> Is it sexism to make assumptions that are probabilistically correct?

Judging people based on group membership rather than individually on axis that are not essentially tied (even if they are correlated with) that group membership is prejudice; prejudiced based on sex is sexism. So, yes, making probabilistic assumptions based on sex and acting on them is, in many cases, sexism.

> ) For a woman I don't know, I buy her frozen yogurt. 2) For a man I don't know, I buy him pizza.

If you want to get people food that they want, you can just ask them. (And, in any case, context is probably more important than sex: IME, women are more likely to prefer pizza to froyo for a meal, and men are more likely to prefer froyo to pizza for a dessert; there're are very few contexts where "pizza" and "froyo" are comparable alternatives where sex would be the prime differentiator on how people would respond to them -- and elevating sex above more relevant decision criteria is a pretty strong indication of sexism.)

yummyfajitas is using the Pizza and Froyo examples from the post they're replying to. I don't think it is fair to grill them on the technicalities of those specific examples.

I wasn't grilling yummyfajitas over the choice of examples, I was supporting the point made in the post they were responding to (and rejected in their post) regarding the example.

> Is it sexism to make assumptions that are probabilistically correct?

Yes it's sexist, because assuming something about somebody based purely on their gender is precisely what sexism is. However, sometimes, in the face of limited data and limited resources to gather more data, being sexist might be the best possible course of action.

The problem with the stereotypes of gender is that you are dealing with huge numbers of people. Even if a well-conducted survey reveals that 99.9% of women prefer frozen yoghurt to pizza, then there's still around 3.5 million women in the world that prefer pizza to frozen yoghurt.

So the best possible way to decide whether to buy pizza or frozen yoghurt for a woman is to ask her. However, if you had to make a snap decision then, yes, you should go with the statistically more likely option and if you get it wrong then she should understand why you did it.

> ...assuming something about somebody based purely on their gender is precisely what sexism is.

I'm going to challenge that definition. There are absolutely biological difference between men and women. When I offer to help a young woman carry her suitcase up stairs, am I being sexist because I assume she would appreciate the help?

The definition of sexism needs to have room for some reality or the definition is useless or even harmful.

Well, think about whether you'd help a young man or leave him in the lurch, and think about whether you'd help if the woman in question is 6'1" and 190 pounds of muscle, handling the suitcase quite easily. Agreed that the definition needs to have some room for reality.

It may be a little old-school (which itself might be code for something else nefarious), but I will almost always stop for a woman [or the elderly] with a broken down car or flat tire on the side of the road (and have often doubled back from an exit a few miles down the road, then past the exit behind me, then returned if I couldn't safely stop as I initially passed). I will virtually never stop for a male in the same situation. (I have done so only when it's been dangerously cold out and in a fairly rural area.)

It comes up a couple times per year; more than half end in "I'm OK, AAA is on the way", the next most common is me changing a tire for them, and the rarest of all (nowadays) is them borrowing my phone to call for help.

If that's sexism, it's a type of sexism that I have no intention of stopping nor apologizing for.

Given the rampant assholism in the world, I have to be unconcerned with those people who believe I'm an asshole for doing this.

Just to give you a hint: How would you feel if you didn't know how to help yourself with something, and someone refused to help you because(!) you happen to be male? You might not have realised it, but that's in essence what you said, and that's at least one reason why people have been telling you that that behaviour is sexist. It's not about you helping women, it's about you deciding who to help based on their gender.

I'm not confused. I just don't agree that it makes me an asshole. I know it's differential behavior based on gender. That's why I thought it was interesting and on-topic to share.

For me, it's a safety issue. I'm pretty certain that I'm going to be able to handle 99.99% of encounters with roadside females (or elderly). I've gotten into an altercation with a roadside male I stopped to help in high school (who was likely high when he ran his car off the road).

I don't view it as particularly different from someone choosing to donate money or time to support the NAACP or NOW. Maybe donating to those organizations is racist and sexist, respectively? (Pedantically, it is, but I don't view those actions as negative or "asshole" either.)

> For me, it's a safety issue. I'm pretty certain that I'm going to be able to handle 99.99% of encounters with roadside females (or elderly). I've gotten into an altercation with a roadside male I stopped to help in high school (who was likely high when he ran his car off the road).

Well, the question is: Do you decide based on physical strength/behaviour of the individual or based on gender? Sexism is when you ignore the individual's characteristics and instead decide how to interact with them based on some supposed characteristic of the gender that they belong to. If you simply decide based on how strong or high the individual is (or appears to be), independently of their gender, and that then happens to correlate with their gender, that's not sexism--if you avoid helping a male despite them appearing completely non-threatening/-intimidating/whatever (and that's how I understood what you wrote--after all, there is nothing "old school" about avoiding situations that seem risky, is there?), then that is sexism.

I'm approaching the car at 30-70 mph from an oblique angle and trying to determine both the car and occupant status and whether I can safely get over and stop. I also wasn't expecting the situation, because I was out going about my business.

I admit that if the car was occupied by 6 elderly females, each with a pistol in one hand, a grenade in the other, and a deranged look in their eyes, that I wouldn't stop, but realistically, I'm making the call on occupant likely threat level to me primarily on gender and age, as if they're in the car, I can only see their shoulders and head.

In a good Samaritan situation like this, I have to be right every single time. One (more) mistake, and I can end up in a really bad spot and that's just not worth it to me or my family. For most false positive cases (where I drive past someone who posed no threat [which of course is most people on the side of the road]), they're going to get help inside of an hour or two from the police or AAA anyway.

It's worth noting that the Good Samaritan stopped for a man, and it's likely the Samaritan was running a real risk of an ambush and mugging. We're all free to choose not to stop, which is the premise of the story, but the point of the story is that stopping is the right thing to do.

Though for people who aren't Christians, it's just a convenient metaphor.

What percentage of drivers stop to help? I would wager it's well under 1%.

I also suspect that some people who judge me negatively for my selective stopping have never in their life stopped to help a fellow motorist of any gender.

Whether or not they understand Bayes, some groups of people are more concerned about what their posteriors will become proportional to whether they choose pizza and/or froyo.

In those cases, despite your information being generally probabilistically correct for your goal, the worst decision could be giving them something that they'll eat (and enjoy) or want to eat but can't. You're predicting for what to bring on a first encounter, and the best answer could be to bring nothing until you have better information.

Please, buy chocolate. Froyo is demeaning.

I know you're just doing your normal schtick of incorrectly assuming that everyone else is stupid (obviously no one, including blowski, thinks "to not be sexist, you need to ignore information and make worse decisions"), but this time you have accidentally stumbled on an interesting point. I know you don't care one way or another, but for the sake of other readers, I'd like to point it out(and BTW, your specific example (taken from blowski) isn't really sexist as I don't quite see how it keeps women away from power, but that's beside the point):

> I need to ignore information and make worse decisions?

Ignoring information and making worse decisions are two different things. Or, more precisely, by not ignoring some statistical data you are ignoring other pieces of information, such as our understanding of social dynamics. Ignoring yesterday's information about the train's arrival time would only lead to a much better decision making if you want to get to the train on time and you know that the train runs on a different schedule on weekends and yesterday was a Sunday. A piece of data is mostly useless (and even misleading) if you don't know what model it samples. Statistical data on human behavior is usually very bad information because history shows us that preferences change all the time (except for obvious things, like that people want power and are afraid of death). In fact, if women preference for frozen yogurt over pizza stayed constant throughout the ages it would be the exception rather than the rule.

If you base your decision only on statistical information of the kind you describe, you're creating a conservative feedback loop. We now know that people's assumptions about others' behavior may affect the other's behavior, and so behaving in accordance with past behavior (because that's what most of your information is really about) simply perpetuates the current condition.

The information you shouldn't ignore is the deeper information that comes from the actual study of social dynamics. That information tells us that changing your behavior to not conform to past behavior would (and constantly does) change society's future state. This is where social dynamics diverge from physics and the train schedule. We can decide to affect them (not necessarily determine them, because it's very hard to predict the behavior of such a complex system, but our actions definitely change the dynamics; we know this for a fact).

The only question is, which way do you want to direct society: do you want to keep it as it is for as long as possible, or change it according to your values?

I guess the conclusion to yummyfajitas' scenario then is to not give a woman frozen yogurt or pizza, but broccoli or spinach instead.

The conclusion is to try to study about society before trying to make unfounded assumptions such as "as in the past" = "greatest happiness".

I'll ignore your ad-hominem and your strange attempts to change the topic to your uncommon definition of sexism (as opposed to the original definition used by blowski, Pauline Leet, Caroline Bird, and everyone else here).

You are correct that it is possible that by giving people snacks they prefer, I'm affecting their future preferences in some way.

So what? All actions might possibly influence people's future preferences in some (probably unpredictable) manner. Why would I trade actual predictable present-day happiness for unpredictable changes in future behavior that might theoretically be somewhat better?

(If you disagree with me that influence on future preferences is unpredictable, and think you can predict social dynamics better than random chance, offer me a basket of bets at favorable odds.)

> your uncommon definition of sexism

My definition of sexism really is the common academic definition (although by "definition" I mean "definition plus corollaries", but it's a good enough approximation for people unfamiliar with research).

> Why would I trade actual predictable present-day happiness for unpredictable changes in future behavior that might theoretically be somewhat better?

Why wouldn't you? The question of what you should do is a philosophical question (perhaps the ultimate philosophical question) and depends largely on your values. I am not trying to dictate what you should or shouldn't do, simply show that your "optimal" decisions are not neutral ones based on "data" but ultimately a result of your assumptions and your (valued) goals. Achieving similar power distribution between men and women is a value of mine and it may not be yours, but you cannot justify your actions by anything other than another value.

> If you disagree with me that influence on future preferences is unpredictable, and think you can predict social dynamics better than random chance, offer me a basket of bets at favorable odds.

Well, my views of social dynamics are a result of years of study. I cannot say that my predictions are better than random because historians don't even try to make predictions (and I wouldn't dare call myself a historian), but we do know one thing for a fact: social action does indeed create social change. When it comes to complex systems, a familiarity with the dynamics does not translate to good predictions -- I'm sorry, but that's a direct result of the math. Non-linear, multivariate differential equations are very hard to predict (and we don't even have the right coefficients). But I know enough to recognize terrible reasoning that ignores what we do know, and denies its own assumptions.

I think I was pretty clear that my goal was to "make everyone happier". Obviously if you want to maximize paperclips you'll come to different conclusions.

I can't figure out what point you are trying to make. Best I can tell is that we should stop talking about bias against people due to gender and talk about your preferred topics instead?

Achieving similar power distribution between men and women is a value of mine...

It's an irrelevant one, however, since you can't actually predict whether any action will change this distribution.

As for your "years of study" that result in no testable predictions, I'm sorry for you. Perhaps it might have been a better use of time to spend a semester studying complex systems instead. I think I know what you are trying to claim (positive liapunov exponent + error m in measurement of x(0) -> error m e^Lt in prediction of x(t)), but the nonexistence of good predictions does not follow this - it merely changes the character of the predictions. Typically the predictions concern probability distributions on x, regions exhibiting certain behavior, phase transitions, etc.

While I could simply argue by authority, based on getting a Ph.D. in nonlinear multivariate differential equations, I'm making claims that can actually be intellectually defended. So feel free to exhibit some actual math and I'll carefully illustrate why you are wrong.

> I think I was pretty clear that my goal was to "make everyone happier"

How does your behavior achieve your goal, then? How do you know what would make people happy if all of your information about what makes them happy is snapshots of the past? Maybe a better study of the model would yield that the best way to make people happy is to offer them the exact opposite of what they had in the past? I don't understand how you can draw conclusions from data without understanding the model the data is a snapshot of.

> It's an irrelevant one, however, since you can't actually predict whether any action will change this distribution.

I didn't say that. I said that we can't predict with certainty and we're not trying to make predictions at all, but we do know from history that political and social action does work. That's how women were allowed to vote, go to universities, and become doctors and lawyers.

> that result in no testable predictions, I'm sorry for you

As I have spent even more years studying mathematics, I can tell you that neither subject was a waste of time, testable predictions or no. I feel sorry for you if testable predictions are the only measure of worth you know, but I hope that you are at least aware that much of the progress humankind has made wasn't due to any scientifically-predictable actions (and some of it was).

Nevertheless, I never said I can't make any testable predictions, only that we can't predict the course of history.

> Perhaps it might have been a better use of time to spend a semester studying complex systems instead.

Well, I did both.

> but the nonexistence of good predictions does not follow this - it merely changes the character of the predictions.

Really? If your measurement time (of initial conditions) significantly exceeds the time the predicted trajectory is within usable error, then yes -- nonexistence of predictions does follow. That doesn't mean, however, that we've seen that societies where social action was taken changed faster than those that didn't have social action. If nonlinear equations is your field then you know that, sadly, sometimes we are reduced to making qualitative predictions only.

No one disputes the necessity of models. Why do you think I did?

I didn't say that. I said that we can't predict with certainty...

Actually you said "I cannot say that my predictions are better than random". Either your predictions convey information (i.e., posterior(world state|prediction) != posterior(world state)) or they doesn't. If they don't, they are useless for prediction since E[utility|choice] cannot vary.

If they do, you can make the basket of predictions I asked for and behave better than random chance.

Really? If your measurement time (of initial conditions) significantly exceeds the time the predicted trajectory is within usable error, then yes -- nonexistence of predictions does follow.

The nonexistence of stable predictions of x(t) does indeed follow. Again, there are a huge number of quantitative predictions which can be made. For example, many systems with positive liapunov exponent allow claims like "a particle's long term distribution will behave like a sample from the distribution f(x) dx" or "a particle with energy < E_0 will travel a distance L only with probability < A exp(-BL)."

> Why do you think I did?

Because you said that ignoring people's past preference leads to sub-optimal decisions and implied that the best decision is merely to reflect past actions. You can't possibly make that claim unless you have a model that states that people's preferences don't change. That would be a very wrong model, though, because we know that they change and we know what may change them.

> Actually you said "I cannot say that my predictions are better than random".

By that I meant predictions of the course of history. I can say with high certainty that social movements often achieve their goals, and without them, those goals arrive much, much later. In particular, the feminist movement (on all its forms) has had tremendous success since the middle of the nineteenth century, and there is little reason to believe that the trend will stop. OTOH, the resistance to change has also taken a familiar path (while historians don't try to predict the future, they do recognize certain patterns) of an unofficial glass ceiling, then appeals to "nature" and "science", a better past and the tyranny of the reformer against a sieged hegemony. All of these are very familiar to us.

> a particle's long term distribution will behave like a sample from the distribution f(x) dx

Sure, we could theoretically make that claim, but we wouldn't be able to verify it because our measurement time exceeds the duration of any reasonable error of the prediction.

Listen, I went into history after studying math (yes, including non-linear equations), with delusions of being the first to come up with a dynamical model of society's behavior (with "power" of course, being a central quantity). But then I learned history and realized how unattainable that goal is. Maybe one day, if we have near instantaneous means of measurements. Otherwise, coming up with a model out of grossly inaccurate measurements made at different times -- and far too few of them (there are so many variables, so many forces, including forces of nature, namely a lot of external energy) -- all mediated by narrators, makes it impossible. We can say that "social movements that grow large enough tend to work", but putting that into any formula won't make the prediction any more accurate. I still have a dream of a semi-accurate model of how power flows in society, but it will require much better measurements.

BTW, if all our models are no better than random, the basis for relying on past behavior as a predictor of future happiness is no less shaky than any other. In that case, we'd have no reason to believe that ignoring past data would lead to any worse decisions than any other assumption.