Think critically when looking at statistics (A valuable lesson, now with porn!)

Today, I finally caught that infographic on Internet porn facts that's been going around like a bad itch in a delicate place. I'm a bit skeptical about how accurate it is—if you scroll to the bottom you will see that the number one source is an article in the Daily Mail. But, it did make me think about how a single soundbite statistic doesn't necessarily tell you everything you need to know to understand the big picture.

Case in point, according to said infographic, Utah has the United States' highest online porn subscription rate per thousand home broadband users. My first thought: "Heh, dirty Mormons."

But wait! The fact doesn't actually tell us that the good people of Utah consume more porn than, say, Californians. It merely tells us that Utahans are more likely to pay a regular subscription for their porn. Sure, this could be an example of religious hypocrisy caught red-handed. (Ahem.) But, it might also be something far funnier: People who are so clean-cut and law-abiding that they don't even feel comfortable illegally downloading their porn. (Or, it could just be that the citizens of Utah are more likely to not be terribly Internet savvy, in which case, as a public service, I'd like to take a moment inform them all the RedTube exists.)

The point: We have a statistic. But we don't have enough information to know what that statistic really means.

Image courtesy Flickr user fixe, via CC

• [Anonymous]

  I realize this will not be a popular opinion on boingboing, but tube sites like redtube depend on stolen videos and the “user submitted” loophole in the DMCA for their business model, and the whole porn industry which employs hundreds of thousands of people in this country is suffering hugely because of the stolen content model of the tubes, which are all run by some of the shadiest business owners on the planet.

  One of the biggest tube owners, the owner of redtube in fact, if I remember correctly, is under multiple investigations for money laundering, as an example.

  Thousands of mom-and-pop porn sites are being driven out of business by the tubes, because it’s too expensive to constantly chase the tubes with takedown requests, when tube employees with dummy “user” accounts immediately reupload the stolen content.

  All of you hoighty toity netkultur types may believe that when a person gets naked and gets banged on camera to try to make a living you have a god-given right to steal that infinitely-reproducible piece of created work, but you are still stealing it.

  • [failix]

    Oh this is just pure comedy gold!

    Btw, it’s copyright infringement, not theft.

  • [SteveT]

    Fair point.
    I just wish you hadn’t used the words “mom-and-pop porn sites”. Apart from the pun, the thought of my parents web-casting home-made porn is somewhat icky!

• [Anonymous]

  Actually, the characterization of Utah as rural or technologically backwards is also a common misconception. The University of Utah was one of the four founding hubs of ARPAnet, the predecessor to the internet, and Utah has one of the highest percentages (by household) of internet use/access.

  I remember when I was in highschool in the early to mid nineties, the national guard laid cabling for all the high schools in the area to insure we had the highest speed of internet connection then available (and no, I was not in what one could call an even REMOTELY rich district).

  I moved to tacoma in 2005 to teach high school and was quite shocked at how limited the online technology and availability was in schools here. We are just now, in 2010, getting to the point my poor Utah high school was at 15 years ago.

• [Anonymous]

  Meant to include (yet another) infographic to support my point on that last post.

  Broadband coverage by state: http://coronainsights.com/2008/09/who-uses-the-internet-part-1-state-by-state/

  Note Utah falls into the top five in multiple categories.

• [Anonymous]

  To say nothing of the selection problem. The statistic referenced broadband internet subscribers. If broadband penetration (heh) is lower in Utah than average, one might naturally expect that those willing to pay for broadband are – ahem – heavy users of certain types of online video.

  So, I suspect it’s not that falks from Utah are a particularly horny lot, it’s that they have lower overall broadband coverage and that partakers of paid porn are disproportionately represented among the broadband users.

• [foobar]

  There are interesting things we can infer from these statistics. For example, if 70% of men aged 18-24 view porn in a typical month, we then know that 29.9% of men aged 18-24 are liars.

• [Rev. Benjamin]

  Very good point Maggie. We have the ‘data’, but what does the data or lack-of-data say?

• [halfsmart]

  or, for example, physical porn (eg magazines and DVDs) could be less available in Utah, forcing non-Mormons to resort to the internet for their wanking.

  • [ink]

    I live in Utah, and you’re spot-on. Pretty much anything more risqué than Cinemax is illegal to sell in a brick-and-mortar store. All those dollars are diverted to online resources — and Utah sees no sales tax as a result.

  • [robulus]

    “forcing non-Mormons to resort to the internet for their wanking”

    You say that like its a bad thing.

• [PaulR]

  Are those stats world-wide, or what the average ‘Mercan thinks is the world (that is, the USA)?

• [DarthVain]

  Its the classic cause VS causality argument.

  Too many times people that use statistics jump to conclusions (many times convenient, or to prove whatever they want to prove) as to what the numbers actually mean in reality.

  Freakanomics is a good book that shows this principle. Statistics is not simple, and usually cannot stand alone, and has to be taken in context of all reasonable factors.

  Many medical “facts” scare me a bit as someone does some random study, and then based on that they make some grandiose determination. It might be reasonable, or even true, however there are proper ways to use statistics and eliminate possible error. In many instances I don’t think this is done, as it takes additional work, and doesn’t make the headline any better.

  Usually it is a number of factors that determine anything, but it is way sexier to say that they found that People that do X have a 20% higher chance to get Y.

  That doesn’t mean X causes Y, it is only an empirical observation, they then jump to the conclusion that it does. It could be that the People that do X have less money, and as a result, eat less healthy, or have more stressful jobs, etc…(any number of related factors) and thus are more susceptible to Y, but it certainly doesn’t mean in any science type way that X is at the cause of Y in any kind of definitive manner.

  Trying to explain this to people over and over again can be frustrating.

• [Wordguy]

  “The average porn site visit lasts 6 minutes and 29 seconds.” Heh.

• [LostCatSoda]

  … Utah isn’t 100% Mormon – a pretty common misconception. It has been steadily dropping and is roughly 60% now. Not that there are pockets of vastly higher and lower percents.

  In Provo, where BYU is, you can go into a chain restaurant (like PF Chang’s) and order a bottle of wine. Right there at your table. In front of everyone. Things be a changin’.

  So maybe it’s all the non-Mormon’s who are whoopping up the internets with fleshy bits.

  • [Brainspore]

    … Utah isn’t 100% Mormon – a pretty common misconception.

    Really? I’ve never met anyone who thought that every single resident of Utah was Mormon.

    • [matt blank]

      Most people don’t think every single one of us are Mormons. But it gets irritating when people ask you where you’re from it comes out “Utah And I’m Not A Mormon” as if that was the name of the state.

    • [Matthew Miller]

      Really? I’ve never met anyone who thought that every single resident of Utah was Mormon.

      Seriously, I’m kinda surprised to hear that it’s as much as 60%.

    • [LostCatSoda]

      ok: I was making the point that the author, Maggie, wrote “Heh, dirty Mormons” which is a fairly typical jab at Utah. It get’s tiresome.

      • [Brainspore]

        ok: I was making the point that the author, Maggie, wrote “Heh, dirty Mormons” which is a fairly typical jab at Utah. It get’s tiresome.

        I’m sure it does, but a 60% majority is more than most stereotypes are based on. For example, less than 37 percent of voters in Massachusetts are registered Democrats and probably even a smaller percentage of Hawaiian people know how to surf.

• [Anonymous]

  Not quite sure when adult dating became considered pornography either. Better not look at e.harmony again or any other such site again.

• [Jer]

  If you’re interested, I believe that the statistics cherry-picked for the Utah “factoid” are from this paper. Page 217 has the table showing the various broadband porn subscriptions by state. Haven’t read through it much but by their methodology Utah and Alaska have the highest porn to broadband access ratio while Montana and Idaho have the lowest.

• [capl]

  There are lies, damn lies, and then there are statistics. All of your posts are great!

• [Anonymous]

  “People who are so clean-cut and law-abiding that they don’t even feel comfortable illegally downloading their porn.”

  Or they could be so backwards they don’t realize that porn is available freely inasmuch as water is.

• [matt blank]

  I can assure you that porn is virtually inaccessible in a non-online medium here in Utah. Of course there are places to get it. Just very very few of them. Just like full strength alcohol. Those of us who seek it out know where to get it easily. Those who do not routinely seek it out must use other more accessible methods, aka the internet. Trust me – if you could order alcohol online in a similar fashion (instant, private delivery) Utah would probably have the highest subscription rate for that too. I know I would sign up. I don’t like giving the socialist State Liquor Store my money.

  • [Gilbert Wham]

    Fuck, if I had a socialist liquor store, I’d go there. Especially if they sold vodka with a crudely-executed woodcut of a rocket on the label.

  • [Anonymous]

    Oh man, someone just thought of a the next multi-million dollar business ;)

• [Anonymous]

  In statistics we were always taught that you needed to see the source data, know where it came from, and only then can you understand the results and know the meaning of the findings. Without that information you know squat. This is doubly true when it comes to subjective questioning like you’ll find in polls. How was the question worded and was it asked in different ways to eliminate as much bias as possible? Because of my background I know that statistics can truthfully lie for you.

  Oh, and the other commentator is dead on with the fact that causality can not be proven by a relationship in the data. We all know that more babies in Britain are born in the spring and that storks migrate there in the spring, but this does not mean that storks are the cause of more babies. (They actually come from cabbage leaves…)

• [phlavor]

  A king calls in a Mathematician, an Accountant, and a Statistician. He tells them, “I have a problem for you to solve. What is two plus two? Come back tomorrow with your answers.

  The next day they all come back and the king asks the Mathematician what his answer is and he says, “I worked out the problem in several different ways and I always got the same answer. Four. I’m absolutely certain of this.” The Statistician chimes in, “Your Majesty, I also got four, but I think I could get it down to three.” The King turns to the Statistician and says “Well? What is your answer?” The Statistician replies, “I dunno… What do you want to hear?”

  Joke courtesy of my college Statistics teacher. The only math class I ever enjoyed.

  • [Matt Deckard]

    Wait, so there was a mathematician and two statisticians?

    I’m guessing the second person to chime in was actually supposed to be the accountant?

    • [Antinous / Moderator]

      That’s just mean.

      • [Felton]

        What are you, a comedian?

  • [Anonymous]

    You do realize that it would take a mathematician a lot longer than that to prove that 2 + 2 = 4, right? Russell & Whitehead’s proof that 1 + 1 = 2 was on page 58 of volume II of the Principia.

• [Anonymous]

  The stats don’t really make sense anyway. Other stats you can derive from the infographic:

  If 28258 people view porn each second and there are 24644172 porn sites then on average each porn site gets 99 visits a day. Seems way too low to make any money. What % of those 99 porn viewers are paying each day?

  If the global internet porn revenues are $4.9 billion (I’ve seen estimates much lower) and there are 24664172 sites that means on average a site makes about $198 per year. Again seems low.

  I’m guessing the most out of whack stat is the number of porn sites. 24 million is way too high.

• [Suburbancowboy]

  Post #14 makes a great point. The article points out many of the possible factors for statistics to not be maningful, but this key social factor (lack of access offline) is a huge one.
  another factor that must be considered is the number of people with broadband in Utah. If that number is very small, then your sample size is very small as well.

• [venicerocco]

  Is the diagram of a baby really necessary? Or is the baby supposed to be doing the searching? I mean, what would be the appropriate image to illustrate “rape” searches? Or “bestiality” I wonder?

• [techdeviant]

  I have access to an incredible amount of statistics for a very large group of porn sites. I can say at least our stats do not match up very well at all to this graphic. I am pretty sure that 1 in 3 users on our sites are not women.

  This does make me want to update our bland reports to look like this though.

  • [Brainspore]

    I am pretty sure that 1 in 3 users on our sites are not women.

    So 2 out of 3 users are women? That’s a lot of randy ladies!

• [Tynam]

  For a solid non-mathematician’s grounding in critical thinking about news statistics, I’d recommend Paulos’s classic A Mathematician Reads the News.

• [Anonymous]

  actually, the first thing I thought when I saw that number weeks ago on something else was “haha, mormons are stupid enough to pay for porn.”

• [Anonymous]

  I doubt the veracity of these statistics and thank you for questioning their interpretation.

  http://wqebelle.blogspot.com/2010/10/pornography-statistics-laundering.html

• [sergeirichard]

  My theory: As you probably know, Mormons feel obligated to induct all people into the Church of Latter Day Saints in order that they might be saved. They’re aware though of just how many people there are in this world and how fast that number is going up. So there’s a clear need to prioritize – and who is in need of more urgent salvation than the workers in the sex industry?

  So that thing they’re doing in front of their computer screens? That’s baptism.

  • [Shelby Davis]

    Note to self: stop reading BB comments in the library.

• [Anonymous]

  “you have a god-given right to steal that”

  Gimme a means of *anonymously* throwing you a few bucks and I will… thing is, I really can’t be having that show up on my credit card statements.

  capcha: “rubbers the”… lol…

• [Daemon]

  Hence the “Lies, Damned Lies and Statistics” thing.

  It’s pretty easy to “prove” almost anything using statistics, if you set things up right.

  Let’s say you want some statistics about the sex trade. (I’ve just finished writing a report on this, so my experience is fresh.)

  Who do you include in the ‘sex trade’? Street prostitutes are definitely in. What about strippers? Porn stars? Dominatrixes?

  Example: The opinions held by some feminist groups are supported by data from street-level prostitutes, but seriously challenged by data from male prostitutes and escorts. I’ve seen a few studies from those groups – they commonly passed off female street prostitutes as the norm, when it’s actually a fairly small minority.

• [Anonymous]

  I realize that Utah has an image problem, but it’s not like we are all pioneer farmers out here. BTW, stuff is great here, you should check it out.

• [dudois]

  Facts 2 and 4 are not consistent with each other.

  If $3075.64 is spent on internet porn every second,then the worldwide revenue would be $97 billion, not $4.9 billion.

  Inconsistent facts like these appear together all the time.

  Perhaps in this case part of the inconsistency arises from tax evasion and undeclared income?

• [JoshP]

  This is also an example of how traditional media can make anything sound like anything. Even if they don’t know they’re doing it. Add in the colossal dimensions of the people involved and the quasi-fluid of public conception and you have the current world state. For good or ill.
  Lies, damn lies, only joss whedon can save us now.

• [Anonymous]

  I’m not in Utah but… I don’t watch pornography online, but I once bought a feminist porn video. Is it so strange that some of us might want to support the performers by paying for our porn?

• [joshhaglund]

  My 10th grade civics teacher taught us to remember “statistics are like loose men and women; you can do with them as you please.” Seems relevant here.

• [Antinous / Moderator]

  I’d pay good money to see Donny Osmond in Cum Fart Tsunami 3. Just sayin.