Image: A bar chart showing relative sexual partner averages between iPhone, BlackBerry and Android users.
OkCupid
A snapshot of 30-year-old smart phone owners and their sexual partners, based on a study of 9,785 members of the online dating community OkCupid.
Image:
By Wilson Rothman
msnbc.com
updated 8/11/2010 3:47:47 PM ET 2010-08-11T19:47:47
commentary

There's no bait 'n' switch headline here. What you read above is indeed a discovery, one made using solid data from a large group of fairly random people. In other words, argue all you want, but it's true.

OkCupid, the self-styled "best dating site on Earth," is by extension also the best repository of anonymous (or, rather, anonymized) data on socially active single people. While conducting a site-wide survey on photography, the Cupideers realized they had a data set of almost 10,000 people that showed two things: A type of smart phone and a number of lifetime sexual partners. The results were strangely non-random.

At every age group, the iPhone-owning posse has significantly more liaisons under their collective belt, with BlackBerry owners coming in second and Android bringing up the rear. In a controlled snapshot of the 30-year-old age group, depicted in the bar chart above, iPhone-owning women had had 12 sexual partners on average, twice the number named by Android-carrying women of the same age. In fact, although the trend applies to men and women, it's 30-year-old iPhone-owning women who come off as particularly promiscuous — statistically speaking, that is.

So, what's the takeaway? First, if iPhone-owning singles are not more sexually promiscuous, they're at least bigger liars and/or braggarts. And second, if you're under the impression that — thanks to OkCupid's study — you can actually gauge your chances of getting lucky by the type of phone your intended mate is carrying, you will most certainly fail. I mean, I'm no Dr. Phil, but just don't go there.

Update: A talented social scientist — who also happens to be my wife — casts some doubt on the statistical accuracy of OkCupid's findings. She writes:

- OkCupid's report shows *correlation* not *causality.* It may be that there's something about iPhones that attracts sexual partners (doubtful, but possible I suppose) or that people who have more sex are more likely to choose an iPhone. 

However, cause and effect are impossible to decipher here.

- There's potential for omitted variable bias. Perhaps there's another factor — such as income — that predicts both iPhone selection AND coital frequency. Then the correlation between iPhones and sex may be spurious. It's like the example used in every "intro to sociology" class: There is data to suggest that murder rates spike as a result of ice cream consumption ... when actually the only connection between ice cream and murder is that they are both more common in the summer, when people are out and about.

I was glad to see you recognize the possibility for measurement error, or "braggart" error as you call it. The population in their study may be fairly representative (reducing the likelihood of sampling error), but their results don't mean anything if self-reports are not reliable, especially if certain groups (i.e., men, younger people, iPhone users) are more likely to mis-report. And we all know about self-reports of sexual activity...

OkCupid's "research" may be headline-grabbing, but it violates nearly all of the norms of social science.

Catch up with Wilson on Twitter at @wjrothman. And feel free to remind him how much smarter his wife is.

© 2013 msnbc.com Reprints

Discuss:

Discussion comments

,

Most active discussions

  1. votes comments
  2. votes comments
  3. votes comments
  4. votes comments