|
|
Post by bernythedolt on Jul 31, 2024 11:14:33 GMT
Whilst I understand what you mean and what you're suggesting here, strictly speaking a person's sexuality is a fixed discrete variable, not a continuous random variable, so it cannot be modelled by the Normal distribution, like say a person's height or weight, which are continuous variables. A person's height or weight can take any finite value anywhere along the real continuous line. A person's sexuality cannot, it's discrete data, it's categorical and cannot be broken into smaller parts. The Normal distribution is an excellent model for height or weight, but pretty useless and inappropriate for discrete data. You could present a bar chart/histogram (for example) of the numbers of people who fit into each discrete category (straight, gay, bi, etc, etc) and there are various non-parametric stats tests you can apply to such data, but the Normal distribution (and the stats that flow from it) is not one. So there's no concept of whether a person falls within 2 or 3 standard deviations of the distribution's mean value, for example, as has been suggested elsewhere. Heights yes, a random person will fall with x standard deviations of the population's mean height, but (in general) we can't apply that logic to categorised data. (There are a few exceptions like exam marks, with a discrete distribution which closely mimics a normal distribution, but that isn't the case here). In a mathematical sense, I would argue, like you and keitha , that being gay or trans (with a prevalence far below <10% of the population) is certainly not normal. In a philosophical/ethical sense, the question is far more nuanced and I'm not qualified to comment further! I would say there is some sort of normal distribution of extremely masculine males to quite feminine males and from extremely feminine females to quite masculine females, both in terms of looks and attitudes, whether that translates into potential sexual preferences I wouldn't know. Homosexual relationships were common in both Greek and Roman society so it's definitely not a new thing. I suspect there is some sort of distribution of sexual preferences with strongly heterosexuals in the centre, sexually curious/uncertain males and females to each side, then bi, then gay, then maybe trans. Whatever, it is good that people don't need to be afraid of being arrested for not being heterosexual and can live with whoever they like. To work this into a normal distribution, you'd need to be very clear what you mean by "to each side". With weight, for example, a person would fall above or below the population mean weight by a measurable figure, which would show the exact degree to which they differ from the mean, either above or below. For a normal distribution, you need to identify very clearly the random variable you are measuring. If you'd distinguish one side of the curve as males and the other side females, that's not like any normal distribution I've ever come across. What you're describing sounds very much like a histogram and there's nothing wrong with that. You just couldn't ascribe the usual normal distribution stats to it. For example, for a population's weight, you can say with great confidence that 68% of the population will fall within one standard deviation of that population's mean. You couldn't do that with the distribution you've described because it's by no means clear what variable you've measured, or how far from the mean is represented by one standard deviation. And I agree with your final sentence. What happened with Alan Turing, who was a hero to this country, was heartbreaking, to take just one example.
|
|
ethel
Posts: 153
Likes: 179
|
Post by ethel on Jul 31, 2024 11:44:55 GMT
But we are animals. Unless you think we were specially created by god and are different? That isn't the only exception but as above such a profound discussion warrants its own thread. Suffice to say I think you would both be wrong not to draw a distinction between human and the remaining animal kingdom. I think if we did behave more like animals, in some respects the world would be a much better place. Animals don't wage wars, torture, rape or maim, or destroy the environment, and kill only to eat. As a "higher" kind of animal humankind is a disappointment, but we have produced some great art, scientific discoveries, novels, films etc. Not sure that makes up for our destructive tendencies in any way though. The planet doesn't need us and would do better without us, as would the other animals.
|
|
michaelc
Member of DD Central
Say No To T.D.S.
Posts: 5,715 
Likes: 2,986
|
Post by michaelc on Jul 31, 2024 12:47:15 GMT
Whilst I understand what you mean and what you're suggesting here, strictly speaking a person's sexuality is a fixed discrete variable, not a continuous random variable, so it cannot be modelled by the Normal distribution, like say a person's height or weight, which are continuous variables. A person's height or weight can take any finite value anywhere along the real continuous line. A person's sexuality cannot, it's discrete data, it's categorical and cannot be broken into smaller parts. The Normal distribution is an excellent model for height or weight, but pretty useless and inappropriate for discrete data. You could present a bar chart/histogram (for example) of the numbers of people who fit into each discrete category (straight, gay, bi, etc, etc) and there are various non-parametric stats tests you can apply to such data, but the Normal distribution (and the stats that flow from it) is not one. So there's no concept of whether a person falls within 2 or 3 standard deviations of the distribution's mean value, for example, as has been suggested elsewhere. Heights yes, a random person will fall with x standard deviations of the population's mean height, but (in general) we can't apply that logic to categorised data. (There are a few exceptions like exam marks, with a discrete distribution which closely mimics a normal distribution, but that isn't the case here). In a mathematical sense, I would argue, like you and keitha , that being gay or trans (with a prevalence far below <10% of the population) is certainly not normal. In a philosophical/ethical sense, the question is far more nuanced and I'm not qualified to comment further! I would say there is some sort of normal distribution of extremely masculine males to quite feminine males and from extremely feminine females to quite masculine females, both in terms of looks and attitudes, whether that translates into potential sexual preferences I wouldn't know. Homosexual relationships were common in both Greek and Roman society so it's definitely not a new thing. I suspect there is some sort of distribution of sexual preferences with strongly heterosexuals in the centre, sexually curious/uncertain males and females to each side, then bi, then gay, then maybe trans. Whatever, it is good that people don't need to be afraid of being arrested for not being heterosexual and can live with whoever they like. And then ... ?
|
|
michaelc
Member of DD Central
Say No To T.D.S.
Posts: 5,715
Likes: 2,986
|
Post by michaelc on Jul 31, 2024 12:53:26 GMT
I would say there is some sort of normal distribution of extremely masculine males to quite feminine males and from extremely feminine females to quite masculine females, both in terms of looks and attitudes, whether that translates into potential sexual preferences I wouldn't know. Homosexual relationships were common in both Greek and Roman society so it's definitely not a new thing. I suspect there is some sort of distribution of sexual preferences with strongly heterosexuals in the centre, sexually curious/uncertain males and females to each side, then bi, then gay, then maybe trans. Whatever, it is good that people don't need to be afraid of being arrested for not being heterosexual and can live with whoever they like. To work this into a normal distribution, you'd need to be very clear what you mean by "to each side". With weight, for example, a person would fall above or below the population mean weight by a measurable figure, which would show the exact degree to which they differ from the mean, either above or below. For a normal distribution, you need to identify very clearly the random variable you are measuring. If you'd distinguish one side of the curve as males and the other side females, that's not like any normal distribution I've ever come across. What you're describing sounds very much like a histogram and there's nothing wrong with that. You just couldn't ascribe the usual normal distribution stats to it. For example, for a population's weight, you can say with great confidence that 68% of the population will fall within one standard deviation of that population's mean. You couldn't do that with the distribution you've described because it's by no means clear what variable you've measured, or how far from the mean is represented by one standard deviation. And I agree with your final sentence. What happened with Alan Turing, who was a hero to this country, was heartbreaking, to take just one example. I think that's right but doesn't it depend on whether you model the distribution as a continuous range (one for each sex) starting at so-called alpha male [female] and moving towards other personaly traits such as being camp. That is what Greenwood seems to be saying. In that case Normal distro seems ok to me ? On the other hand if you model being gay/trans/+ as discrete traits then a histogram is likely what you want. I could see it both ways (so to speak). To GW's other point, of course its right these things are legal. I just don't want it being actively encouraged particularly around young children. Edit: On reflection maybe you'd put the average joe, with average build, average testosterone in the centre (mean) and have one direction moving towards alpha male at the extreme on one side and "less and less alpha" going on the opposite direction. Seems highly contrived though.
|
|
michaelc
Member of DD Central
Say No To T.D.S.
Posts: 5,715
Likes: 2,986
|
Post by michaelc on Jul 31, 2024 13:03:52 GMT
We are going to get very philosophical very quickly and I'd rather not. it is hardly a philosophical question. Its a matter of biological classification as defined by ummmm humans. Living organisms are divided into five kingdoms: o animals (all multicellular animals) o plants (all green plants) o fungi (moulds, mushrooms, yeast) o protists (Amoeba, Chlorella and Plasmodium) o prokaryotes (bacteria, blue-green algae) Of course with the advent of social media, it has become more evident that some of us who identify as humans often exhibit behaviours and views more akin to a protist, but that is bye the bye. And even though I might like to think that occasionally I'm a Fun Guy, I'd be upset if you labelled me as a slime mould. You are an animal. Get over it. Better still, perhaps if more us could come to not just understand but feel how close we are to our closest relations in the animal kingdom, and realised what that means at a deeper level, perhaps we'd show a bit more empathy towards them and others. There you are: All your answers are here :  www.catholic.com/magazine/online-edition/do-you-know-what-separates-man-from-the-animalsAs I said before, create a new thread if you want to go into more philosophical detail.
|
|
|
|
Post by bracknellboy on Jul 31, 2024 13:46:50 GMT
it is hardly a philosophical question. Its a matter of biological classification as defined by ummmm humans. Living organisms are divided into five kingdoms: o animals (all multicellular animals) o plants (all green plants) o fungi (moulds, mushrooms, yeast) o protists (Amoeba, Chlorella and Plasmodium) o prokaryotes (bacteria, blue-green algae) Of course with the advent of social media, it has become more evident that some of us who identify as humans often exhibit behaviours and views more akin to a protist, but that is bye the bye. And even though I might like to think that occasionally I'm a Fun Guy, I'd be upset if you labelled me as a slime mould. You are an animal. Get over it. Better still, perhaps if more us could come to not just understand but feel how close we are to our closest relations in the animal kingdom, and realised what that means at a deeper level, perhaps we'd show a bit more empathy towards them and others. There you are: All your answers are here : www.catholic.com/magazine/online-edition/do-you-know-what-separates-man-from-the-animalsAs I said before, create a new thread if you want to go into more philosophical detail. Oh there are many things that distinguish us from other animals, even our closest relatives. If there weren't then we wouldn't be top of the pile on this planet. There are many excellent and thought provoking tomes on the subject. I would definitely recommend you research and read some of them. Although it won't surprise you to hear that I wouldn't count any derived from a theist basis among them. There are also of course many things that distinguish other higher order animals from lower order animals. But we are still all animals.
|
|
Greenwood2
Member of DD Central
Posts: 4,388
Likes: 2,787
|
Post by Greenwood2 on Jul 31, 2024 13:56:13 GMT
I would say there is some sort of normal distribution of extremely masculine males to quite feminine males and from extremely feminine females to quite masculine females, both in terms of looks and attitudes, whether that translates into potential sexual preferences I wouldn't know. Homosexual relationships were common in both Greek and Roman society so it's definitely not a new thing. I suspect there is some sort of distribution of sexual preferences with strongly heterosexuals in the centre, sexually curious/uncertain males and females to each side, then bi, then gay, then maybe trans. Whatever, it is good that people don't need to be afraid of being arrested for not being heterosexual and can live with whoever they like. To work this into a normal distribution, you'd need to be very clear what you mean by "to each side". With weight, for example, a person would fall above or below the population mean weight by a measurable figure, which would show the exact degree to which they differ from the mean, either above or below. For a normal distribution, you need to identify very clearly the random variable you are measuring. If you'd distinguish one side of the curve as males and the other side females, that's not like any normal distribution I've ever come across. What you're describing sounds very much like a histogram and there's nothing wrong with that. You just couldn't ascribe the usual normal distribution stats to it. For example, for a population's weight, you can say with great confidence that 68% of the population will fall within one standard deviation of that population's mean. You couldn't do that with the distribution you've described because it's by no means clear what variable you've measured, or how far from the mean is represented by one standard deviation. And I agree with your final sentence. What happened with Alan Turing, who was a hero to this country, was heartbreaking, to take just one example. I did say some sort of distribution, I deliberately avoided normal distribution. perhaps one for males and one for females, with heterosexual to the left and the tail of other preferences to the right. Yes probably better as a bar chart as it's difficult to get any in between options. Difficult to get any numbers really hence the I suspect.
|
|
Greenwood2
Member of DD Central
Posts: 4,388
Likes: 2,787
|
Post by Greenwood2 on Jul 31, 2024 13:57:14 GMT
I would say there is some sort of normal distribution of extremely masculine males to quite feminine males and from extremely feminine females to quite masculine females, both in terms of looks and attitudes, whether that translates into potential sexual preferences I wouldn't know. Homosexual relationships were common in both Greek and Roman society so it's definitely not a new thing. I suspect there is some sort of distribution of sexual preferences with strongly heterosexuals in the centre, sexually curious/uncertain males and females to each side, then bi, then gay, then maybe trans. Whatever, it is good that people don't need to be afraid of being arrested for not being heterosexual and can live with whoever they like. And then ... ? Indeed....?
|
|
|
|
Post by bernythedolt on Jul 31, 2024 16:09:44 GMT
To work this into a normal distribution, you'd need to be very clear what you mean by "to each side". With weight, for example, a person would fall above or below the population mean weight by a measurable figure, which would show the exact degree to which they differ from the mean, either above or below. For a normal distribution, you need to identify very clearly the random variable you are measuring. If you'd distinguish one side of the curve as males and the other side females, that's not like any normal distribution I've ever come across. What you're describing sounds very much like a histogram and there's nothing wrong with that. You just couldn't ascribe the usual normal distribution stats to it. For example, for a population's weight, you can say with great confidence that 68% of the population will fall within one standard deviation of that population's mean. You couldn't do that with the distribution you've described because it's by no means clear what variable you've measured, or how far from the mean is represented by one standard deviation. And I agree with your final sentence. What happened with Alan Turing, who was a hero to this country, was heartbreaking, to take just one example. I think that's right but doesn't it depend on whether you model the distribution as a continuous range (one for each sex) starting at so-called alpha male [female] and moving towards other personaly traits such as being camp. That is what Greenwood seems to be saying. In that case Normal distro seems ok to me ? On the other hand if you model being gay/trans/+ as discrete traits then a histogram is likely what you want. I could see it both ways (so to speak). To GW's other point, of course its right these things are legal. I just don't want it being actively encouraged particularly around young children. Edit: On reflection maybe you'd put the average joe, with average build, average testosterone in the centre (mean) and have one direction moving towards alpha male at the extreme on one side and "less and less alpha" going on the opposite direction. Seems highly contrived though. ...contrived, agreed, and I'm still unclear what random variable you'd be measuring. Remember the normal distribution is a probability function. Eg. the probability behind one's weight being less than or greater than the average. It's easily defined and measured, and maps beautifully onto the probability function best described by the normal distribution. How you'd measure the differences between alpha male, camp male, gay male, trans male, etc, is beyond me. One possible way might be to measure testosterone levels, which probably would map nicely to a normal distribution, but I'm not convinced that would lead to any meaningful result. Can we say testosterone levels somehow infer 'gayness' likelihood, I very much doubt it. I'm still not sure what random variable one would attempt to measure in order to map your data (whatever that is here!) to a normal distribution. You first of all need data, that's crucial for any distribution. If those data fall into counts in different boxes, eg.number of men camp, number of men gay, number of men alpha, etc, then that data set is in almost every case unsuited to a normal distribution. If the data set is measurable on the real continuous line, like height, weight, testosterone level, then it very often fits the normal distribution. (One exception is time-based data sets, which normally fit better the Poisson distribution, for example).
|
|
michaelc
Member of DD Central
Say No To T.D.S.
Posts: 5,715
Likes: 2,986
|
Post by michaelc on Jul 31, 2024 16:18:05 GMT
I think that's right but doesn't it depend on whether you model the distribution as a continuous range (one for each sex) starting at so-called alpha male [female] and moving towards other personaly traits such as being camp. That is what Greenwood seems to be saying. In that case Normal distro seems ok to me ? On the other hand if you model being gay/trans/+ as discrete traits then a histogram is likely what you want. I could see it both ways (so to speak). To GW's other point, of course its right these things are legal. I just don't want it being actively encouraged particularly around young children. Edit: On reflection maybe you'd put the average joe, with average build, average testosterone in the centre (mean) and have one direction moving towards alpha male at the extreme on one side and "less and less alpha" going on the opposite direction. Seems highly contrived though. ...contrived, agreed, and I'm still unclear what random variable you'd be measuring. Remember the normal distribution is a probability function. Eg. the probability behind one's weight being less than or greater than the average. It's easily defined and measured, and maps beautifully onto the probability function best described by the normal distribution. How you'd measure the differences between alpha male, camp male, gay male, trans male, etc, is beyond me. One possible way might be to measure testosterone levels, which probably would map nicely to a normal distribution, but I'm not convinced that would lead to any meaningful result. Can we say testosterone levels somehow infer 'gayness' likelihood, I very much doubt it. I'm still not sure what random variable one would attempt to measure in order to map your data (whatever that is here!) to a normal distribution. You first of all need data, that's crucial for any distribution. If those data fall into counts in different boxes, eg.number of men camp, number of men gay, number of men alpha, etc, then that data set is in almost every case unsuited to a normal distribution. If the data set is measurable on the real continuous line, like height, weight, testosterone level, then it very often fits the normal distribution. (One exception is time-based data sets, which normally fit better the Poisson distribution, for example). Well quite. It would be impossible just as it would to measure "attractiveness" or strength of smile. A panel rating system maybe? Anyway, this is getting a bit daft now. Oh whilst harder to map to a Normal distribution I'm not convinced a discrete rv is the answer either as I suspect some people might feel "a little bit" of one attribute whereas others might feel 100% that attribute.
|
|
|
|
Post by bernythedolt on Jul 31, 2024 16:18:58 GMT
To work this into a normal distribution, you'd need to be very clear what you mean by "to each side". With weight, for example, a person would fall above or below the population mean weight by a measurable figure, which would show the exact degree to which they differ from the mean, either above or below. For a normal distribution, you need to identify very clearly the random variable you are measuring. If you'd distinguish one side of the curve as males and the other side females, that's not like any normal distribution I've ever come across. What you're describing sounds very much like a histogram and there's nothing wrong with that. You just couldn't ascribe the usual normal distribution stats to it. For example, for a population's weight, you can say with great confidence that 68% of the population will fall within one standard deviation of that population's mean. You couldn't do that with the distribution you've described because it's by no means clear what variable you've measured, or how far from the mean is represented by one standard deviation. And I agree with your final sentence. What happened with Alan Turing, who was a hero to this country, was heartbreaking, to take just one example. I did say some sort of distribution, I deliberately avoided normal distribution. perhaps one for males and one for females, with heterosexual to the left and the tail of other preferences to the right. Yes probably better as a bar chart as it's difficult to get any in between options. Difficult to get any numbers really hence the I suspect. With respect, your words were "some sort of normal distribution". (Btw, there is only one sort of normal distribution - THE normal distribution  ). "difficult to get any in between options" - and that's exactly the point. It's a discrete data set, ie. formed from categories, while the normal distribution requires a continuous data set, ie. formed from a random variable moving along a continuous line.
|
|
adrianc
Member of DD Central
Posts: 10,037
Likes: 5,155
|
Post by adrianc on Jul 31, 2024 17:06:55 GMT
Mmmm. If they're all the answers, I think I must have misunderstood the question. The strongest argument against the existence of any kind of deity is the behaviour of their self-appointed salespeople. Especially down the Abrahamic timeline. Even leaving aside all the history of war and murder and conflict, quite how an organisation as mind-meltingly wealthy as the Catholic church can pontificate about poverty, when their own strictures against contraception are one of the main reasons many in the developing world are so poor...
|
|
Greenwood2
Member of DD Central
Posts: 4,388
Likes: 2,787
|
Post by Greenwood2 on Jul 31, 2024 18:29:39 GMT
I did say some sort of distribution, I deliberately avoided normal distribution. perhaps one for males and one for females, with heterosexual to the left and the tail of other preferences to the right. Yes probably better as a bar chart as it's difficult to get any in between options. Difficult to get any numbers really hence the I suspect. With respect, your words were "some sort of normal distribution". (Btw, there is only one sort of normal distribution - THE normal distribution ). "difficult to get any in between options" - and that's exactly the point. It's a discrete data set, ie. formed from categories, while the normal distribution requires a continuous data set, ie. formed from a random variable moving along a continuous line. I said some sort of normal distribution for extremely masculine men (alpha males if you like) to quite feminine men (and similarly for women), it might well not actually be THE normal distribution, might well be a skewed (normal) distribution, long time since I played with these things. For sexual preference I said I suspect some sort of distribution leaving out normal. All just spur of the moment musing really. And I agreed with you that a bar chart might be better for this. It would also be difficult to quantify how masculine a man is, but I'm sure some clever person might be able to figure out a way of determining it (possibly already have), personality testing?.
|
|
michaelc
Member of DD Central
Say No To T.D.S.
Posts: 5,715
Likes: 2,986
|
Post by michaelc on Jul 31, 2024 19:34:31 GMT
With respect, your words were "some sort of normal distribution". (Btw, there is only one sort of normal distribution - THE normal distribution ). "difficult to get any in between options" - and that's exactly the point. It's a discrete data set, ie. formed from categories, while the normal distribution requires a continuous data set, ie. formed from a random variable moving along a continuous line. I said some sort of normal distribution for extremely masculine men (alpha males if you like) to quite feminine men (and similarly for women), it might well not actually be THE normal distribution, might well be a skewed (normal) distribution, long time since I played with these things. For sexual preference I said I suspect some sort of distribution leaving out normal. All just spur of the moment musing really. And I agreed with you that a bar chart might be better for this. It would also be difficult to quantify how masculine a man is, but I'm sure some clever person might be able to figure out a way of determining it (possibly already have), personality testing?. Would you include the taboo, absolutely immoral and illegal in that distribution?
|
|
|
|
Post by bernythedolt on Jul 31, 2024 19:48:29 GMT
With respect, your words were "some sort of normal distribution". (Btw, there is only one sort of normal distribution - THE normal distribution ). "difficult to get any in between options" - and that's exactly the point. It's a discrete data set, ie. formed from categories, while the normal distribution requires a continuous data set, ie. formed from a random variable moving along a continuous line. I said some sort of normal distribution for extremely masculine men (alpha males if you like) to quite feminine men (and similarly for women), it might well not actually be THE normal distribution, might well be a skewed (normal) distribution, long time since I played with these things. For sexual preference I said I suspect some sort of distribution leaving out normal. All just spur of the moment musing really. And I agreed with you that a bar chart might be better for this. It would also be difficult to quantify how masculine a man is, but I'm sure some clever person might be able to figure out a way of determining it (possibly already have), personality testing?. Well there's something we definitely agree on! Final year of university nearly 40 years ago in my case. Surprised I can remember any of it TBH, given I can't remember what I had for breakfast some days...
|
|
).