Watching is better than doing

October 31, 2006 8:33 AM

Internet porn reduces rape. Violence on the silver screen reduces it in the streets.
posted by MonkeySaltedNuts (88 comments total)

"Violence on the silver screen reduces it in the streets."

Let's a spend a billion or so building movie theaters in Iraq.
posted by orthogonality at 8:37 AM on October 31, 2006

But don't blame the cats if you leave uncovered meat on teh Internets.
posted by orthogonality at 8:38 AM on October 31, 2006

The bottom line on these experiments is, "More Net access, less rape."

No offense, but that's bullshit. The study didn't show that pornography reduced rape, it showed that internet access, which correlates with money and a million other things reduce rape.

More money, less crime. Unless the authors of the study adjusted for income, it's not conclusive.

Studies like this are bullshit a lot of the time regardless of which direction they show things going.

It's as logical as saying "Most rapists are men, most porn customers are men, therefore porn causes rape!" I'm sure you could dig up a ton of studies that say things just as ludicrous.
posted by delmoi at 8:39 AM on October 31, 2006

If not Wikipedia, then what? Maybe rape is down because former rapists have found their true loves on Match.com. But professor Kendall points out that the effects are strongest among 15-year-old to 19-year-old perpetrators—the group least likely to use such dating services.

Wtf is this guy talking about? Has he heard of myspace, hotornot, facethejurry?

The author of the study is trying to claim that porn is the only explanation. It's stupid. It could be a number of things, maybe just reducing physical social contact. Rather then going on "dates" where girls can be date-raped, they just chat online where they can't be. There are an infinite number of possible explanations.
posted by delmoi at 8:42 AM on October 31, 2006

More money, less crime. Unless the authors of the study adjusted for income, it's not conclusive.

If you had read as far as the third paragraph of the linked article before posting, you would have seen:

And, according to Clemson professor Todd Kendall, the effects remain even after you control for all of the obvious confounding variables, such as alcohol consumption, police presence, poverty and unemployment rates, population density, and so forth. [emph. mine]

Which is not to say that there couldn't be other factors which the authors didn't take into account which lead to the observed result, but money does not seem to be one of them.
posted by DevilsAdvocate at 8:44 AM on October 31, 2006

Oops, I quoted the wrong thing. The "More Net access, less rape." thing may not be bullshit, but the "Porn reduces rape" thing probably is.
posted by delmoi at 8:45 AM on October 31, 2006

But professor Kendall points out that there is no similar effect of Internet access on homicide.

So delmoi, do you hypothesize that poverty has nothing to do with homicide rates?
posted by public at 8:46 AM on October 31, 2006

Maybe rape is down because the rapists are all indoors .. vandalizing Wikipedia.

I think that is the case.
posted by stbalbach at 8:49 AM on October 31, 2006

see also
posted by caddis at 8:52 AM on October 31, 2006

This actually sounds logical to me. I've always believed that people with violent tendencies are less likely to commit the act if they are given a chance to experience it vicariously.

I also realize that there are plenty of sickos that will want to experience it firsthand no matter what, but I'm betting that there are fewer of them than of the other type.
posted by tadellin at 8:52 AM on October 31, 2006

Weak.
posted by KirkJobSluder at 8:52 AM on October 31, 2006

I would like to take this opportunity to present a pet theory of mine, supported by absolutely nothing but anecdotal evidence, that in America before the internet, most kids were first exposed to sex in horror movies or violent action movies, which if you think about it, is psychologically pretty screwed up.
posted by Pastabagel at 8:54 AM on October 31, 2006

Maybe it's all that humpworthy prose.
posted by weapons-grade pandemonium at 8:54 AM on October 31, 2006

What about video games? Hopscotch? Contact sports? You're all thinking so small! Lets kill ourselves before we can hurt anyone!
posted by prostyle at 8:57 AM on October 31, 2006

I love how whenever one of these studies gets posted, some jerkass will, within 10 minutes or so, criticize a strawman version of the study. RTFA please. You're not clever just because you can make incorrect assumptions about the logical rigor of the FPP.
posted by rxrfrx at 9:01 AM on October 31, 2006

I love how whenever one of these studies gets posted, some jerkass will, within 10 minutes or so, criticize a strawman version of the study.

Please, the study is self strawmanizing.
posted by delmoi at 9:04 AM on October 31, 2006

I first read that as 'introduces rape', which I found flawed at best. I can see how this theory might hold water. maybe that's because I already thought of it, and it's always nice to have some reaffirming study come out to support your preconceptions. but hell, we are becoming the ultimate recluse society, no? rape? have to leave the home for that!

okay, back to reading...
posted by Busithoth at 9:06 AM on October 31, 2006

see also

Here's another.
posted by PeterMcDermott at 9:06 AM on October 31, 2006

I read this early today and considered posting it just to watch his logic get shredded by the MetaFilter sophistry meat-grinder.
posted by dead_ at 9:12 AM on October 31, 2006

Look, even thought they did correct for some economic factors, the simple fact is that showing internet access reduces rape is not the same thing as proving porn reduces rape. Not by a long shot. They authors dismiss a few possibilities out of hand "it can't be vandalizing wikipedia because wikipedia has nothing to do with sex!" and "It can’t be dating sites because people in the rapist demographic don’t use dating sites that much!" which is just a stupid response anyway.

Those aren't the only two explanations. Unless there is a credible theory as to why porno watching is the only possible cause then I still say the study is B.S.

Yes, I did miss that the factored for some economic factors, but I still say it's BS.

How many times do we hear people saying "Correlation != Causation!" I think they do it more then they should, (It would be more proper to say correlation does not imply causation).

The internet changes society a great amount, and an internet connected society has less rape and more porn, but that doesn't mean porn reduces rape.

Porn showing people in healthy relationships might reduce rape, but I would imagine that porn showing rape as normative might increase it. Or perhaps there is no effect at all. Since you can find any kind of porn you want online. I don't think internet porn is the cause. Rather I would say just spending more time in doors physically diffrent other people is the cause.
posted by delmoi at 9:15 AM on October 31, 2006

the study is self strawmanizing.

self-strawmanizing? is that a new kind of fetish porn? if it isn't it should be.
posted by jonmc at 9:15 AM on October 31, 2006

Any natural experiments that we can test? A study that uses a time series of internet penetration (heheheh... that's what they call it in the ICT literature!) in various countries would provide a quasi natural experiment...
posted by stratastar at 9:17 AM on October 31, 2006

Maybe it's just that without as much rape, horny guys have to go online and look at porn! Did you ever think of that? Huh? Huh?
posted by delmoi at 9:17 AM on October 31, 2006 [1 favorite]

I've never believed that violent entertainment turns people into Mad Max or Alex DeLarge, but I do suspect that the constantly repeated presentation of violence either as a difficult but necessary masculine responsibility or as a consequence-free carnival has a psychological effect on the viewers. I think these effects appear less in street rampages and school shootings, and more, say, at the ballot box. I can't think of any way to check this empirically, so I guess it will just have to remain a suspicion.
posted by stammer at 9:22 AM on October 31, 2006

Let's a spend a billion or so building movie theaters in Iraq.

I dunno, it seems to be working in India.
posted by GuyZero at 9:23 AM on October 31, 2006

Didn't we fisk this already? Man, I wanna write for Slate so I can toss off any ol' bullshit. Maybe I should get ahold of Wendell and ask him how he does it...
posted by klangklangston at 9:25 AM on October 31, 2006

I have an alternative way to look at it. What about this sequence?

1. Male viewer watches porn.
2. He feels compelled to spend "quality time" with himself.
3. He feels somewhat relaxed.
posted by jgbustos at 9:30 AM on October 31, 2006

I'm tired of being blamed for society's ills. I may induce lordosis and a rash of groupies, but it's all good clean fun.
posted by Humpable Prose at 9:33 AM on October 31, 2006

Wow. I think Pastabagel's theory may have some traction.

I would spculate that there was far more sexual violence BEFORE films. Given the traditional status and treatment of women.

Spending "quality time" with your self yeilds diminishing returns. Most sexual compulisves that trend to violence already masturbate all the time... how relaxed do they need to be?

Maybe it's the alledged vanishing testosterone we've been hearing about and not the pRon.
posted by tkchrist at 9:41 AM on October 31, 2006

Personally, I'm going to join the fight against rape right now.
posted by TheOnlyCoolTim at 9:46 AM on October 31, 2006

So, it sounds to me that this is research we can really ACT upon.

All we need to do is get a few million dollars from the treasury, and use them to fund about 20 more Saw and Hostel movies.

Then ship them to Gitmo.

And then we'll be in line with the geneva conventions again.
posted by InnocentBystander at 9:56 AM on October 31, 2006

Look, even thought they did correct for some economic factors, the simple fact is that showing internet access reduces rape is not the same thing as proving porn reduces rape.

Science isn't about "proving", and nowhere does the paper claim to "prove" anything. What you can do is strengthen a theory or weaken competing ones, by exposing it to contradiction through separate lines of testing. This process makes theories gradually weaker or stronger, as they survive this process.

He presents six original, fairly conducted, tests in support of his theory. Your knee-jerk nay-saying has been uniformly half-assed, wrong ("He didn't control for teh moneys!") and/or lame. You haven't presented any competing evidence or reasonable alternate interpretations of the data. You haven't even bothered to read the damn study, which is online.

Those aren't the only two explanations. Unless there is a credible theory as to why porno watching is the only possible cause then I still say the study is B.S.

This theory is discussed in the paper. If rape is, in part, a function of frustrated sexual or female dominating desires, and porn partially diminishes or satisfies these desires, then porn could reduce rape. There is some evidence for these separate premises.
posted by dgaicun at 10:01 AM on October 31, 2006

There are an infinite number of possible explanations.

"There are an infinite number of possible explanations" for any scientific theory. Evolution, etc. This fact alone says nothing about the validity or strength of any theory any more than the phrase "Things can be false" is a real challenge to the fact that cats and lions are related. You have provided no alternative data interpretations except that women go out on less dates b/c of Myspace and that states adopted the Internet in the order they did due to the demands of cock-blocked rapists. These theories are both obviously screwy. You haven't seriously attempted to think about this, you're just knee-jerking.
posted by dgaicun at 10:15 AM on October 31, 2006

This is really funny and Halloween-esque.
posted by kayalovesme at 10:18 AM on October 31, 2006

Media doesn't necessarily breed immediate and complete imitation; it has large and lasting effects, however. I'd say that pornography neither directly encourages nor directly discourages rape. It's more likely to directly encourage men to think of women in a relatively distant and idealized way. Maybe that leads to rape.

It's a mistake to think that the incidence of rape is a function of "sexual frustration." It ignores the whole dynamic behind most rapes; although "rape isn't about sex, it's about power" is an oversimplification, it points up the fact that rapists usually ignore the other sexual "outlets" around them in favor of a violent and controlling one. We can surmise from this that the thrill rapists crave is more about domination than it is about the mere experience of sex.

I'd also say that it's a mistake to assume that pornography assuages "sexual frustration." First of all, sex is not a need to be met, but a changing and turning impulse and a living part of us; we are encouraged to think of it as a need by those who create a product which purports to meet that need, but, between normal, real human beings, filled quotas are the dullest, deadest kinds of sex. Sex is not a product. Second, of course it's not in the interest of pornographers to assuage whatever frustration they tell us needs to be assuaged; if they did, they'd be out of a job.
posted by koeselitz at 10:22 AM on October 31, 2006 [1 favorite]

Wtf is this guy talking about? Has he heard of myspace, hotornot, facethejurry?

posted by delmoi at 8:42 AM PST on October 31 [+] [!]

Please don't pretend to be hip with something you obviously do not understand. While I know you yourself have a MySpace, you are not in the 15 to 19 age demographic nor do you understand this age group (or you wouldn't be so confused by the author's point).

None of the sites you listed are used as dating sites in the way that E-Harmony and Match.Com are used, especially the latter two. MySpace is a social networking (the other two don't fit into this category and shouldn't have even been mentioned)--it is used by highschoolers to primarily communicate with real friends and is not a site whose mission statement is to find you your one true love. Sure, some people do use MySpace to find dates, but those people are as a general rule not in this demographic.

Don't make things up to support your "argument" because you have trouble admitting when you're wrong.
posted by nonmerci at 10:22 AM on October 31, 2006

Just in case this study is correct, I'm going to increase my Internet porn viewing, and prevent even more rapes.
posted by Astro Zombie at 10:35 AM on October 31, 2006 [1 favorite]

Also, dgaicun: delmoi has a point. "What causes or reduces rape?" is a question for psychology, not statistical sociology. Compare, for example, the method in medicine: hundreds of studies are done every day, proving by correlation that eating eggs causes cancer, that eating eggs cures cancer, and a billion other things. Studies are not in and of themselves a useless tool, but they mean very little without physiological and scientific examination of the subject. I'll take a small set of psychologists' case-studies which carefully considers the causes involved over a study as broad as this one any day.
posted by koeselitz at 10:36 AM on October 31, 2006

It's a mistake to think that the incidence of rape is a function of "sexual frustration." It ignores the whole dynamic behind most rapes

It doesn't "ignore it", some scholars/evidence simply do not agree with this viewpoint:

Thirdly, a very controversial element of rhetoric and academic research on rape concerns the question of whether rape is driven by sexual desire or not. Sociological and feminist scholarship over the last 25 years has typically treated rape as a crime of violence or “power”, not lust. Under this theory, therefore, consensual and masturbatory sex are not substitutable for rape. However, this view has been criticized by other scholars on empirical grounds (Ellis and Beattie, 1983), and through the arguments of evolutionary biology (Thornhill and Palmer, 2000). The findings of this paper are consistent with the possibility that at least some part of the cause for rape is sexual in nature, though the results could also be consistent with the “power” theory if viewing pornography on the internet satisfies cravings for power that otherwise would be acted out as rape. (p 5)

Ellis, L., and C. Beattie (1983) “The Feminist Explanation for Rape: An Empirical Test” Journal of Sex Research v. 19, pp. 74-93.

Thornhill, R., and C.T. Palmer (2000) A Natural History of Rape: Biological Bases of Sexual Coercion, Cambridge, MA: MIT Press.

posted by dgaicun at 10:36 AM on October 31, 2006

stammer : I've never believed that violent entertainment turns people into Mad Max

Hey, I resemble that remark.

/stares coolly off into the post-apocalypse wasteland distance with my faithful dog sitting next to me. I shift gears and my supercharged V8 roars us off to another exciting adventure.
posted by quin at 10:40 AM on October 31, 2006

dgaicun: What you can do is strengthen a theory or weaken competing ones, by exposing it to contradiction through separate lines of testing. This process makes theories gradually weaker or stronger, as they survive this process.

Which is why scientific theories are not built and demolished by single papers. It's an interesting paper, but it alone does not establish a solid direct correlation between reported rapes for the age range, and access to pornography.
posted by KirkJobSluder at 10:40 AM on October 31, 2006

dgaicun: Personally, I don't think that you can create a single model that covers all rapes. It is quite possible that sexual frustration, and power, and drug/alcohol abuse are all factors that contribute to rape.
posted by KirkJobSluder at 10:43 AM on October 31, 2006

"What causes or reduces rape?" is a question for psychology, not statistical sociology

Ummm, rape has nothing to do with sociology?? No. Wrong. No. False. It's a question related to any number of scientific disciplines. And the paper cites psychology research (e.g. the infamous Thornhill and Palmer). Now this work from separate disciplines can reinforce each other.
posted by dgaicun at 10:46 AM on October 31, 2006

dgaicun: And the paper cites psychology research (e.g. the infamous Thornhill and Palmer).

That was a psychology paper? I thought it was a biology paper with some of the typical EV-Psych speculation added on for spice?
posted by KirkJobSluder at 10:51 AM on October 31, 2006

Bottom line: Slate sucks. So does Salon.
posted by supremefiction at 10:54 AM on October 31, 2006

Ech. Read what I wrote, dgaicun. Statistical sociology is a very, very imprecise field. It's not likely to produce a very compelling thesis on this question. The paper may cite psychologists, but it adds nothing to them but an example of correlation. And don't just quote the paper at me; I already read the damned thing. You might've noticed that I made some arguments against it.

Conversations go better when you read the comments you're responding to.
posted by koeselitz at 10:55 AM on October 31, 2006

Actually, I was wrong, Thornhill and Palmer published a book, which has much lower standards for evidence and review.
posted by KirkJobSluder at 11:01 AM on October 31, 2006

I read the article, and unfortunately, it does not include much information about the author's methods beyond what's in the tables included in the appendix (i.e., there is no separate "methods" section). I also did some back-of-the-envelope calculations, and it appears that the coefficient that measures the relationship between Internet usage and rape rates is not even statistically significant. If there are any statisticians here who can check my calculations, that would be great. Here's what I get:

95% Confidence Interval = Unstandardized Beta Coefficient +/- [(z-score for 95% CI)*(standard error)]

= -.730 +/- [1.96*2.86]

= -.730 +/- [5.6056]

= [-6.3356, 4.8756]

Based on the 95% confidence interval, the relationship between Internet use and rape rates could be a negative relationship as low as -6.3356, but it could also be a positive relationship as large as 4.8756. Or it could just as easily be zero. When I did these calculations, I thought that they can't be right. Am I missing something? If I'm right, it appears that the author of the "Internet porn" paper hasn't even done the basic job of disproving his null hypothesis.
posted by jonp72 at 11:10 AM on October 31, 2006 [1 favorite]

Please don't pretend to be hip with something you obviously do not understand. While I know you yourself have a MySpace, you are not in the 15 to 19 age demographic nor do you understand this age group (or you wouldn't be so confused by the author's point). -- nonmerci

Please, are you saying that you're somehow "hip" to the 15 to 19 demographic in a way that I'm not? What makes you think that's true?

None of the sites you listed are used as dating sites in the way that E-Harmony and Match.Com are used, especially the latter two. MySpace is a social networking (the other two don't fit into this category and shouldn't have even been mentioned)--it is used by highschoolers to primarily communicate with real friends and is not a site whose mission statement is to find you your one true love. Sure, some people do use MySpace to find dates, but those people are as a general rule not in this demographic.

First of all, let me say this the specific nature of the websites is irrelevant to my central argument which is that there are other reasonable explanations for the change. That said, I do want to defend myself, because I am right about those sites, and you are wrong, but I am going to make the font smaller.

<sub>First of all hotornot. When you say that hotornot does not say that it's mission is to "help people find their true love" you're wrong. They say that very thing on this page and they even have testimonials of people who did find their true love, or at least started dating. I also personally met girls in real life who I met via hotornot's meetme system years ago.

Face The Jury allows paid members to see the screen names of people if those are public. A friend of mine chatted up lots of girls off of FTJ and met at least a few of them. He also met girls on hotornot and myspace. That's how I know it's used, because this guy is always telling me about how he's meeting all these chicks online. He graduated at the same time as me, but he's about two years younger and the girls that he's interested in a few years younger still. So he's dated a lot of girls in that age group over the past seven or eight years.

As far as Myspace goes, I know at least two people who have gone out with people that they met by browsing profiles on myspace (which you can do, for 18+ people just like on a dating site). My above mentioned friend and a woman that I know. To say that it doesn't happen is false.

Of course myspace, hotornot, and FTJ are different then match.com. But they are certainly "dating" sites. They're not about finding "true love" they are more hookup oriented. But people do meet and they do date on those sites.</sub>

Now that that's out of the way, let me get back to my main complaint about this article, which most people here would probably agree with if it said that the internet (and therefore porn) contributed to excess rapes rather then reducing it.

"There are an infinite number of possible explanations" for any scientific theory. Evolution, etc. This fact alone says nothing about the validity or strength of any theory any more than the phrase "Things can be false" is a real challenge to the fact that cats and lions are related.--dgaicun

I think you're making a false comparison. When you do science, you try to find causations you want to be able to say "when this happens, then this happens." We know a huge amount about genetics and DNA, which is the result of tons and tons and tons of experiments. When we look at cat and lion DNA and say they are related, all we are saying is that when you look at their DNA it looks similar. That's something anyone can do with the right equipment. The rest of the theory is filled in with those thousands and thousands of experiments

All this study actually proves is that the internet reduces rape. Anyone can collect and run that data themselves, and the result should be the same. The study doesn't say why that is, they just guess that it's due to porn. there is no evidence to suggest that that the correct theory I'm not saying it's wrong, I'm saying there is no evidence. We can think of experiments that could prove or disprove this, by, for example actually analyzing how much porn people actually view rather then how much they have access too. And then seeing how many of those people rape people. That could provide some evidence that the theory is true, or disprove it.

but right now, all reasonable explanations are equally likely because there is no evidence

You have provided no alternative data interpretations except that women go out on less dates b/c of Myspace and that states adopted the Internet in the order they did due to the demands of cock-blocked rapists. These theories are both obviously screwy. You haven't seriously attempted to think about this, you're just knee-jerking.--dgaicun

People thought evolution was "screwy" when it came out, they were shocked to find out aether didn't exist. How "screwy" something seems is no indication of whether or not it's true. And just to clarify, this is my this is my specific theory:

when people spend less time physically in the same room with each other, rapes decrease. The internet causes people to spend less time together, therefore the internet reduces rape

That's my alternative theory. I'm not saying that alternative theory is true just that it exists and it is equally likely as the porn theory. In the article they gave theories that were designed to sound silly "wikipedia vandalism" or "they found their true love on match.com" but there also other theories that are not silly. That's the whole point, and that's why the conclusions about porn are ridiculous.

The reason I brought up hotornot/myspace is because I wanted to give examples of 'dating sites' used by people in younger demographics.
posted by delmoi at 11:17 AM on October 31, 2006

Based on the 95% confidence interval, the relationship between Internet use and rape rates could be a negative relationship as low as -6.3356, but it could also be a positive relationship as large as 4.8756. Or it could just as easily be zero. When I did these calculations, I thought that they can't be right. Am I missing something? If I'm right, it appears that the author of the "Internet porn" paper hasn't even done the basic job of disproving his null hypothesis.

Wow, that's terrible. I had just assumed that at least the math was right. What a bunch of crap.
posted by delmoi at 11:18 AM on October 31, 2006

The study assumes that the number of reported rapes is an accurate reflection of the actual number of incidents. I'm just saying. Fewer women reporting does not mean that the incidence of sexual assault has been reduced; it might just mean that the climate has become more hostile towards women who try to press charges.
posted by jokeefe at 11:25 AM on October 31, 2006

Internet is a demotivator, film at 11.
posted by Mitheral at 11:29 AM on October 31, 2006 [1 favorite]

A. This study is not a causal analysis, and so cannot, by definition, speak to whether there is a causal relationship.


B. On the other hand, those that say study is merely anecdote are mistaken: the analysis was based on a large sample and was adjusted for potential confounders.

C. Short of a randomized experiment, it would be difficult to prove conclusively that Internet porn caused, prevented, or was totally independent of rape. So happily we can all continue to live with our preconceptions and argue vociferously for the righteousness of our stance.
posted by Mental Wimp at 12:00 PM on October 31, 2006

Does this at least nullify the argument that porn causes rapes?
posted by SBMike at 12:09 PM on October 31, 2006

#SBMike:Does this at least nullify the argument that porn causes rapes?

The causal relation between porn and rape and that between watching violent fiction and being violent in real life has never been firmly established.

For example I have seen studies where higher watching of fictional violence correlates with actually being violent. Two interpretations are: 1) The viewing makes people more violent, and 2) people who are inherently violent are more likely to enjoy watching more violence than than less violent people.

I.e. whether watching violence is a cause or a symptom.

I'm sure the same can be said of many porno/rape studies.
posted by MonkeySaltedNuts at 12:28 PM on October 31, 2006

Wow, that's terrible. I had just assumed that at least the math was right.

No, you had just (originally) assumed that the paper was making a particularly stupid accusation without backing it up. You were quickly called out on this assumption. jonp72 actually did the work before deciding that the paper was wrong in its conclusion.

Thank you, jonp72.
posted by rxrfrx at 12:40 PM on October 31, 2006

For example I have seen studies where higher watching of fictional violence correlates with actually being violent. Two interpretations are: 1) The viewing makes people more violent, and 2) people who are inherently violent are more likely to enjoy watching more violence than than less violent people.


Yes, but this study seems to show at least a correlation between increased availability of porn and decreased rapes. Does this correlation disprove the hypothesis that increased porn consumption leads to more rapes? It seems to me that it does, but it's been a while since I took statistics.
posted by SBMike at 12:58 PM on October 31, 2006

This stuff certainly sheds doubt on the idea that more porn = more rape, and more movie violence = more real violence, if the math is right.

But there's been very little evidence for that idea, anyway-- other than the fulmination by many people about how horrible it is that "young people today" see ever-more-explicit sex and violence on screen and a few rather artificial experiments (Slate makes a good point about the lab experiments on porn being a bit preposterous if they don't allow completed masturbation).

If there is a relationship, it clearly is not a very strong one for most people. a far stronger link exists between being severely beaten as a child and becoming violent and between being sexually abused and becoming a child molester. If media has an impact, it probably is on those already at risk of offending and the direction of that risk is clearly not yet known.

also, one wonders whether passive viewing v. active participation as in video games v. films makes a difference.
posted by Maias at 1:11 PM on October 31, 2006

No, you had just (originally) assumed that the paper was making a particularly stupid accusation without backing it up. You were quickly called out on this assumption. jonp72 actually did the work before deciding that the paper was wrong in its conclusion.

Uh no, the point I made was that the paper did not do enough to prove or even show compelling evidence for the idea that porn decreases rape. The paper made two claims, one that I accepted (increased internet access correlates with decreased rapes) and one that I did not (because of that, porn reduces rape). I never claimed that the first claim was wrong, which is what jonp72 did.

I'm not exactly sure what I would have to do to "back it up" in your mind. It's not like you're even trying to back up anything that you've said in this thread. I also said one thing which was wrong, which I corrected. The fact that he did figure in poverty and unemployment does not fix the central problem with the second claim.
posted by delmoi at 1:12 PM on October 31, 2006

Please don't pretend to be hip with something you obviously do not understand. -- nonmerci

Please, are you saying that you're somehow "hip" to the 15 to 19 demographic in a way that I'm not? -- delmoi

FWIW, I'm 19 and I agree with nonmerci. Dating may occur on those sites, but in my experience, people in my demographic don't think of them as "somewhere to go when you're looking for a date" at all.
posted by booksandlibretti at 1:41 PM on October 31, 2006

While this is not a watertight scientific case, the evidence that porn reduces rape far outweighs the evidence that porn leads to rape.

The evidence for the latter is mostly based on the opinion of the sort of people that wish to impose their moral views on others. Some of them view pornography as an evil in itself, and the link to rape provides convenient leverage to convince others.
Whether this is true or not appears to be less important, since the "porn=rape" viewpoint has been around longer and is more loudly propounded, many believe it by default, see anchoring bias.
posted by spazzm at 2:34 PM on October 31, 2006

All of these technical conversations should be done in small text, it just looks more sophisticated
posted by InfidelZombie at 2:53 PM on October 31, 2006

Internet masturbators support dubious research showing that they're leading the fight against rape. Film at Nerdporn.com
posted by klangklangston at 3:06 PM on October 31, 2006

FWIW, I'm 19 and I agree with nonmerci. Dating may occur on those sites, but in my experience, people in my demographic don't think of them as "somewhere to go when you're looking for a date" at all.

Well, that's fair, but remember that you don't have the exact same perspective as everyone in your age group. Most people that age don't read metafilter.

It was just frustrating that someone was saying I was just "making stuff up" when in fact I knew of actual real people who had hooked up on all of those web pages. I don't think five years in age makes that much of a difference really.
posted by delmoi at 3:38 PM on October 31, 2006

Based on the 95% confidence interval, the relationship between Internet use and rape rates could be a negative relationship as low as -6.3356, but it could also be a positive relationship as large as 4.8756.

It's a good twenty-five years since I last took a statistics class, so I may well be getting things wrong here, but I thought that the whole point of a 95% confidence interval was that you could say with 95% of the time, you could be confident that the relationship as described will be replicated in the general population?

Also, I'm pretty sure that a 95% confidence interval is standard.

Surely there must be a real statistician out there who can enlighten us on this issue?
posted by PeterMcDermott at 5:23 PM on October 31, 2006

Ugh, delmoi, as much as I'd love to wade through paragraph after paragraph of your bullshit, I'm going to opt out. Let me just say this--you have missed the point. You do NOT understand the nature of sites like highschoolers, and why would I? Let me echo another's comment:

"FWIW, I'm 19 and I agree with nonmerci. Dating may occur on those sites, but in my experience, people in my demographic don't think of them as "somewhere to go when you're looking for a date" at all.
posted by booksandlibretti at 1:41 PM PST on October 31 [+] [!]"

Thanks, booksandlibretti.

Delmoi, you may be using MySpace to hook up with chicks (or make a pitiful attempt at doing so), but 15 - 19 year olds DO NOT. Cite what facethejury says all you want--no highschoolers or college-aged kids I know even use that site, and the same goes for hot or not.

I can't explain to you what you don't understand or will not try to understand. Your stubborness is inexcusable in this situation as you clearly don't know what you're talking about. Stop word-masturbating and stick with subjects you're familiar with.
posted by nonmerci at 5:54 PM on October 31, 2006

"You do NOT understand the nature of sites like highschoolers, and why would I?"

My bad, that's a really messy sentence. It should read "You do NOT understand the nature of these sites as highschoolers do, and why would you?"

Anyway, I'm not going to repeat the sentiment in my first comment. It still holds true. Sorry you have such difficulty grasping it.
posted by nonmerci at 5:55 PM on October 31, 2006

People thought evolution was "screwy" when it came out, they were shocked to find out aether didn't exist. How "screwy" something seems is no indication of whether or not it's true.

No, your alternative theories are screwy because they are glib, fatuous, false and stupid. Unlike the theory in the paper you are trying, with such a priori determination, to lamely "debunk". And certainly unlike evolution, for which Darwin made a very strong case, even in the age before any knowledge of genetics.

Order of state Internet adoption was due to largely nonmysterious factors, not to the horniness demands of a small percentage of would-be rapists. This "theory" can't be compared to evolution, it is incoherent bullshit.

And just to clarify, this is my this is my specific theory: when people spend less time physically in the same room with each other, rapes decrease. The internet causes people to spend less time together, therefore the internet reduces rape

And this theory doesn't meet its own predictions, unlike the one you are trying to replace. It doesn't explain why crimes of other types of interpersonal violence aren't reduced as well, only ones of a sexual nature.

Anyway, Happy Pornoween!
posted by dgaicun at 9:02 PM on October 31, 2006

remember that you don't have the exact same perspective as everyone in your age group. Most people that age don't read metafilter. -- delmoi

Well, yeah, and most of my friends don't read MeFi either. In fact, none of my friends do.

I go to a good college in a big city, and we all use the internet to do everything -- we often have to. My friends aren't computer experts at all, and many aren't even power users, but most of them are extremely familiar with how best to get stuff done on the internet. People generally have accounts on many different sites, even if they don't use each site to the maximum. This is the demographic I mean when I say those sites aren't considered for dating purposes. I'm also referencing my high school -- a rural, small public school where, again, nobody would consider those to be dating sites (but I graduated in 2004, so I suppose things could've changed there).

Also, nonmerci's myspace is linked from her profile, and that says she's 19, too.

I don't have a dog in this fight -- my only real opinions here are that internet porn is generally good and that rape is bad -- but I just wanted to let you know that hotornot/facethejury definitely aren't dating sites from this perspective.
posted by booksandlibretti at 9:51 PM on October 31, 2006

But professor Kendall points out that there is no similar effect of Internet access on homicide.

If we believe the main thesis, I guess we need more internet snuff.
posted by owhydididoit at 9:57 PM on October 31, 2006

I just know that I look at porn on the internet and I've never raped anyone. *shrug*
posted by Kickstart70 at 10:26 PM on October 31, 2006

Seems to me that the percentage of people with computers increased wildly at the same time as the internet availability did. The internet really sold lots of computers to people that previously didn't see a need for one.

The computer provides a lot of solitary entertainment. Reading all kinds of stuff off the net, playing games, maybe even learning to write some programs, keep a guy occupied. If he gets horny, instead of heading out, he can just grab some appealing content and 'spend quality time with himself' (I love that expression).

The scenario described applies to all, not only to the potential rapists, so one would expect some other effects on measurable activities. Fewer evenings spent away from home? Less money spent on other entertainment? Maybe people are out driving less?

I wanted to suppose that perhaps fewer people were getting hooked up IRL, however, I know that the internet has made hooking up a lot easier for many. I met my own partner of 9 years on the internet.

Another thing about internet and sexuality, the oppurtunity to talk about whatever kind of sex one can think of. This has got to be reducing the incidence of lonely perverts sitting at home growing sick and desperate. Instead they can learn early that they aren't that unusual, and how to deal with their desires in ways that don't harm others and/or land them in prison.
posted by Goofyy at 10:58 PM on October 31, 2006

Ugh, delmoi, as much as I'd love to wade through paragraph after paragraph of your bullshit, I'm going to opt out.

Why is this so personal to you? It's just bizarre. In your first comment you called me a liar, and when I explained that I was relating things that I had personal experience with you called it all "bullshit." You have your experiences, and I have had mine. I don't appreciate being called a liar. If you want to run around claiming that you know how every teenager in the world thinks and feels, and that the world is totally different then it was in 1999 go ahead I guess. I don't think it is.

Order of state Internet adoption was due to largely nonmysterious factors, not to the horniness demands of a small percentage of would-be rapists. This "theory" can't be compared to evolution, it is incoherent bullshit.

First of all you are the one who brought up evolution, secondly I didn't say that internet adoption was caused by horny rapists.

When two things are correlated, it means that either the first causes the second, or that the second causes the first or that some third thing causes both. It seems like you don't understand this. My argument was that the third thing (internet access) caused both increase in porn viewing and decrease in rape.

I don't have a dog in this fight -- my only real opinions here are that internet porn is generally good and that rape is bad --

I agree completely. I'm against rape, and I'm for Internet porn. I just thought this study was crap, but when I said so people just freaked out and started hurling personal insults at me it was truly bizarre.

but I just wanted to let you know that hotornot/facethejury definitely aren't dating sites from this perspective.

That's fine. People's experiences are different. I know most high school kids are probably more interested in dating other kids from their school. Thanks for the comments.
posted by delmoi at 1:00 AM on November 1, 2006

secondly I didn't say that internet adoption was caused by horny rapists.

Yes, you did. You suggested it as an alternative explanation for the data:

Maybe it's just that without as much rape, horny guys have to go online and look at porn! Did you ever think of that?

It's hard to tell if you were 'kidding', since your seemingly serious alternative explanations are just as makeshift and faulty.

When two things are correlated, it means that either the first causes the second, or that the second causes the first or that some third thing causes both. It seems like you don't understand this.

Oh, fuck you. I understand correlation fine, thank you, and I have revealed not one trace of ignorance of it here. In this case, we know that lower rape did not cause Internet adoption, and the paper actively controls for variables associated with differential state adoption of the Internet. Some such like state sales taxes have no compelling or theoretical relationship to rape.

Again, science isn't about "proving", but enduring falsifiability and competing paradigms. So far this theory has endured several novel tests that could have invalidated it, and that you simply ignore in your "criticisms" (for instance your alternative theory that people are simply more isolated to harm each other, ignores that it's only sex victimization that goes down, and not other kinds of interpersonal violence)

. I just thought this study was crap, but when I said so people just freaked out

You still haven't shown why this study is "crap"! It's this intellectual carelessness and obliviousness that "freaks people out" (whatever that means; you did note there is a discussion on Meta right now about how a user unrelated to this thread would like to block out your comments in particular!). Your first objection was that the paper was "bullshit" because it didn't control for income. When pointed out that you were wrong, you simply held out on your original assessment.

Anyway that there could still be other reasons for the association, is something both I and the author of the paper fully explain and allow. You think this is called "crap", but it's called science (which you don't understand). If you understood science, you woiuld understand why a study is not "crap", simply because it doesn't eliminate every other unspoken or unarticulated possibility forever and ever, amen. The important thing is that paper has used several tests to invalidate the obvious and/or major competing explanations for the association. It opened up its own hypothesis to several real instances of falsification - science. The significance of this is that you keep repeating alternative theories you like better, that the evidence in the paper already contradicts. This fact alone reveals the paper, while not "conclusive", is stronger than you realize or are willing to admit.
posted by dgaicun at 6:31 AM on November 1, 2006

By the way, delmoi, have you even bothered to actually read this "crap" study yet?

No?

Okay, didn't think so.
posted by dgaicun at 6:45 AM on November 1, 2006

dgaicun: So far this theory has endured several novel tests that could have invalidated it...

Such as...? I only see a link to two papers, of which one and only one addresses this theory.
posted by KirkJobSluder at 8:46 AM on November 1, 2006

I also noticed that nobody has yet discussed the economic theory that underlies the Internet porn paper. I am not an economist, but the era of Freakonomics forces a lot of us in the social sciences to be conversant with economic terminology and theory anyway. The main argument of the paper seems to be that Internet porn and rape are substitute goods. The basic idea behind the concept of a substitute good is that the two goods are not usually "purchased" together (e.g., if you're buying a CD, you're probably not buying an 8-track tape). An increase in the price of one substitute good (all things being equal, of course) will lead to an increase in the demand for the other substitute good. In other words, the basic idea of the paper is that the decrease in the cost of pornography, as a result of free downloadable Internet porn, makes rape more "costly" in relative terms to potential rapists. Thus, rape rates go down as a result of Internet porn. Unfortunately, I don't think the full economic implications of the theory were examined as deeply by the author as they could have been.

I'm using the Wikipedia entry for "substitute goods," so bear with me here if you have more substantive economic knowledge than I do. The entry distinguishes between perfect substitute goods and imperfect substitute goods. According to Wiki, "One good is a perfect substitute for another only if it can be used in exactly the same way, at exactly the same cost, and with exactly the same quality of outcome." Using the criteria in the Wiki definition, you can measure how "perfect" or "imperfect" two substitute goods may be. I think that, if you systematically go through these criteria, the "substitutability" of Internet porn and rape is much more imperfect than the author lets on.

The first reason that Internet porn and rape make poor substitute goods is that these two "goods" are not mutually exclusive, but occasionally "purchased" at different times by the same persons. (Please bear with me. I'm only using "good" as an economic term here, not a moral or ethical term.) We know that some married men rape their wives, even though they may engage in consensual sex with their spouses at other times. We know that some men who have no trouble obtaining consensual sex with desirable women still commit acquaintance rapes against other women. If consensual sex does not function as a "substitute good" vis-a-vis rape, then how can a lesser quality good such as pornography succeed as a "substitute good"? Even if you reject the feminist argument that rape is motivated by the desire for power not sex, the demand for sex among many men is so inelastic that a low-cost source of sexual stimulation will not reduce demand for more high-cost sources. In addition, rape may occur as a result of the production of pornography, even if rape is not caused by the consumption of pornography. For example, the most commonly cited argument against child pornography assumes that any increase in pornographic images of children also automatically leads to an increase in sexual assault against children, because otherwise those images could not be produced.

The second reason that Internet porn and rape make for a poor example of substitute goods is that the "cost" of rape can fluctuate just as much as the price of porn. If we ignore the problems with the math in this paper, the most interesting empirical finding is that Internet use is correlated with decreases in rape rates, but not with decreases in other crime rates. The author used several variables favored by criminologists to measure deterrence (e.g., concealed carry weapons laws, prisoners per population). The only problem is that the paper does not examine causal factors that might deter rape, but do not deter other crimes (e.g., women's self-defense classes, changes in evidentiary standards for convicting rapists). The decrease in rape rates may not be a result of a decline in the cost of porn, but due to an increase in the "cost" of committing rape.

The third reason that Internet porn and rape do not function as "substitute goods" is that they do not share "similar quality of outcome." Rape is a very "personal" crime for both its perpetrator and its victim, whereas pornography is by its nature impersonal. In this sense, prostitution would be more likely to function as a "substitute good" for rape, because prostitution is presumably less impersonal than staring at an inanimate pornographic image. Unfortunately for the author's argument, I believe previous attempts to show a negative correlation between prostitution and rape have failed (e.g., legal prostitution in Nevada has not given the state lower rape rates than states where prostitution is aggressively prosecuted).

I still think it's possible to find empirical data that show a negative correlation between Internet porn use and rape rates, but I think it would be more attributable to short-term opportunity costs (e.g., if you're busy using the Internet, you're not outside assaulting people) than because rape and Internet porn are substitute goods.
posted by jonp72 at 8:51 AM on November 1, 2006

jonp72, that was certainly superior to delmoi's farting. Although you reach the same problematic conclusion. Why would busy-time on the Internet reduce the time people spend raping but not hitting or stealing? It is the differential impact on sexual variables (not just rape), and only sexual variables that demands attention.

You make an interesting point about prostitution availability and rape, although I couldn't find your study on Google Scholar. Luckily though, while seeking it out, I did find the wider literature on porn availability and sex crimes. As it turns out papers from Europe and Asia reinforce the conclusion of this study, as does another recent one from the US using a whole different kind of data set:

Looking at increased porn availability in three separate European countries, this paper found either no change or decreases in rapes and sex crimes.

This paper from 1999 looked at the relationship between availability of porn and sex crimes in Japan, finding dramatic decreases in sexual victimization.

And this recent (PDF) economics paper, published last year, looked at the question by dividing regions of the US by PO box availability. Those places where this feature became available had increased rates of private Penthouse subscription through the mail. Controlling for population features the author found an increase in subscription resulted in lower numbers of forcible rapes.

It could be that there is no relationship, instead of the long-standing received wisdom that porn increases rape, or it could be that this newer body of psychological theory and empirical evidence is revealing the opposite to be true. It appears to be the best supported conclusion right now.
posted by dgaicun at 10:59 AM on November 1, 2006

Delmoi, you wrote:

"It would be more proper to say correlation does not imply causation."

No it wouldn't. It's precisely true that correlation implies causation but that correlation is not equivalent to causation.

Also, you wrote:

"When two things are correlated, it means that either the first causes the second, or that the second causes the first or that some third thing causes both."

You leave out the final possibility, which is that the two things are completely independent of each other and have no causal relationship whatsoever, no matter how indirect.

This is not unimportant because the universe of all datasets is much, much, much larger than the universe of all causal relationships¹. It's far more likely for a correlation to be a meaningless coincidence than it is to indicate a causal relationship. The only reason this doesn't mean that correlation is close to being worthless for discovering causal relationships is because we pick datasets to examine for correlation that we already suspect have some causal relationship.

The more we randomly "datamine" using larger and larger sets of arbitrarily selected data, the more we will find that correlations fail to predict causal relationships.

1. This is intuitively obvious. But it also has a more rigorous demonstration via relativity.
posted by Ethereal Bligh at 12:22 PM on November 1, 2006

No it wouldn't. It's precisely true that correlation implies causation but that correlation is not equivalent to causation.

I'm pretty sure you're wrong. If correlation implied causation then "!co ∧ ca &and co → ca" would be a true statement, in other words if one thing was caused (ca) by another, they might not be correlated (co). Then you could say that, say, smoking weed causes monkey attacks even if there was no correlation at all. Do you really think thats true? I guess if you think about it, there may be causes that aren't correlated very well with their results, but I kind of doubt they're very common. It can be difficult to mix pure logic with probability theory; in actuality correlation changes the probability of causation, rather then any sort of true/false relationship.

You leave out the final possibility, which is that the two things are completely independent of each other and have no causal relationship whatsoever, no matter how indirect.

You're totally right, that is a possibility. The larger your dataset is, the lower the probability that the correlation in the data is just a coincidence. But with a 95% confidence interval, 5% of the correlations discovered will just be coincidences. And you're also right that if you just go hunting around for correlations, you're much more likely to find coincidences then if you're testing a legitimate hypothesis, that's a big problem in Machine Learning.

And more then that, according to jonp72 they can't even claim it's true to the 95% confidence interval.
posted by delmoi at 4:28 PM on November 1, 2006

Yes, you did. You suggested it as an alternative explanation for the data:

Maybe it's just that without as much rape, horny guys have to go online and look at porn! Did you ever think of that?

It's hard to tell if you were 'kidding', since your seemingly serious alternative explanations are just as makeshift and faulty.

Oops, you're right, but yes I was kidding. I thought that was obvious. I suppose I should stop overestimating people's sense of humor.
posted by delmoi at 4:30 PM on November 1, 2006

Do I have to read all those footnotes to get the gist of this thread?
posted by UbuRoivas at 6:27 PM on November 1, 2006

Luckily though, while seeking it out, I did find the wider literature on porn availability and sex crimes. As it turns out papers from Europe and Asia reinforce the conclusion of this study, as does another recent one from the US using a whole different kind of data set:

Yes, thank you for mentioning this. This has generally been found to be true, at least since Berl Kutchinsky did some studies of the introduction of hard-core pornography into Denmark in the mid-60s. On the other hand, there isn't necessarily consensus about why this correlation arises, or whether there is a correlation at all. It may have nothing to do with pornography per se, but may because liberalization of pornography laws often occurs at the same time as liberalization of gender roles (such as in the U.S. "sexual revolution"). If that's the case, it may be the liberalization of gender roles that reduces sex crimes (because women are often the target of these crimes), whereas porn has no causal connection.

Also, dgaicun, I think you have detected a weakness in my off-the-cuff "opportunity cost" theory. The correlation of Internet use with lower rape rates, but not with reducations in other crime rates still needs to be explained. On the other hand, I still think author did a poor job of applying economic theory to the discussion of the causal relationship between rape and pornography.

If I had to come up with another theory to explain the pattern of correlations, I might come up with a behaviorist argument that Internet porn conditions Internet porn consumers to prefer "virtual" sexuality to real flesh-and-blood sexuality, whether that seexuality is consensual or non-consensual. That would be consistent with the pattern of correlations, but it still clashes with feminist theories that rape is motivated more by power than by sexual desire. My view is that this is a false dichotomy--that rape is motivated by the desire for power and sex--but I figure I might as well put those theoretical objections out there.
posted by jonp72 at 6:28 AM on November 2, 2006

I'm using the word implies in its normal nontechnical English sense, not its mathematical sense. For more clarity, replace it with suggests.
posted by Ethereal Bligh at 8:34 AM on November 2, 2006

Some correlations are causations does not mean that some causations are non-correlations. Nor does Some correlations are non-causations mean that, either.
posted by Ethereal Bligh at 9:19 AM on November 2, 2006

Mefites, mefites

I am a professor at a major research university and I teach causal inference to doctoral students. I have to say this discussion is off the rails.

If you want to discuss causal inference from observational data, then you MUST read Judea Pearl's book Causality to do so coherently. [He (Pearl) is the father of Daniel Pearl, the journalist who was beheaded by Al Quaida, but that's another story.] This book addresses in rigorous mathematics the calculus of this field. He will straighten out misconceptions such as the one that correlation can only come from direct causality or mutual causes (e.g., note that selection bias can create correlation between unrelated events). Without educating yourself first, you're just engaging in bloggal masturbation, which may or may not keep you from raping.
posted by Mental Wimp at 2:23 PM on November 2, 2006

I appreciate the recommendation and because of it I just ordered the book and I'll get it in a few days. But this discussion between myself and delmoi isn't technical and doesn't need such a treatment. If you'd like to point out something I've written that is untrue, I'd be happy to be corrected. As it is, there's hardly a topic in the philosophy of science as old as correlation and causation. It's not as if Pearl came up with the foundations. One can discuss this without having read him.
posted by Ethereal Bligh at 4:31 PM on November 2, 2006

If you want to discuss causal inference from observational data, then you MUST read Judea Pearl's book Causality to do so coherently.

Thanks for the recommendation. I'll add the book to my amazon queue.

I'm using the word implies in its normal nontechnical English sense, not its mathematical sense. For more clarity, replace it with suggests.

Hmm, well I meant it in a technical sense when I used earlier.
posted by delmoi at 8:04 AM on November 4, 2006