1. -4
  1.  

  2. 14

    The role and extent that sex plays in interest and behaviors is hardly established science. This is reminiscent of Tom Buchannan in The Great Gatsby cherrypicking scientific evidence to support his (very racist) claims and ignoring the evidence to the contrary. The gross lack of epistemological responsibility in the tech community is extremely concerning.

    1. 2

      The role and extent that sex plays in interest and behaviors is hardly established science.

      It is established, it’s just impolite to mention it: https://en.wikipedia.org/wiki/Sex_differences_in_intelligence#Researchers_in_favor_of_males_in_g_factor

      This is reminiscent of Tom Buchannan in The Great Gatsby cherrypicking scientific evidence to support his (very racist) claims and ignoring the evidence to the contrary.

      It wasn’t just the character that was embracing scientific racism, but the author himself - https://books.google.it/books?id=9lMbAgAAQBAJ&lpg=PT3975&ots=UfUIE6dKHW&pg=PT3975#v=onepage&q&f=false

      1. 1

        “There are differences” != “We know exactly how these differences manifest as well as their magnitude, while controlling for all reasonable non-biological factors”

        1. 2

          The logical progression here is:

          1. We bravely maintain that men and women are different
          2. We assume (1) implies that biology determines more men than women in programming not social factors
          3. From (2) we conclude men are better programmers than women
          4. From (3) we conclude diversity programs are left wing authoritarianism
          5. We congratulate ourselves on our bravery and declare any disagreement to be denial of science. QED.
        2. -2

          Your scientific illiteracy has been well documented .

        3. 2

          Ten points for the Tom Buchannan reference. People believe this crap because it is convenient. But the popularity of racist and sexist drivel shows both an appalling level of scientific illiteracy in tech and a social unwillingness to stand up for basic decency.

        4. 9

          Despite how it’s been portrayed, the memo was fair and factually accurate. Scientific studies have confirmed sex differences in the brain that lead to differences in our interests and behaviour.

          Hume’s Guillotine - just because something is a certain way in nature, doesn’t mean it ought to be that way.

          1. 2

            Hume’s Guillotine - just because something is a certain way in nature, doesn’t mean it ought to be that way.

            Are you sure you understand Hume’s criticism? He advocated for a clear separation of observation (“is”) and moral judgement (“ought to”).

            It makes no sense to ignore reality just because you think it “ought to” be different. Separate impassioned observation (descriptive) from passionate calls to action (prescriptive). It’s basic intellectual hygiene.

            1. 3

              I get what youre trying to say but as per Moore’s naturalistic fallacy you can’t derive a normative statement from a descritive statement, thats the point i was making. Im not advocating ignoring reality obv

              1. 2

                you can’t derive a normative statement from a descritive statement

                OK, but can you simply wish a change and expect it to happen? Can you just wish for sexual dimorphism to somehow skip the brain and expect to be taken seriously?

                1. 4

                  I don’t think anyone in their right mind would say sexual dimorphism skips the brain. Certainly there are psychological sexual differences, every feminist I’ve read states this plainly (Beauvoir, Irigaray, Paglia). The question is, taking into account what we know about sexual differences (descriptive), how do we construct ethical norms (normative)? From Moore’s and Hume’s work, we know that it does not follow from these psychological differences that we should separate the sexes into completely distinct or opposing roles; doing so would be to confuse “is” for “ought” (and it supposes a binary logic when actually sexuality follows something like a probabilistic logic). Instead, we need to have a societal discussion about what the parameters are of the sexual function in today’s modern society. Now, this is an extremely complex function, given that it’s basically being evaluated in a distributed manner, across all the individuals of our society and interacting with other societal structures such as marriage, workplace norms, familial norms, and so on. So anyway, this is where the discussion of diversity enters the picture, with all of this background already in place. Kapish?

                  1. 2

                    Can we at least use this reality to explain why differences in outcome don’t imply differences in opportunity?

                    In Italy we have laws stating that all administration boards of companies above a certain size must have at least a third of their members female. Governments brag about the percentage of women ministers. Some elections where you can write a preferred candidate from a list allow you to write in a second one, but only if it has the opposite sex of the first one.

                    Capisci?

                    1. 2

                      Can we at least use this reality to explain why differences in outcome don’t imply differences in opportunity?

                      Yes, we can do so as one part of the discussion; it is a part of the question “what is the biological and psychological origin for these differences?” But, you also have to ask things like:

                      • what are the institutional origins for these differences?
                      • are these origins just?
                      • who actually controls the institutions responsible for these differences?
                      • who is being affected most by these differences? (who benefits, who doesn’t?)
                      • what is their perspective on their opportunities and outcomes?
                    2. 0

                      Clearly this is not an argument involving logic: since no matter how many times it is pointed out to defenders of the google memo that everyone agrees there are gender differences, they cannot progress beyond claiming that the existence of differences proves them to be correct..

              2. 2

                You’re telling me brave new world wasn’t an instruction manual?

              3. 4

                I”m not sure where in the comment hierarchy below to put this…

                Seems to me that the point of studies is to gain predictive power–but individual humans are wildly unpredictable entities.

                Throughout my own life (a very narrow subset of experiences) I certainly haven’t failed to notice the averages: most people suck at technical thinking, period. Among those that don’t, most are male. This is less true among younger people, and less true on the campus of a liberal school, etc. There is variation, but the observation remains.

                So what is wrong with that observation? Two things.

                A. I’ve heard that when you tell a bunch of children that a test is hard and they aren’t expected to score well, they don’t. On average, ya know? (Let’s not go trying to predict the future actions of individual humans again…) So, the observation can have a detrimental effect on the young. I think!

                B. The observation about the average technical skill of women vs the average technical skill of men is not only not useful, I’ve found that it is actually wrong in practice. This observation only applies to groups, but technical skill is applied by individuals. That is, knowing the center of a distribution isn’t that helpful if the distribution is wide and isn’t even normal, which technical skill obviously isn’t. Succinctly: I’ve never worked with “women” in general, I’ve only worked with individual women and each of them was very skilled technically. (I’ve previously guessed that only truly brilliant women can survive the hiring bias, so those are the only type of female co-worker I’ve worked with: the brilliant ones.)

                Does any of that make sense? For those of you that want to apply this information about averages to your life or work life, does this influence your thinking?

                1. 0

                  Of course you can’t use statistical information about groups to predict something about individuals: https://en.wikipedia.org/wiki/Ecological_fallacy

                  What you can do with that information is explain why non-random samples do not mirror the composition of the general population and why a small average difference between two groups can translate in a big difference in the top or bottom 1%: https://lobste.rs/s/q9ewda/no_google_manifesto_isn_t_sexist_anti#c_5lqoxd

                2. 7

                  Perhaps it is all three of those things.

                  1. 4

                    It isn’t in line with the scientific community consensus if that’s what you’re implying.

                  2. 6

                    No the Google manifesto is a resentful political document. Attempts to claim it is scientific are dishonest hackery. This is how the document starts:

                    Google’s political bias has equated the freedom from offense with psychological safety, but shaming into silence is the antithesis of psychological safety. This silencing has created an ideological echo chamber where some ideas are too sacred to be honestly discussed. The lack of discussion fosters the most extreme and authoritarian elements of this ideology.

                    People who can claim such nonsense to be science don’t understand science or are being openly dishonest.

                    1. 2

                      At least some scientists find that memo pretty scientifically accurate (as already posted somewhere in this thread):

                      The author of the Google essay on issues related to diversity gets nearly all of the science and its implications exactly right.

                      For what it’s worth, I think that almost all of the Google memo’s empirical claims are scientifically accurate.

                      As a woman who’s worked in academia and within STEM, I didn’t find the memo offensive or sexist in the least. I found it to be a well thought out document, asking for greater tolerance for differences in opinion, and treating people as individuals instead of based on group membership.

                      1. -1

                        What a collection of worthless hacks.

                        1. 3

                          Jordan Peterson in his interview with author of that memo also thinks that the facts in it are essentially correct.

                          Can you clarify why do you call scientists quoted in my previous message “worthless hacks”?

                          1. 2

                            The manifesto begins with

                            Google’s political bias has equated the freedom from offense with psychological safety

                            None of that is at all concerned with science. It is just a hackneyed wingnut political opinion. Your first “scientist” has made a career of explaining that stereotypes are right and he wades right in by asserting that the manifesto’s tedious political assertions are scientifically valid. He’s a fake scientist attempting to promote a political ideology as if it were blessed by the science gods.

                    2. 7

                      This stuff is no better than Phrenology, a post-hoc rationalisation of pre-existing prejudice. It’s sad that the self-styled “smart guys” are falling into the oldest and dumbest of bias-traps.

                      1. 8

                        There’s a bit of a difference between Phrenology and the sorts of scientific studies carried out on this topic today. Further, everybody claiming there isn’t a difference is either a) ignorant of the state of current research or b) pushing a false equivalence to further their own agendas for what they believe is correct.

                        Scientists have voiced that the science is accurate, others have put up much more thoroughly sourced articles. The science isn’t in question, really, by people that actually know the science.

                        This leaves us with the policy conclusions. The conclusions of:

                        • “Hey, maybe we should see how we can better serve women”
                        • “Hey, maybe we should have greater transparency into the diversity black-box”
                        • “Hey, maybe we shouldn’t make conservative viewpoints verboten”

                        Without spiraling off into “but but but muh soggy knees”, any reader should be able to look at those conclusions and say “Those are pretty reasonable things to consider, we can have a productive discussion about them even if we disagree”.

                        The fact that we can’t do so, because people immediately zoom off into talking about institutional bias and pseudoscience and namecalling and so forth, should be alarming to you. It is alarming to many of us, but most of us are keeping quiet having seen both the attacks on that person and the tacit support in our industry for those attacks.

                        You’re going to get upvotes, of course, because the rot is clearly set in even in this community but that doesn’t make your statement factually accurate–not that that matter anymore in public discourse. ¯\_(ツ)_/¯

                        1. 6

                          Th e science isn’t in question, really, by people that actually know the science.

                          You’ve missed the point entirely. Even if there are consistently repeatable physiological differences (and spoiler warning, the science on this is absolutely not settled), why is that a good excuse for manifesto guy to denigrate his co-workers? Why is it Good Actually™ for him to use a 0.002% difference in spatial reasoning in a study from 1971 in a contest of domination over his peers?

                          Let’s put it a different way: there are major physiological differences between you and I, and anyone else in this thread. Does that mean some of us shouldn’t be in this industry?

                          And that’s the connection to Phrenology, the misapplication of science in service of naked prejudice.

                          There are also female engineers within google who are massively more skilled than manifesto guy. should they be made to feel unwelcome in their own profession because of the bizarre psyco-sexual prejudices of a junior engineer?

                          1. 2

                            why is that a good excuse for manifesto guy to denigrate his co-workers?

                            It wouldn’t be one, but that isn’t what he wrote. Please cite, with direct quotes, where he does so. Here, have a copy with the hyperlinks and figures intact.

                            Does that mean some of us shouldn’t be in this industry?

                            Of course not. Then again, that wasn’t the point the author made in his memo. Please cite where he makes that point.

                            […] should they be made to feel unwelcome in their own profession because of the bizarre psyco-sexual [sic] prejudices of a junior engineer?

                            I don’t know…if they choose to interpret that memo as unwelcoming (you know, instead of noting all the parts that say things like “hey, maybe we can reward pair programming and work to make the environment less stressful”) there’s not much to be done. Again, please cite the bits that you find unwelcoming and factually incorrect.

                            1. 6

                              His thesis is that google is too diverse. Or trying to be too diverse. Right? How can a company be too diverse unless some of the diverse people don’t belong?

                              Pair programming sounds great, but how does it work if there aren’t any women to pair with? If you divide up into pairs with a guy who codes and a woman who talks, you still need an equal number of women.

                              1. 0

                                That wasn’t his thesis, but I can’t tell if you’re trolling or not.

                                1. 1

                                  Then I guess I misunderstood it.

                          2. 6

                            “Strawman” is one of the most overused arguments online, but this is a classical case. Absolutely nobody claims there are no differences between men and women.

                            1. 5

                              Except of course, lots of people do claim there is no difference between men and women(’s brains):

                              One could easily be forgiven by a cursory scan of these articles to assume that is EXACTLY what is being claimed, not some strawman setup here, but a legitimate out in the world idea.

                              1. 2

                                Of course they do not. I read the first link you posted - did you read it? Here is the first paragraph.

                                The study, published in the journal PNAS, argues that if there were really such a thing as male and female brains, there wouldn’t be much overlap in the characteristics of the two—people would show either only male or only female characteristics. However, after examining the brains of 1400 people aged 13 to 85 years old in terms of their composition of gray matter, white matter, and connections, the researchers found that very few people were clustered on the extreme ends of the spectrum of features typically associated with males and females. Rather, there was a lot of overlap. *While some features were more common in female brains and others in males, most people have a mix of the two. *

                                People who don’t know any science or statistics are often confused by the difference between distributions of traits and absolute classifications. Is that your issue here?

                                1. 4

                                  My issue is you claimed the poster used a “strawman” (an intentionally misrepresented proposition that is set up because it is easier to defeat than an opponent’s real argument). I don’t believe it is one, I have been in the room / had it sincerely argued to me.

                                  That doesn’t mean I think it is at all accurate, it just means that it is the sincerely held belief of some, not a strawman. The fact that there are literally hundreds of articles about it and fierce debate around it I think validates this perspective.

                                  1. 3

                                    Really, I have yet to read a single claim that men and women are identical and nobody has been able to link to one. Probably someone believes it, there is a believer in anything, but this argument is not about whether men and women are identical, it is about policies to increase representation of women in engineering and management. Such policies are not based on the theory that there are no difference - the name “diversity” indicates a belief in differences. It is really annoying to see repeated citation or arm wave towards results that show there are differences between men and women as if such, totally uncontroversial information, had any bearing on the efficacy or value of diversity programs.

                              2. 1

                                Where in the post you are replying to does the word “strawman” appear? ctrl+f fails me.

                                Unless you are…you know…arguing a point I hadn’t posited there. What was the term for that again?

                                Also, note that the “difference” I’m referring to, in context, was about phrenology and contemporary research:

                                There’s a bit of a difference between Phrenology and the sorts of scientific studies carried out on this topic today. Further, everybody claiming there isn’t a difference is either a) ignorant of the state of current research or b) pushing a false equivalence to further their own agendas for what they believe is correct.

                                1. 4

                                  No I was pointing out that your argument uses a strawman: the imaginary people who argue that men and women are identical.

                              3. 0

                                “The science isn’t in question, really, by people that actually know the science”

                                And yet, the blog post by SSC you cite begins with an attempt to minimize the conclusions of a peer reviewed article by Hyde.

                            2. 4

                              Luckily some of the press starts to see it too - The Atlantic.

                              1. 3

                                The article should point to the original memo instead of Gizmodo’s version that was stripped of hyperlinks, images and relevant formatting.

                                Misrepresenting the source in a clickbait title is one thing, but Gizmodo crossed the line when it stripped it of references and edited it with this absurd note:

                                The text of the post is reproduced in full below, with some minor formatting modifications. Two charts and several hyperlinks are also omitted.

                                1. [Comment removed by author]

                                  1. 0

                                    Even if it were true that women, on average, were worse than men at engineering or whatever, the difference is small enough that it would be dwarfed by even a tiny amount of prejudice or sexism in hiring practices.

                                    Until you realise that a company like Google hires from the top 1% of the workforce. You know that small average difference? It’s enough to become a huge difference in the fringes.

                                    1. 0

                                      Your comprehensive ignorance of basic statistics is offensive. {2,4,6,101,103,105,107, 109,111, 400} 60% of these numbers are odd, 85% of numbers larger than 10 are odd, 100% of numbers over 111 are even. Amazing!

                                      1. 1

                                        Your comprehensive ignorance of basic statistics is offensive.

                                        I guess I need to thank you for motivating me to write my first R script just to show that “basic statistics” prove my point:

                                        #!/usr/bin/env Rscript
                                        
                                        sample_size <- 1000000 # the population is twice this
                                        mean_a <- 99
                                        mean_b <- 101
                                        top_percentage <- 0.01
                                        
                                        # generate 2 random normal distributions with a standard deviation of 5 sigma
                                        # and a slightly different mean
                                        a <- rnorm(sample_size, mean_a, 5)
                                        b <- rnorm(sample_size, mean_b, 5)
                                        
                                        # put them both in a data frame and label them accordingly
                                        all <- data.frame(value = c(a, b),
                                        		  label = c(rep("a", sample_size), rep("b", sample_size)))
                                        # get the top 1%
                                        top <- subset(all, value >= quantile(value, 1 - top_percentage))
                                        top_rows <- nrow(top)
                                        # now separate the top 1% by label
                                        top_a <- subset(top, label == "a")
                                        top_a_rows <- nrow(top_a)
                                        top_b <- subset(top, label == "b")
                                        top_b_rows <- nrow(top_b)
                                        # print the distribution of the top 1%
                                        message("top ", top_percentage * 100, "% of the population is made of ",
                                        	top_a_rows / top_rows * 100, "% 'a'(mean=", mean(a), ") and ",
                                        	top_b_rows / top_rows * 100, "% 'b'(mean=", mean(b), ")")
                                        

                                        Output of the script:

                                        top 1% of the population is made of 25.59% ‘a’(mean=98.9968034208468) and 74.41% ‘b’(mean=101.010232597726)

                                        So I got a more or less 1:3 ratio at the top just by making the mean values differ by 2 units. Q.E.D.

                                        1. -4

                                          It’s like when Otto tells Wanda he reads Socrates.

                                          1. 1

                                            Is this the extent of your “scientific literacy”? You can’t even argue against a simple statistical model you disagree with?

                                            1. -2

                                              I suck as a remedial teacher.

                                                1. -2

                                                  At last a relevant citation although you have the target backwards. You’ve made a ridiculous, but predictable, assumption that the claimed biological distribution in the general population is more pronounced in a sample of higher quality programmers. The claim for a biological basis for the imbalance in the general population of programmers is shaky, at best, but your added theory is pure fiction.

                                                  1. 2

                                                    a ridiculous, but predictable, assumption

                                                    You never got past those sorting algorithms, did you? You’re looking at well commented source code and you still don’t get it.

                                                    1. -2

                                                      You think I should have to debug your pathetic R code instead of pointing out you have no idea what you are talking about?

                                                      1. [Comment from banned user removed]

                                                        1. -1

                                                          I found the bug!!! You used a normal distribution.

                                                          I invoke that whole 10x programmers conversation.

                                                          Might as well give my stance on that: 10x programmers aren’t a myth, they are those which made it past some inflection point, allowing them to use their technical skill to gain more technical skill. After that point, they can ‘drive’ way out to the right of the graph and become major outliers… so called 10x programmers.

                                                          This phenomenon means your model is exceedingly naïve.

                                                          1. 1

                                                            This phenomenon means your model is exceedingly naïve.

                                                            Wait, you’re serious? How would the “long tail” of extraordinary programmers modify the distribution of programming skills in society?

                                                            My model is actually skewed towards the opposite of the point I’m trying to make. Not only are differences in average IQ by sex bigger than 2, but the distribution in males has a larger standard deviation than in females.

                                                            http://onlinelibrary.wiley.com.sci-hub.cc/doi/10.1348/000712605X53542/abstract

                                                            http://doi.org.ololo.sci-hub.cc/10.1016%2Fj.intell.2010.04.006

                                                            http://doi.org.ololo.sci-hub.cc/10.1016%2Fj.intell.2006.09.003

                                                            Now lets rerun the script with mean values of 97.5 and 102.5 and standard deviations of 6 and 6.7, respectively, to make the model closer to the measured data:

                                                            top 1% of the population is made of 4.07% ‘a’(mean=97.5071385480722) and 95.93% ‘b’(mean=102.496303410369)

                                                            Looks familiar?

                                                            1. 2

                                                              More word salad. You started with numbers about “interest” and are now citing studies on IQ tests and SAT tests. These are different things and can’t be substituted for each other and none can treated as an indication of programming talent without some evidence - evidence that is totally lacking in your posts.

                                                              Your second cite writes

                                                              . It is extremely likely that sociocultural factors played a role in the rapid decline from a 13.5 to 1 ratio in the early 1980s to a 4 to 1 ratio by the early 1990s in the top 0.01% of SAT-M scores.

                                                              Your third cite writes:

                                                              Of course, these sex differences in school and college subject choices are likely to be influenced by differences in original ability, or by socialisation, or by a combination of these and other factors. Thus, it will remain methodologically difficult to ascertain the nature and causes of any sex differences in abilities among adults

                                                              But it’s emotionally important to you to make the case.

                                                              1. [Comment from banned user removed]

                                                                1. 0

                                                                  Oh, “generic numbers”. Love those. But aside from your continuing inability to understand basic probability theory (such as what a random variable does) your flipping back and forth between “generic numbers”, SAT-M scores, and IQ as if not only were they all interchangeable but each one was a proxy for programming ability is genuinely pathetic. If one wants an explanation for the paucity of women in programming the oversupply of people desperate to lecture as you have is a sufficient. You make the Creation Museum seem like a scientific stronghold.

                                                                  1. 0

                                                                    You make the Creation Museum seem like a scientific stronghold.

                                                                    Excuse me, but am I your biological father? Why the fuck do you foam at the mouth every time you reply to me?

                                                                    1. [Comment removed by author]

                                                                      1. 1

                                                                        I assume it’s because you won’t stop measuring skull diameter.

                                                                        That was Cesare Lombroso. My thing is pointing out various elephants in the room.

                                                              2. 1

                                                                Gah, we’re six days into this.

                                                                Look, the neural networks inside women aren’t fundamentally different from the neural networks inside men.

                                                                If you wonder which definition of ‘fundamentally’ I’m using, it’s this one: A fancy multi-core 64-bit x86 CPU isn’t fundamentally different from a fancy multi-core 64-bit ARM CPU.[1]

                                                                Women can be Turing complete, too!

                                                                What do people who oppose this idea have to lose?

                                                                1. I’m no biologist, but I suspect as far as architecture goes, the difference between male and female human brains is vanishingly small compared to the difference between AARCH64 and AMD64.

                                                                1. 1

                                                                  Look, the neural networks inside women aren’t fundamentally different from the neural networks inside men.

                                                                  I’m not interested in explanations and mechanisms for sex differences in intelligence, at this point. I’m interested in observations of the distribution of intelligence and statistical mechanisms for an increased imbalance in the top and bottom percentiles.

                                                                  Am I making myself clear, can we get on some common ground, or would you like to go on preaching about shitty brain-computer analogies?

                                                              3. -1

                                                                Naive and post-hoc are different.

                                                                1. 1

                                                                  I’d never heard that term before. Thanks. Of course that characterization applies to my contribution to this “conversation”, too… :)

                                        2. 0

                                          This comment seems entirely off topic. The OP doesn’t mention job performance. The OP mentions job preference.