfarm blogs?

Psychology Today is not great in general, but the article I linked to is written by a social psychology PhD (Lee Jussim).

Ok, good to know, thanks!

Setting this aside for a deeper read, but it appears at first glance that the concerns revolve around statistical technique rather than methodological soundness?

It can't be "still reputable" after "falling into disrepute".

“was falling” in the imperfect tense.

> Our small study showed what you'd expect- students- men and women students rated male writers more highly than female writers, and initials were right in the middle, though they tended to be more like responses for men.

How much more? That's really important.

Yes, but also is that "what you'd expect", why?

Most things I read I have no idea of the sex of the author. Do people even look at author names before they start reading (online, say), and even then they can be nom de plumes, or non-gendered names (sometimes surprisingly).

We announced the name of the author. The script went something like "We're going to ask you to read this piece, written by J. Smith.".

During debriefing, we would tell the subjects what we were actually looking for (perceptions based on gender) and most reported not remembering/caring about the gender of the author, but nonetheless the results were clear that there was a gender bias, regardless of whether they reported remembering or caring about the gender of the author.

We didn't have a huge study, <40 participants, so whatever we found wasn't publishable, but I remember the results surprised me.

At the time, I felt like "Sexism is dead" and the university I was in was 2/3rds women to 1/3rd men, so we'd never find a bias.

As a young man (18/19) I learned to question my assumptions. I thought "It's the 90s, sexism is dead.", but it just wasn't true.

Sorry, but there are several things problematic about that study, if your description is accurate.

Just one of the issues is the specific perception regarding that particular sport (basketball) and relative perceptions between sex differentiated similar sports (men’s and women’s basketball are not the same sport, albeit similar). A sport should have been chosen where there are as few differentiating perceptions as possible, which is nearly impossible, because men are inherently more competent at sport due to physiological realities. It bakes in a bias towards men, which is what you likely actually confirmed with the study you described.

Introducing sports alone essentially corrupted that research and likely biased the responses for relative innate competency of males in sports.

For example, ignoring its own challenges, what could have been done, is show video of a relatively broadly positively evaluated sport (not basketball), an activity that is also relatively broadly and positively evaluated female clustered competencies, e.g., effective child rearing, communicating effectively, conflict resolution, creative outputs, or even beauty pageants, etc., along with the inverse, i.e., poor performance of men in the same sport and poor performance of women in whatever is chosen, e.g., screaming and yelling at poorly behaved children.

It is in the past and by no means are or were you the only one who has long engaged in this type of poorly executed “research” that has merely been confirming researcher biases and often, thereby immensely damaging society, but maybe my illustration of a few of the several issues with what you described, will cause some changes in thinking.

There is a real hidden epidemic of not only single order thinking, but what really should be called negative order thinking in research, where research is not only not finding the truth, but rather even doing damage through confidence in false findings.

I say these things with an expensive relevant background, in the trenches of bad and … frankly … destructive research, if you will.

Why are you ignoring the stronger non-sports-relates results that PP reported? Your bias seems stronger than the bias you are complaining about.

If it is known that most discoveries are made by males it is reasonable to give larger weight to new male-done research, if the only thing that is known about the researchers is their gender and if the reader is not enough of an expert on the subject matter to treat it 100% on its own merit.

I don't see this "bias" as inefficient or counter-productive. It is just an artifact of the way you designed your flawed experiment.

To find any real bias you would have to assure the subjects that the researchers were equally accomplished.

An experiment like OP's would 100% show bias. There weren't multiple researchers, each reader read the exact same paper, but with a different gendered (or non-gendered) name as the author.

> To find any real bias you would have to assure the subjects that the researchers were equally accomplished.

What do you mean? I think you didn't understand the experiment. Subjects were shown the same text with, at random, a male name, female name, or initials. These names were not real researchers, nor familiar to the subjects.

> If it is known that most discoveries are made by males it is reasonable to give larger weight to new male-done research

I'm sure you can see the two different fallacies in this short sentence:

- There have been more male researchers than female, therefore more "discoveries" were made by men than women. No shit. This does not mean male researchers are better than female, obviously.

- There are more men being researchers therefore men's research should be given more weight therefore there are more men researchers therefore...

You will find it beneficial to learn what a "prior probability distrubution" is.

It's actually you who doesn't understand probabilities. If a woman is twice as likely to make a discovery but women are only 5% of researchers, then it's both true that:

1. most discoveries are made by males

2. it is reasonable to give larger weight to new female-done research

contradicting you point.

But you pulled those number out of thin air. The fact of the matter is that if I pick one research paper at random out of all the papers written by men and I pick another research paper at random out of all the papers written by women the former will be of higher average quality.

Did you just pull that out of thin air?

You will find it beneficial to learn about "selection bias".

> If it is known that most discoveries are made by males it is reasonable to give larger weight to new male-done research

No it isn’t. Dont you see this is just circular reasoning?

Even if one sex was 10x as likely as the other to produce good research, it is still sexist bias to judge research on the basis of the author's sex. Fairness and meritocracy mean everyone gets a chance; no one should be stopped from succeeding because they belong to a low-performing group.

I think it's safe to say that this study can be treated as an anecdote because we don't know the effect size or exact methodology and it was never peer-reviewed. And I'd agree with your point if we were talking about the Nobel prize study, which evaluates people as individuals. But arguing that 'real bias' is bias that doesn't come from empirical evidence doesn't make a whole lot of sense- all bias comes from empirical evidence of varying quality.

the sports thing seems like a weird addition. did the text have some relevance to sport? if not it seems like the kind of thing that would skew results just by reminding the participants they’re in a weird study

Would be interesting to hear from OP, but I assumed it was a reminder to check ones stereotypes -- eg 'remember women can be good at traditionally male pursuits' (would work for some sports only).

It wasn't anything so complex as that.

The simpler version of the study had been done before, and the class I was a study design class. Our class assignment was to design, execute, and analyze a study.

We were simply combining two existing studies into a new study.

It bears mentioning- this was ~1996/1997 (I don't remember), I was a freshman, it was a single credit supplement, and we never published the results.

Can you elaborate on the sports thing? What effect did showing the sports had and do you have a thoory on the mechanism?

This is a really long time ago but if I remember, we ready a study about how sports effected perception. In the other study they did the male sport vs female sport thing, but not the name part, and so we just were combing two previous studies.

What p value did you get?

From a one credit course I did in 1997, I don't remember :)

It’s one thing to observe a pattern in a certain subset of people. However, why do you think there is this apparent bias or perception?

If there is a bias, is it good, bad, or neutral? Are there biases in other directions?

> If there is a bias, is it good, bad, or neutral? Are there biases in other directions?

Bias isn’t inherently good or bad - they are normative judgements of distance from rationality.

Some biases are probably evolutionary advantageous to the individual but harmful to the society or others.

Sticking with the group is great, except when it isn’t, for example.

Or, a “gut check” is great, but you can’t always make decisions based on your gut:

https://en.m.wikipedia.org/wiki/Affect_heuristic

The example described is the textbook definition of bias. Given the exact same text, the participants rate it higher if a male name is attached.

Maybe someone in your field would know, is there any difference in the actual reliability/reproducibility difference in research done by men and women? Maybe you are simply measuring a real phenomenon but in a round about way?

The OP mention it is the same text which is rated differently depending on the gender of the author.

This was one of the few factors we kept the same- same text.

I think if it had been a larger study (not a one credit class run by 4 undergrads but an actual study with funding) we might have had multiple texts and randomized them, but we had one text- I don't even remember what the topic was.