« Testosterone economics | Main | An early start in life is a good start »
Do women talk more?
Today, study published Science magazine: 396 subjects wear tiny microphones. Result: whoops. Women emit 16,125 words per day, men 15,669. Statistically, even-steven.
Here is the story. I am pretty quiet, though, so I do my part to uphold the honor of men.
Posted by Tyler Cowen on July 5, 2007 at 05:11 PM in Data Source | Permalink
Comments
Hmm, I've had Deborah Tannen's:
Women (25,000 words per day) and Men (10,000 words per day) in my head ever since premarital counseling.
I am certainly willing to concede a move to parity based on the post-feminist-modern-metrosexual counterrevolution. But that also leads my confirmation bias baggage to suspect that guys in college are macking on chicks big time, so of course they need to talk more then they want to. Once they reel in a victim, they can dial it back down to a fraction of the woman's word count. And the unattached guys would be more available for this study to begin with, or need the money to buy more women drinks.
Nothing in the article helps me overcome my bias. But it helps me check it quite a bit.
Posted by: caveat bettor at Jul 5, 2007 5:36:04 PM
The article quickly jumps to a conclusion based on the mean, but look at the accompanying graph: the median and mode are substantially higher for women; the male average is driven by small percentage of outspoken men. Based on the graph, it seems equally reasonable to conclude that the typical female is far more loquacious.
Posted by: blink at Jul 5, 2007 6:05:47 PM
I probably shouldn't do this, but in order to answer "blink"'s commments, I typed in the data from the plotted histogram. I reproduced the means given in the article.
I then computed the medians by finding the bins that contain half of the sample of men (93) or women (107.5). I interpolate through the bins to get a better guess at the median (in other words, the bin the contains the 93rd male starts at the 86th male and ends at the 100th male, hence the 93rd male is the middle of that bin).
The median values are ~13000 for men and ~14000 for women, half a bin apart. The average in both cases are skewed by outliers and by almost the same amount.
Discussing the mode when the bins have a maximum of ~25 points in them is pretty meaningless. Poisson errors yield 20% errors on the expectation value with those size numbers making the mode even more sensitive to outliers than the mean.
The interquartile ranges are the same, ~10000 words.
I will leave computation of the biweight center and scale as an exercise for the reader.
Posted by: Brad Holden at Jul 5, 2007 7:28:06 PM
Perhaps it's that most of the words that men speak are monosyllabic, so it sounds like we talk less. Gr, argh.
Posted by: Trieu Truong at Jul 5, 2007 8:37:37 PM
This reminds me of something I heard in a meeting:
Senior VP of Marketing: "What 5 words can we use to describe [new product]."
Male voice from back of room: "Can they be really big words?"
Posted by: chug at Jul 5, 2007 11:13:54 PM
Men tend to like to express themselves in public (as in blogs and blog comments), women in private.
Posted by: Steve Sailer at Jul 5, 2007 11:47:13 PM
It's the nature of the talk, the kind of talk, too. Men punch out a lot of jokes, information and opinion, where women go on about feelings, experiences, and pulling apart relationships and characters. If men can look like assertive bullies to women, women can seem to go on and on forever about sheer drivel to men. My wife jokes that mens' natural form of self-expression is the op-ed piece, where women's natural form of self-expression is yakking about feelings and people over coffee and cookies with a girlfriend.
Posted by: Michael Blowhard at Jul 5, 2007 11:53:21 PM
Where the hell did they conduct this study, San Francisco?
Posted by: Yancey Ward at Jul 5, 2007 11:55:51 PM
I'm inclined to go with blowhard's general viewpoint. Women complain men don't share their feelings, while men aren't entirely thrilled that women talk a lot about theirs.
This doesn't mean that men don't talk as much, but perhaps men don't talk as much in the company of women? Get some men together without women present, or with a large enough component of men, and I can pretty much guarantee you that at some point the men (overall, there are outliers) will be talking an awful lot about sports with minimal input for them women.
Posted by: happyjuggler0 at Jul 6, 2007 1:22:40 AM
I'm not sure what the relevance of this is, but if you put a microphone on
anybody, and tell them you're going to measure how many words they speak,
it will definitely skew the results.
Of course some people will talk more than normal and others less. I think
there would have to be a different type of study to get results that could
be relied on.
Posted by: Gary Bourgeault at Jul 6, 2007 2:29:20 AM
I'm sick of seeing media coverage of gender studies that treats not rejecting the null hypothesis the same as proving it.
Posted by: James at Jul 6, 2007 2:45:50 AM
There has never been a day of my life where I said 16,000 words, and there never will be.
Posted by: jb at Jul 6, 2007 4:58:19 AM
Caveat: the study used college students. Their habits may or may not be representative.
Posted by: Constant at Jul 6, 2007 6:31:29 AM
This is an amazing statement:
"Despite the flaw [college students], says lead author, Matthias R. Mehl, University of Arizona psychologist, 'Our paper puts to rest the idea that the female brain evolved to be talkative and the male brain evolved to be reticent.'"
One paper puts to rest an entire debate? Really?
I don't have access to the actual article - by the way, the title of the article is "Are women really more talkative than men?", which it would be nice if the NY Times article would mention - but do they control for text messaging? Perhaps women have stopped talking as much but still "communicate" more via the text message.
Posted by: AZ at Jul 6, 2007 7:45:10 AM
I feel obliged to point out Language Log here:
http://itre.cis.upenn.edu/~myl/languagelog/archives/004691.html
They have been hitting this topic for a long while, and that link is to their latest posting, noting that the blog is cited in "Are women really more talkative than men?"
Posted by: Zubon at Jul 6, 2007 8:28:00 AM
Not in my house.
College students are typical of, well, college students.
Posted by: save_the_rustbelt at Jul 6, 2007 9:34:55 AM
Not in my house.
College students are typical of, well, college students.
Posted by: save_the_rustbelt at Jul 6, 2007 9:35:02 AM
Not in my house.
College students are typical of, well, college students.
Posted by: save_the_rustbelt at Jul 6, 2007 9:35:06 AM
Upholding the "honor" of men, huh? Please remind me why it is necessary to assign "honor" to a misplaced, and perhaps disproven, stereotype. While I too would rather put a fork in my eye than listen to mindless chatter, I don't assume, or find, that it comes from one sex more than the other, much less that men alone are honorable when they refrain from it.
Posted by: Nancy at Jul 6, 2007 9:43:05 AM
Men speak more in social settings. The study should be conducted in cars. 1 man, 1 woman, 2 men, 2 women. Monitor how much is said in a 4 hour drive with no stops.
Posted by: 8 at Jul 6, 2007 10:38:50 AM
~16,000 words per day sounds too much. Assuming 8 hours of sleep each day, a person would speak ~1000 words per hour, or 16.67 per minute, or one word every three and a half seconds during their hours awake.
The numbers appear very inflated (even for college students). Either the algorithm that counted the words is broken or they just happened to find 396 people who just never shuts up.
Posted by: Robert at Jul 6, 2007 10:50:08 AM
I love Mark Liberman (of Language Log) for the unimprovable 'sex-linked linguistic budgets', but this looks like a question that begs for economic analysis.
Suppose I want food I don't want to cook myself. I select a restaurant. It has a menu with options and prices. The prices may vary widely from restaurant to restaurant, but the amount of speech required to elicit a desired meal doesn't show much variation. If I travel, the range of meals on offer may change, the range of prices may change, but the speech requirement remains pretty much the same. There's unlikely to be much difference between sexes in quantity of utterances.
But suppose I want sex with someone other than myself. Sentences certainly exist capturing preferences - I can find them in any number of books on the subject. But I can't just identify a practice and a sentence and then deploy the sentence on a likely-looking prospect as a way of satisfying the demand. I probably have to deploy quite a large number of sentences - most having nothing whatsoever to do with the desired practice - in the hope of ultimately getting the desired prospect to the point where the thing desired can be indicated (quite possibly not with a sentence).
The number of sentences uttered tells us nothing about the relative propensity of persons with XX and XY chromosomes to utter sentences. Social constraints give different sentence prices to different sexual preferences; the level of demand may, of course, vary by sex and be reflected in the number of sentences a person may be required to pay, given the social constraints.
That's not to say, of course, that some people may not find the activity of talking a benefit rather than a cost. That preference might or might not be more prevalent among women than men. It's hard to see how a study such as the one described could offer enlightenment.
Posted by: Helen DeWitt at Jul 6, 2007 11:39:08 AM
"I don't assume, or find, that it comes from one sex more than the other, much less that men alone are honorable when they refrain from it."
Women are MORE honorable when they refrain - for it is harder for them to.
Posted by: Tom at Jul 6, 2007 12:58:04 PM
Grunt. I wonder how many garden paths psychology has been led down by the fact that researchers use middle-class 18-22 year olds with IQ between 100 and 140 as their baseline for normal psychology.
ALso, unless they planted these bugs on these kids without their knowledge, there is a serious distortion due to observation. I can't imagine going around with a microphone all day and not having it affect what I say. Some people would 'perform', others would be intimidated. It wouldn't surprise me if there is a gender bias - socialized or inborn - in how many perform and how many don't.
Posted by: rvman at Jul 6, 2007 4:58:58 PM
Perhaps men speak as much as women....but no one's listening.
Posted by: elvis at Jul 6, 2007 10:30:45 PM
