quote: quote:Maybe it's just because I hang out in Christian circles, but There are a whole lot of people that, if they are not lying, have waited until marraige.
Agreed. [Smile]
Agreed. In my youth, I hung out in mostly Christian circles and as I got older, learned that many if not most of them were lying. Or at least misrepresenting themselves. Imagine my disillusionment when I found out what exactly passed for "chaste", "pure", and "virginal".
quote:Originally posted by MrSquicky: Lyr, If you draw randomly from a population that has characteristics in some distribution, then, as your sample size increases, you become more and more likely to approach the true distribution in your sample, assuming that you have no selection bias. That is to say, it is possible that this sampling yielded a slice of the population with much different aspects from the general population, but it is very, very, very unlikely unless there was some sort of strong selection/response bias.
Well, first of all, they said right in the article that I read that the majority of respondents were women, so already it isn't 50/50 gender wise. That makes me wonder what else is different.
Second, let's assume you're right about the truth of the statistic. Like my previous post said, without asking more questions, that number doesn't tell us anything, and it doesn't help anything. Both sides of any particular argument regarding sex will be able to use it to support their point. It's about as helpful as an AOL poll.
Posts: 21898 | Registered: Nov 2004
| IP: Logged |
posted
Lyr- It doesn't matter if the responses are 50/50 genderwise. It just matters that you have a large enough sample of each.
Have you considered taking a basic statistics class? Since you're into this kind of thing, I think it would really help you.
Posts: 4655 | Registered: Jan 2002
| IP: Logged |
posted
It seemed kinda high to me, but I'm not shocked either. I hang out in slightly more conservative and religious atmospheres, so I'm sure there's more selection bias in my own sample than in the survey's. Also, I don't exactly go around asking my friends about their sexual habits.
Anyhow, I think the whole virginity thing might be a bit overemphasized. I don't consider it something to be proud of or ashamed of. When it comes to bases, I'm sitting in the dugout, but I don't really consider myself more "chaste" or "pure" than other people. I just happened to have not engaged in certain behaviors, due to lack of interest and opportunity. I don't know why people care about it that much.
Posts: 3546 | Registered: Jul 2002
| IP: Logged |
quote:Originally posted by blacwolve: Lyr- It doesn't matter if the responses are 50/50 genderwise. It just matters that you have a large enough sample of each.
Have you considered taking a basic statistics class? Since you're into this kind of thing, I think it would really help you.
A little quick off the trigger there.
If the gender imbalance is due to Non-response bias, then it is a legitimate concern.
quote:When it comes to bases, I'm sitting in the dugout, but I don't really consider myself more "chaste" or "pure" than other people.
This makes me wonder what the word "chaste" means to you.
I'd guess more along the lines of these definitions:
quote:4. free from obscenity; decent: chaste conversation. 5. undefiled or stainless: chaste, white snow.
Rather than these:
quote:1. refraining from sexual intercourse that is regarded as contrary to morality or religion; virtuous. 2. virgin. 3. not engaging in sexual relations; celibate.
posted
The 95% makes sense to me. I hang out with folks in their late 20s and early 30s. We're geeks. We play D&D. And I still don't know any virgins.
I know plenty of folks who don't get any very often. But I don't know any who never have.
Yes, having sex even once can be dangerous. But I doubt folks who can count on one hand the number of encounters they've had can really be much of a health risk. My buddy who's done it once ever doesn't seem to compare well to the guy I knew who considered it a badge of honor to rack up a high partner count over the weekend. That's just apples and oranges.
If they want to study the number of partners someone has and the age they started, why don't they try asking that instead? I have a buddy who didn't start until 23 and has far surpassed myself who started at 15. Ask the question you actually want to know the answer to.
Posts: 2283 | Registered: Dec 2003
| IP: Logged |
quote:I read in Cosmo that the average woman has 3 partners in her life.
I think Cosmo is a liar.
Actually, I think that's probably pretty close to right -- again, speaking anecdotally. There are the statistical extremes on either end, and there's the question of whether we're dealing with a median or a mean, but three rings relatively true to me.
Posts: 37449 | Registered: May 1999
| IP: Logged |
posted
I dunno. Even in my circle of relatively promiscuous friends, I can count on one hand the number of women who've had sex with more than five men. Now, admittedly at least two of them have had sex with more than twenty, so they do weird things to the average, but I wouldn't be at all surprised to learn that a slim majority of women sleep with just one or two men -- ever -- in their life.
Posts: 37449 | Registered: May 1999
| IP: Logged |
quote:When it comes to bases, I'm sitting in the dugout, but I don't really consider myself more "chaste" or "pure" than other people.
This makes me wonder what the word "chaste" means to you.
In the sexual sense, more of an active avoidance of sex, I suppose. Rather than just being an engineer .
Posts: 3546 | Registered: Jul 2002
| IP: Logged |
posted
I'm not a mathematician, so I'm having trouble with something. If the average women only has 3 partners, wouldn't that have to be relatively the same for men? Or am I apporaching this wrong? *puzzled*
Posts: 208 | Registered: Jun 2005
| IP: Logged |
quote:Originally posted by Hitoshi: I'm not a mathematician, so I'm having trouble with something. If the average women only has 3 partners, wouldn't that have to be relatively the same for men? Or am I apporaching this wrong? *puzzled*
It depends on what they mean by average. If they're talking about the mean, I think so, since the total number of sexual acts is the same on both sides. If they're talking about the median, not at all, since one person having sex with 146 people and four people having sex with one each will result in a dramatically different median than five people who all have had sex with 30 partners.
And this is all assuming we're talking about single-partner sex!
Edit: not sure how "30" didn't get in there. I distinctly remember typing it. Posts: 4313 | Registered: Sep 2004
| IP: Logged |
posted
Not necessarily, especially if we're dealing with median instead of mean.
Imagine a group of 10 women and 10 men. Let's say that 8 of the women have sex with nobody, but two of the women have sex with all 10 men.
In this case, the median number of partners for the men would be 2, while the median number of partners for the women would be 0.
Posts: 16551 | Registered: Feb 2003
| IP: Logged |
posted
I should've waited a few minutes to post, since mph's explanation was much easier to read than mine.
Posts: 4313 | Registered: Sep 2004
| IP: Logged |
quote:Originally posted by blacwolve: Lyr- It doesn't matter if the responses are 50/50 genderwise. It just matters that you have a large enough sample of each.
Have you considered taking a basic statistics class? Since you're into this kind of thing, I think it would really help you.
A little quick off the trigger there.
If the gender imbalance is due to Non-response bias, then it is a legitimate concern.
Non-response bias had already been talked about earlier in the thread. Everyone agreed that it might be a problem. Lyrhawn seemed to be saying that the poll couldn't be accurate because it wasn't 50/50 male/female. That's incorrect.
Posts: 4655 | Registered: Jan 2002
| IP: Logged |
posted
If the men in the poll where much more likely to ignore it except when they had lots of sexual experience to report it would skew the results. Something as simple as time of day could do the same thing. I do not trust sampling because what it is really good at is finding interesting unexpected strange attractors in samples, not revealing truth.
I know from interaction with young men that they have more and earlier sexual opportunities then I ever had, but it seems to be with a fraction of the girls who are promiscuous and they seem to be much more relaxed about it. Of course if I had sex in Jr high I would have been pretty relaxed about it as well.
The one trend I see that is strange to me is that the act of 'breaking in' a new participant, which used to be a practice that attracted certain men and boys has crossed gender lines and now there seems to be a significant number of older girls who seek to initiate the younger men. I know of one school where half the eighth grade boys were initiated by a single senior girl and her clique. It is significant because the numbers can get so high, a virgin girl is likely to be timid but a virgin boy is likely to be eager to the point of desperation. A handful of such girls or in one case a mother of school age girl can seriously change the sexual climate in small town for several years with effects for decades.
Posts: 475 | Registered: Aug 2006
| IP: Logged |
quote:Originally posted by blacwolve: Lyr- It doesn't matter if the responses are 50/50 genderwise. It just matters that you have a large enough sample of each.
Have you considered taking a basic statistics class? Since you're into this kind of thing, I think it would really help you.
A little quick off the trigger there.
If the gender imbalance is due to Non-response bias, then it is a legitimate concern.
Non-response bias had already been talked about earlier in the thread. Everyone agreed that it might be a problem. Lyrhawn seemed to be saying that the poll couldn't be accurate because it wasn't 50/50 male/female. That's incorrect.
I only scanned the thread, but it just seemed that Squickly and SR mentioned it and no one really responded, but I could be wrong. Although I'm sure that I would have noticed if *everyone* chimed in and agreed, because thats an awful lot of people.
In any case, it seemed to me that Lyrhawn suspected that the sample was unrepresentative of the population and cited the gender bias as evidence that there might be a bias, saying that if certain men didn't respond then there might be a reason why they didn't respond.
That seemed perfectly reasonable to me, and if *everyone* already agreed that nonresponse might be a problem, it might have been a little redundant, but hardly any cause to tell Lyr to take a basic statistics class.
Posts: 7593 | Registered: Sep 2006
| IP: Logged |
That seemed perfectly reasonable to me, and if *everyone* already agreed that nonresponse might be a problem, it might have been a little redundant, but hardly any cause to tell Lyr to take a basic statistics class.
I don't think suggesting that Lyr take a statistics class was meant at all meanly (ie "Why don't you just let the people who know what they're talking about discuss this, OK?") so much as she (he?) seems to have an interest in the subject and might enjoy learning more about it. I could be wrong, though.
Posts: 1547 | Registered: Jan 2004
| IP: Logged |
quote:So is postmarital sex.....divorcing your spouse....but still having sex with him/her?
No. Its having intimate relations with a fence you've promised to love, honor, and recieve tax benefits with unto death.
Posts: 11895 | Registered: Apr 2002
| IP: Logged |
posted
I think it's been shown in a number of studies that women, on average, when asked about their sexual history, lie on the low side. Men exaggerate the other way.
I suspect that if we took an average of male and female claims combined, we'd have an accurate number.
Posts: 5264 | Registered: Jul 2002
| IP: Logged |
posted
Personally, if I was 40 and still a virgin, I hope I'd've come to terms with it enough to be honest with it.
Posts: 3658 | Registered: Jan 2002
| IP: Logged |
quote:Originally posted by Human: Personally, if I was 40 and still a virgin, I hope I'd've come to terms with it enough to be honest with it.
Oh I agree with you there, but I know plenty of people who at least SAY they want to get laid for the sake of saying they did.
Posts: 14316 | Registered: Jul 2005
| IP: Logged |
quote:Originally posted by blacwolve: Lyrhawn seemed to be saying that the poll couldn't be accurate because it wasn't 50/50 male/female. That's incorrect.
I think Lyrhawn was saying the sample is already non-uniform, since if it were there would be a 50/50 gender split; he's wondering whether there are other sampling biases, such as religion, class, etc.
For instance, say Jedis make up 10% of the total population (they don't, but say they do). Say no Jedis engage in pre-marital sex (they do, but say they don't). Say the study didn't do a random sample across Jedi-ness, skewing the sample by only including 2% Jedi respondants. The 95% extrapolation to the broader population would be quite inaccurate (off by appr. 7.5%). The fact that the study did not do a random sample across gender makes one (particularly Lyrhawn, but also me) wonder what other characteristics they didn't take a random sample across.
Posts: 2926 | Registered: Sep 2005
| IP: Logged |
quote:The fact that the study did not do a random sample across gender
Who says they didn't do a random sample across gender? If you flip a coin 100 times, it probably won't be heads 50 times.
If they made sure it was a 50/50 gender split, that would not be a random sample.
But it's much more likely that it will come up heads fifty times than that it will come up heads 75 times; so if a coin came up heads 75 times you're best hypothesis (with some degree of uncertainty) would be that the coin was biased. I could calculate the actual uncertainty, given an hour of down time <edit>or 20 minutes; the standard deviation (a good measure of uncertainty) at 100 samples is about 0.0216; at 1000 it is about 0.0041), meaning the likelihood of getting anything other than a 50/50 split is vanishingly small for that sample size.</edit>. The uncertainty would go down significantly if you flipped a coin 1000 times and got 750 heads. This is a study with (IIRC) 44,000 participants, so the fact that it isn't 50/50 gender split indicates that it is very unlikely that it was a uniform sampling.
What actually happened (as far as gender sampling) is this (just went to the website and found out): they combined four surveys, taken in 1982, 1988, 1995, and 2002 and combined them. The final year was the only one to include male respondants in the sample group, resulting in a males being underrepresented in the sample set. I don't know if the study corrected for this by working on the gender subsets and then extrapolating to the larger population based on a true 50/50 gender split. That would be (I believe) the best practice, but it would do some funky things to your error bounds (because you're drawing statistics from two non-identically sampled populations).
That seemed perfectly reasonable to me, and if *everyone* already agreed that nonresponse might be a problem, it might have been a little redundant, but hardly any cause to tell Lyr to take a basic statistics class.
I don't think suggesting that Lyr take a statistics class was meant at all meanly (ie "Why don't you just let the people who know what they're talking about discuss this, OK?") so much as she (he?) seems to have an interest in the subject and might enjoy learning more about it. I could be wrong, though.
This was what I meant.
Posts: 4655 | Registered: Jan 2002
| IP: Logged |
posted
So I downloaded the data set (the 2002, I don't think I can download the older sets) and rummaged through it a bit. I haven't recreated any of the original experiment, but I did think these numbers might be of some interest:
Male percentages for 1st sex: married = 2.6%, unmarried = 80.8%, no 1st sex (virgin) = 16.6%
Female percentages for 1st sex: married = 7.6%, unmarried = 81.1%, no 1st sex (virgin) = 11.3%
So, of all 1st sex events in females, 91.5% were premarital. Of all 1st sex events in males, 96.7% were premarital.
Presuming there is no correlation between current virgin status and waiting until marriage, that would lead me to conclude (using an equal male/female waiting <edit>err, that's weighting</edit>) that appr. 94% of all 1st sex is premarital. However, I doubt the non-correlation assumption is true, and believe 94% is more of an upper bound. (81% would be the radical lower bound, where the assumption is that all current virgins wait until marriage to have 1st sex; 95.1% would be the corresponding radical upper bound that assumes no current virgins wait until marriage to have 1st sex).
These presumptions are, of course, only as good as the underlying sampling, which I've previously stated I have some concerns with (although fewer after reading a bit more of the methodology). Overall, this has been a pretty depressing exercise for me.
quote:Male percentages for 1st sex: married = 2.6%, unmarried = 80.8%, no 1st sex (virgin) = 16.6%
Female percentages for 1st sex: married = 7.6%, unmarried = 81.1%, no 1st sex (virgin) = 11.3%
So, of all 1st sex events in females, 91.5% were premarital. Of all 1st sex events in males, 96.7% were premarital.
I am I reading this right? The virgins are counted in the "having pre-marital sex" category? *raises eyebrow*
Posts: 26077 | Registered: Mar 2000
| IP: Logged |
posted
No, the 91.5% for female and 96.7% in males is not counting virgins at all. It's a coincidence that counting them as having sex before marriage leads to a combined male and female 95.1%. Or at least, that's how I read Senoj's post.
(Keep in mind, he's only using numbers from the most recent of the four polls.)
Posts: 1547 | Registered: Jan 2004
| IP: Logged |
posted
I think he may be excluding the virgins from the numbers and calculating the premarital from just the total of those that have had sex. Added: which creates the upper and lower bounds, and the number of virgins that have premarital sex will fall somewhere in the midde.
Posts: 1256 | Registered: May 2005
| IP: Logged |
quote:Originally posted by katharina: I am I reading this right? The virgins are counted in the "having pre-marital sex" category? *raises eyebrow*
I don't understand your question exactly; are you wondering about the original study's "having pre-marital sex" category? If so, I'll just repeat that I haven't tried to replicate the original study's methodology. However, if you assume that current virgins will participate in pre-marital sex at the same rate as non-virigins, the number still comes out to be 94% of all people's first sex is premarital. If you make a different assumption, say that no current-virgins will participate in pre-marital sex, the number comes out to 81%; and if you make the assumption that all current-virigins will participate in premarital sex, the number will be 95.1%
Posts: 2926 | Registered: Sep 2005
| IP: Logged |
posted
Wondering how the number could be so high when clearly debating statistical methods is so much more fun!
Posts: 11187 | Registered: Sep 2005
| IP: Logged |
quote:I think it's been shown in a number of studies that women, on average, when asked about their sexual history, lie on the low side. Men exaggerate the other way.
How could that be shown? If they lied about their history, how do you know?
Posts: 16551 | Registered: Feb 2003
| IP: Logged |
posted
Perhaps by doing studies where people are asked face to face and doing a separate study where people are asked completely anonymously on paper and comparing the results? I don't know, but I thin it may be possible.
Posts: 1547 | Registered: Jan 2004
| IP: Logged |
posted
Why don't they just do the poll completely anonymously in the first place then, if they get more accurate results?
Posts: 16551 | Registered: Feb 2003
| IP: Logged |
quote:I think it's been shown in a number of studies that women, on average, when asked about their sexual history, lie on the low side. Men exaggerate the other way.
How could that be shown? If they lied about their history, how do you know?
Trivial. Every sex act by a man with a woman involves a sex act by a woman with a man. If men and women report different total numbers of sex acts, something is wrong. To be sure, you'd have to compensate for out-of-population effects: Maybe all the boys are having sex with one girl whom you didn't happen to poll, and if you got that one in the numbers would match up again. But I'm going to wave my hands around that problem because I used the word 'trivial' and then came up with some pretty nice elegant math, and therefore any real-world problems with the theory are just a conspiracy by sociologists and suchlike scum; and besides, for an effect the size we speak of here, sheesh, that's one really skanky girl, as in, porn-star-gangathon-every-day skanky.
quote:Why don't they just do the poll completely anonymously in the first place then, if they get more accurate results?
Because the question you asked was "How do you show that anonymous polls are more accurate", for which you do actually need to do at least one non-anonymous poll. Duh.
Posts: 10645 | Registered: Jul 2004
| IP: Logged |
posted
In one test I saw (there were a few to choose from), Port, there were fake lie detectors involved. Women answered the questions anonymously, then were asked again, thinking they were hooked up to a lie detector.
The jump in that particular test was from an average of, I think, 2.6 "anonymously" to 4.4 on the fake lie detector.
Posts: 5264 | Registered: Jul 2002
| IP: Logged |
posted
That's really interesting. I wonder why people would feel the need to od that (lie in an anonymous poll). Are they trying to represent their group in a more favorable light?
Posts: 16551 | Registered: Feb 2003
| IP: Logged |
Printer-friendly view of this topic
![[Razz]](tongue.gif)
![[Wink]](wink.gif)
.
![[Confused]](confused.gif)
![[Dont Know]](graemlins/dontknow.gif)
I could be wrong, though.
