Friday, September 28, 2007

Economists on the Happiness Gap

MSNBC reports that men are happier than women (video). MSNBC makes the uncited claim that women enjoy time with friends and family less than their male counterparts. The New York Times gives more detail, claiming that this is a shift from 20 years ago, when women reported more happiness than men.

Both studies sparking recent coverage are by economists. Betsey Stevenson and Justin Wolfers at University of Pennsylvania compared general quality of life assessments. Alan Krueger at Princeton used time survey data (Krueger, 2007, pdf) to suggest that women are unhappy for 90 more minutes per week than men. Krueger notes that men are spending less time doing "unpleasant" things and more time relaxing than 20 years ago, whereas this shift has not occurred for women.



Oh, those busy, busy women. Seriously, I would really like to see more studies that compare single, childless men to single, childless women. Maybe I'll make a point of finding more of those next week. Yes, I know it's valid and generalizable to the human experience, but the whole "women are like X because of child-rearing responsibilities" bit is starting to get really old. Now, maybe my complaint is with fathers for spending less time contributing to the well-being of the family unit. Maybe my complaint is with mothers for setting unreasonable achievement goals. But I feel like my real complaint is with the use of the words "men and women" to mean "father and mothers."

livejournal version

Thursday, September 27, 2007

Corporal Punishment: parent gender

According to Helsingin Sanomat (2007), a recent Finnish study has found that women are more likely than men to use physical discipline on their children, despite holding more negative views towards it. Nobes et al (1999) also found that women were more likely to use physical punishment in a British sample; Wolfner and Gelles (1993) also found higher rates of mother punishing in a US study. However, some notable exceptions found no gender differences in prevalence of physical punishment, such as Wissow (2001) and Holden et al (1999). Many studies note "time spent with children" tends to be higher for women, and this probably explains much of the difference in punishment rates.

The gender of the punishing parent may have differential effects on children. Harper et al (2006) found that when one parent was supportive while the other punished, child outcomes varied by parent gender. A punishing father was linked to more aggressive children, whereas a punishing mother was linked to more depressed children.



I've started this paragraph twice to say that I couldn't speak from personal experience.

I think that's not entirely correct. I have long been unconvinced of my ability to restrain my temper; that's one of many reasons I've chosen to neither bear nor adopt children. Granted, I haven't had supervisory power over a child since my early teens, as a babysitter. I am deeply ashamed to recall that I did actually spank a child, more than once. I usually had the parent's permission, but that doesn't make me feel much better about it. While I can believe in a difference between "punishment" and "hitting", I never punished a child with the cool-headedness that I think creates that distinction. It wasn't so much about teaching the child as about releasing my frustration with a situation that had gotten out of my control.

It's probably not fair for me to compare my self-restraint at 14 to the self-restraint of an adult parent. Several of the above studies did note that younger parents were more likely to hit their children.

livejournal version

Wednesday, September 26, 2007

Height and Disease

Davey Smith et al (2000) used a 20-year Scottish cohort study to assess the observation that height is inversely related to respiratory disease, cardiac disease, and stroke -- that is, that shorter people are at greater risk. Louis Israel Dublin (1949) noted the correlation between short stature and shorter life in his statistical studies in the early 20th century. Davey Smith's study found that the strongest association between height and death was for respiratory disease, and that when forced expiratory volume (FEV) was controlled for, this correlation was reduced, suggesting that respiratory function may simply be related to overall size. However, for certain cancers, cardiac disease, and stroke, shorter height still showed a correlation with higher risk, even after adjusting for socioeconomic factors, which had long been thought to be the cause of both the shorter height and the higher disease risk. Davey Smith et al found that height was associated equally with stroke risk and cardiac risk for women, but more so for stroke than cardiac risk for men.

Jousilahti et al (2000), in a similar Finnish study, were not able to disassociate the risk-height relationship from socioeconomic status. Interestingly, the Finnish study found that, for men, risk of violent death also increased with shorter height.



I was curious about this because of a conversation I had about six years ago where I was complaining about my height to another short friend of mine. He mentioned that shorter people lived longer, and at the time, I believed him. Since we already know that women tend to be shorter and women tend to live longer, I thought that perhaps some of the women's mortality advantage might be tied to height, but it appears my friend was just wrong. The socioeconomic impact on height and later health makes a lot of sense to me; I'd have to take a much closer look at how the Scottish study controlled for this to be really convinced.

livejournal version

Tuesday, September 25, 2007

Unsolicited porn

Grimes et al (2007) expected to find gender differences in attitudes towards unsolicited advertising emails, or spam. The survey actually found much greater differences based on age than gender, although men were more likely to report having received sexual and financial spam than women, and tended to rate their computer expertise higher. Older computer users received more spam despite lower overall computer use, and were more likely to have purchased something from a spam email. Similarly, a PEW/Internet survey (2007) found that women reported less pornographic spam, and that women are "most upset by porn spam" (although what groups they are comparing "women" to here is unclear).

The difference in receipt of porn spam may be tied to action taken with porn spam. A CTV story reports that men are more likely to click on links in pornographic spam. The CTV story links to the Canadian Journal of Human Sexuality article (Nosko et al, 2007, PDF), which claims that men were responded more positively to sexually explicit emails and pop-ups than women, not surprisingly considering the finding that they were far more likely to be searching for sexually explicit content online.



I think an additional factor is probably the use of online dating services, which is a topic I don't think I've covered. In paper personals, women's ads are often free, because they need to encourage women to participate. Online, most ads are free, but responding to ads often carries a fee. While I can't find any specific research to back this up, I suspect that men are far more likely to have personal ads than women. Email accounts I have associated with personal ads have seemed to receive more porn spam than non-personal-ad accounts.

livejournal version

Monday, September 24, 2007

Bicycling Safety

Traffic and transport psychologist Ian Walker is passionate about cycling safety. So much so, in fact, that he donned a wig to see if cars passed women or men more closely while cycling, according to a New York Times, (2006) story from last year. He found that when dressed as a man, cars gave him less room than "when they thought he was a woman." (His results were later published: Walker, 2007). Cars gave him less room when he was wearing a helmet.

Krizek et al (2005) found that men were more likely to commute by bicycle than women, which they suspect may be related to women's tendency to combine their commutes with errands and carrying passengers. Krizek also found that women were more likely to describe environmental factors as threats to bicycling safety (lack of bike paths, road conditions) while men were more likely to name behavioral factors (unsafe driver behavior, unsafe bicyclist behavior).



When I read the Times article on Dr. Walker's work, I said out loud: "This has to be a British study." I mean, what other country is going to produce a psychologist who cross-dresses "in the name of science." Not that I don't think his results are fascinating, but even I can't suppress a giggle -- and wonder whether cars were giving him more space because they thought he was a woman, or because they thought he was a transvestite.

Still, it brought to my attention my own habit of giving women extra room -- not just on bicycles, but on the subway, when crossing the street, and just in general. I go out of my way to make sure that women do not feel threatened. I'm much less likely to do this for someone I perceive as a man. Besides, if I keep mowing down all the hot chicks in my car, who am I going to inappropriately ogle? Who will produce the next generation of hotties for me to even more inappropriately ogle in my dotage?

livejournal version

Friday, September 21, 2007

HCI: Human computer interaction

Laura Beckwith thinks girls aren't using tools. An Associated Press story dated September 24th tells about the many experiments Beckwith, with her advisor Margaret Burnett, have designed to try to figure out why women do not seem to use advanced functions in the software they use. Beckwith suspects that this is down to a difference in the way men and women solve problems.

Beckwith et al (2006) found that confidence did not predict how much men used a debugging tool in an experiment asking subjects to correct broken spreadsheets. Confidence did predict how much women used the tool. When women did not use the debugging tool, the AP article says that they ended up with "more bugs than when they started." After many manipulations of the software, Beckwith was able to create a version of the debugging tool that women used just as often or more often than men did.



I do correct a lot of spreadsheets. I fact, I think it's safe to say that I <3 spreadsheets. Spreadsheets are love. However, I've never seen a debugging tool for one, and I'm generally distrustful of the "suggestions" that any software makes to me, whether it be for spelling errors, formatting, or anything else where that damned paperclip pops up and makes me want to strangle it. Yes, I know, I could use OpenOffice -- or even just turn off the Paperclip. That's not as much fun as having a focus for my frustration.

Beckwith's main suggestion is that there are differences in the way men and women solve problems, and that current computer tools are not set up to make that intuitive for women. However, what I find really interesting is her suggestion women over-tinker. The more "tinkering" men do, the more bugs they fix, according to Beckwith. However, women seem to overtinker. This strikes me as related to introducing more errors into the spreadsheets, but I'm not sure how to make the pieces fit.

livejournal version

Thursday, September 20, 2007

Wine, Women, and fear of being Wrong

A new survey conducted by Harris Interactive (2007) that men and women are equally likely to drink wine (63% and 61%), but that women were more likely than men to feel uncomfortable making wine choices. The survey, commissioned by Robert Mondavi Private Selection, found the biggest gaps in wine attitudes by age, not by sex. Most of those surveyed (69%) expressed that they wanted to know more about wine. Women expressed more discomfort about choosing a wine from a wine list, pairing wines with food, or interacting with a waiter when selecting wine. Women were also less likely to have sent back a bottle of wine after tasting it (17% to 10%). Barber et al (2006) also found that women expressed more concern about making a good wine choice than men. Barber suggests that self-confidence is the predicting factor in wine anxiety.

It is important to note that although men and women drink wine with roughly equal frequency, this is the only alcoholic beverage for which this is true; men are more likely to drink beer or spirits than women (e.g. Chomak & Collins, 1987). Although the gap in men's and women's drinking rates seems to be narrowing ("Girl Drinks", 10/3/06), a woman who drinks is far more likely to drink wine than a man who drinks.



Excuse me for a moment of unsurprise. Women expressed more anxiety over making a choice? Women seemed to have less self-confidence? Wine selection is built up into this weird cultural juggernaut, where confidence is pretty much the name of the game. We've had multiple discussions over the tendency of women to have less confidence than men, and I don't know if any of your have tripped over a solution yet. I certainly haven't. However, I do think I have some insight into the illogic that makes women (or anyone with low self-confidence) uncomfortable dealing with the boggling selection available on most wine lists.

I've never sent back a bottle of wine. I've only sent back anything at a restaurant once, a mixed drink, last month. Mojitos are supposed to be sweet. And now I feel really self-conscious about admitting I ordered a Mojito. As Brian said in Family Guy, "what? it's not a gay drink!" It was bad enough ordering a Mojito in front of all my friends (yeah, I'm not particularly secure in my masculinity. would you be?) . It was even more embarrassing sending it back, because I felt like that set me up as some sort of expert on Mojitos. In fact, I wouldn't have sent it back, except for the conflicting presence of a (male) friend of mine who is a big proponent of sending things back. He wouldn't have let me off the hook if I hadn't.

livejournal version

Wednesday, September 19, 2007

Anne & Mary: A Piratey Portrait

So, in this mornins' post, I mentioned a paintin' I wished I coulda used ter capture the spirit of the famous pirate women Mary Read and Anne Bonny. Normally, I don' like ter make a second post in a day, let alone on th' same subject, but in this case, an' fer this day, I'll make an exception.

Ye see, mateys, I heard back from the artist, and I this paintin's got under me skin like a badly infected rat-bite. I know the lot o' ye are too lazy to click on such a link when it be provided ye, so I'm glad ter present "Anne & Mary" by Erica Chappuis. Erica Chappuis also asks that I note that this paintin' appeared in Subcomandante Insurgente Marcos' book, Conversations With Don Durito: The Story of Durito and the Defeat of Neo-Liberalism, published by Autonomedia (a "radical" publishing house That ought ter catch th' attention of at least one er two of me more politic'ly-minded readers).

Now, I have a mighty suspicion that the artist didn't think the credit she was askin' fer would be quite so elaborate-like, but she don't know Danny Linkbeard. It's just not in me bones ter post about a book wi'out providing a link, and a touch more information. On t'other hand, th' more I read about th' book, th' more I think this weren't quite such a benign subject after all.

livejournal version

Lasses wi' Cutlasses


No peg legs (exceptin' the broom) in
this photo by flickr user jurvetson
It be that time o' th' year again me hearties!

Accordin' ter th' fine lasses at Beagle Bay Books, (a fierce band if ever I seen one), th' few ladies o' th' high seas -- the actual pirate lasses o' history -- were every bit as bloodthirsty an' ruthless as any o' their male counterparts. Maritime archaeologist Jesse Ransley (2005) (now don't that sound like a name fer a slightly-built cabin boy?) points out that while womenfolk was by'n'large banned from th' manly maritime arts, th' few women what did make a name fer themselves was "mythologized" in popular tales o' the time.

Ransley also rightly spies that women pirates were made out ter be unfeminine: wanton, masculine, and even deviant, citin' Rictor Norton (1997) suggestion of th' likelihood that Anne Bonny and Mary Read were haulin' each others keels, as it were.



Apart from th' stealin', the killin' an' th' rampant beheadin', what's not ter like about the fierce women o' the sea? While the Tampa Florida group "The Bonney-Read Krewe" claims that Bonny and Read were th' only female pirates "known" ter have terrorized th' Western hemisphere, it's clear that "known" be the operative word. Bonny and Read both spent considerable time posin' as men, and it seems unlikely beyond reason that every woman who did the same was discovered.

I'd been wishin' ter illurstrate today's post wi' this fine portrait of Read and Bonny by artist Erica Chappuis, but I weren't able ter reach her fer permission in time. Still, I can't be believin' that a mere link would be oversteppin' me bounds.

livejournal version, me hearties!

Tuesday, September 18, 2007

Speed Dating and Assortative Mating

According to Wikipedia, "Speed Dating" was introduced in 1998, and by 2000 had gained national recognition, partially thanks to an episode of Sex in the City wherein a successful lawyer pretends to be an airline stewardess to stop scaring away potential dates. The repetitive format of speed-dating made it a natural choice for the study of mating choices, providing many interacting pairs in a brief period of time. Finkel et al (2007) offer a "methodological primer" for those interested in performing speed dating experiments.

Speed dating studies have released many interesting, if not particularly unexpected, results. Pentland et al (2004) found that mutual attraction was best predicted by female vocal cues. Fisman et al (2006) suggests that women place more importance on race than men do, although this may be confounded by social status. Todd et al (2007), which found that women were choosier than men. The more attractive a woman was, the pickier she was. The 4 subjects who rejected all the other participants (1 men, 3 women) were all more than 1 SD above the mean for attractiveness; the 5 individuals who received no offers for second dates (2 women, 3 men) were more than 1 SD below the mean for attractiveness. However, these less-attractive men made more offers to compensate; the unchosen women did not make more offers than the women who were chosen.



Now, I'm confused. I was absolutely, positively sure that I'd already cited Fisman et al, but apparently not. Well, damn. [edit: Found it. I posted it in November as "Kamenica and Simonson (2006)"]

At any rate, it seems that one of the main purposes of the Todd study was to examine whether people were aware that they had to offer something to attract a mate. The "trade-offs" hypothesis is the status/wealth for attractiveness/youth model that seems to dominate evolutionary psychology assortative mating papers. I wonder how many evolutionary psychologists are wealthy, somewhat-funny-looking men. I'm certainly not surprised that no one admitted to "trading off" attributes when selecting the best partner. Of course we're all looking for "kindred spirits" and no one is any "better" than anyone else -- just a better match. I think it's very comforting to think that way, and probably only partially false.

livejournal version

Monday, September 17, 2007

Scent of a Man

Androstenone, a testosterone derivative present in sweat and urine, can smell very differently (or not at all) to people depending on a single gene, according to Keller et al (2007). It is, of course, stronger in men's sweat and urine. Some smellers report a "sweet, floral" odor, while others report a more logical "sweaty, urinous" odor, and some can not detect the scent at all. Twin studies two decades ago (e.g. Wysocki & Beauchamp, 1984) implicated a genetic component to this ability, but Keller's study found a strong effect of a single gene. Dorries et al (1989) suggest that most, if not all, children are able to detect androstenone, but that this ability decreases with age, and is more likely to decrease in men. Brand and Millot's review (2001) confirms that many studies find that women are more sensitive to androstenone than men, but that men are more likely to find the smell pleasant.

Because some mammals use androstenone to communicate social signals, many consider the variation in human sensitivity "intriguing." Leslie Vosshall, one of the co-authors of Keller's study, asks "what happens to humans who can't get the signal because they have the nonfunctional copy of the gene? Or the hyperfunctional one? What could be the social and sexual implications of this on one's perception of the smell of fellow humans?" (RU News, 9/16/07)



As I have mentioned with no small amount of embarassment before, I'm a smoker. The fact that I can smell anything amazes and concerns me. In the past, when I have temporarily quit smoking, I've actually found the smells and tastes of the world to be fairly overpowering and largely unpleasant -- however, I recognize that my continued smoking makes me overpowering and unpleasant to a lot of people. *shrug*

I bring this up because, even with my senses dulled by smoking (and possibly by testosterone, but I'm not sure), I still get a very strong sense of "my people" or "not my people" when I meet someone. Traditionally, I've identified it as "they smell right" or "they smell wrong." Discussing this sense, I've encountered other people who say they smell people as "right" or "wrong" -- and they are all women. (Granted, my n is 3 or 4)

livejournal version

Friday, September 14, 2007

Measuring scientific influence

Claire Cain Miller (2007), in Forbes, reports that gender diversity in R&D may not just be good for people, but good for companies and science in general. "Who Invents IT?", a report released this week by the National Center for Women & Information Technology, examines the relationship between IT patent-holder gender and the number of times that patent is cited in subsequent patents. NCWIT found that mixed-gender patents received up to 42% more citations than single-gender patents. Women were named on only 6.1% of U.S. IT patents in 2005.

The underrepresentation of women in patenting is not unique to the IT field. Ding et al (2006) found that academic women in the life sciences applied for patents only 40% as much as their male peers. Robert Fisher (2005) suggests that women may choose less risky research subjects, leading to fewer patentable breakthroughs. However, Ding's study, which also used "number of citations" as a measure for the influence of research, found that women's papers were cited slightly more often than men's (there is no detail as to mixed-gender groups provided). Ding suggests that concern about maintaining their academic careers is a higher priority for women scientists, which they choose not to jeopardize by pursuing commercial ventures.



How do you measure the scientific impact of an idea? The number of citations seems like as good a place as any to start, but I wonder how much the politics of citing comes into play. Most of the scientists I know are fairly cynical about the business of publication in general, and about citations in particular. Still, I find it extremely promising to see results-based arguments about women's inclusion in science, rather than morally based ones. If mixed-gender teams are coming up with more influential work, that seems like the best reason to keep both men and women involved in science (and everything else, for that matter). Even after more than a year of researching for Difference Blog, I'm hesitant to say I could quantify the differences in the way men and women think -- but I do believe that there is one, and that it has an effect on the way they solve problems.

livejournal version

Thursday, September 13, 2007

Walking speed

Browning et al (2006) found no difference in preferred walking speed between men and women of normal and obese weights. All subjects preferred a walking speed of around 1.42m/s (about 3 mph), which was determined to be the speed that maximized distance for effort. This is in contrast to previous results wherein Spyropoulous et al (1991) reported a preferred walking speed of 1.09m/s for obese men while Ohrström et al (2001) found a preferred speed of 0.75m/s for obese women.

Oberg et al's (1993) reported that "significant sex differences exist in all gait parameters" in their study of normal men and women aged 10-79. Women reached their fastest average walking speed (1.29m/s) in the 30-39 age group, whereas men topped out in the 15-19 age group, at 1.35m/s.



My coworker, the Evil Ganome, is fond of ranting about the slow walking speed of people blocking the sidewalks in front of him, which he attributes to their flip-flops. I tend to be offended by flip-flops for more aesthetic reasons, but agree that this particular choice in footwear will slow you down. I note this because I tended to believe that the stereotype of women walking slowly was based on the sadistic nature of women's shoe design, but this does not appear to be entirely the case. One thing that is not reported in any of today's studies is whether the differences in walking speed are proportional to differences in length of stride, although Oberg does note that women's step frequency is higher than men's, which suggests (to me) that women are used to trying to keep up with people with longer strides.

livejournal version

Wednesday, September 12, 2007

Swagger and Sway

Johnson et al (2007), in three experiments, examine the relationships between body shape, gait, sexuality, and perceived sexuality. They conclude that when the gender of body type (hourglass vs. tubular) and gait (hip-sway vs. shoulder-swagger) conflict, a person is more likely to be perceived as homosexual, and one of their experiments suggests that this perception is likely to be accurate. In one experiment, undergraduates were asked to identify the sexuality from video of actual people on a treadmill, and in another, asked to identify sexuality of computer-generated animations (presumably, these did not have an orientation). Observers were good at identifying the gender of actual walkers, but accuracy in identifying sexual orientation was just above chance for male walkers, and below chance (although not significantly so) for female walkers. Although the walker videos presented to observers were divided evenly between gay and straight walkers, the observers guessed "straight" for most of the men, and more than half of the women.



Johnson begins with the assumption that "people can accurately judge the sexual orientation of others." I, for one, haven't found this to be true -- but bear in mind that I don't think a lot of people can accurately judge the sexual orientation of themselves. Whatever, I'm projecting, but I don't really buy the whole "I knew since I was a kid, but I fought it for 40 years" story. At some point in there, the people who come out in later life thought they were straight. Anyway, that's off the point.

This is another "gay men are like women, and women don't have a sexuality" paper, essentially. One bit I really chuckled over was the assertion that men had to work at walking femininely, but that when women walked masculinely, it was because their body types were more masculine. One of the main arguments of the paper is that "don't ask, don't tell" is pointless because people can spot the Gay. If men are having to work at their Gay Walk, it seems like that's a voluntary signal -- if it exists at all. I'm curious how they got p < .0001 with a 55% Gaydar success rate.

livejournal version

Tuesday, September 11, 2007

Potluck research

Neuroscience and literature author Jonah Lehrer (2007) drew my attention to King-Casas et al (2005), in another economic-game experiment, this time using fMRI. The researchers observed that reactions in the head of the caudate nucleus were greater for benevolent partner behavior than malevolent. Lehrer goes on to quote Colin Camerer's "Neuroeconomics" page:
The difference in brain activity in the two genders is like the kind of behavior you might see after a couple gets home from a potluck dinner and rehashes the event. The man wants to turn on the TV and catch some sports scores (his medial cingulate is turned off). The woman is more likely to rehash the evening's events, and worry about whether she said the right thing and whether the hostess was happy with the dish she brought, and whether plans for having lunch later in the week are genuine.

In fact, most of the points raised in Lehrer's post seem to be drawn from Camerer's Neuroecon page, and not from the study he links to (which doesn't actually reveal the gender split of the subjects). Camerer's Neuroecon page also doesn't seem to point to any published research about gender differences in fMRI results in economics games, although there are interesting links to further discussion of neuroeconomics.



I feel like I'm being unnecessarily harsh on Lehrer and Camerer. I have to admit I'm disappointed, and that's probably making me mean. Lehrer's post opens with a condemnation of bad gender brain science (and he's not too hard on the eyes, either) so I was pretty excited to see what he considered good research. It appears, unfortunately, to be another example of "this theory fits my experience, and therefore is science." Lehrer is at least conscientious enough to admit this in his closing. Now, Camerer may have actually seen the gender differences he describes, but I can't find any published proof of this. I think I may have to agree with Ariel Rubinstein, who reportedly called neuroeconomics "a field that oversells itself." Whether the feeling originates in my caudate or my gut, I have to say that I'm not buying it.

livejournal version

Monday, September 10, 2007

Ambulance Use

Ron Winslow (2007), reporting in the Wall Street Journal, describes (but does not identify) a "Minnesota study" presented at the 2007 Annual Meeting of SAEM (Society for Academic Emergency Medicine). The study suggests that rural men are less likely than rural women to call 911 at the sign of a heart attack. Urban men and women called 911 at equal rates (65%), but rural men drove themselves or had a friend/relative drive them to the ER in almost 2/3 of the cases of heart attack presentation to ERs, compared to about half of women.

Brown et al (2000) did not find any association between use of emergency transportation and sex in a sample from 20 U.S. communities. However, they did report geographic differences -- but these were largely based on how each community paid EMS services. Living alone was a significant predictor of using emergency transportation.



I have a little personal exposure to this phenomenon. The only time I've been transported in an ambulance was in a rural area. There was little choice in my case -- I'd been in car accident (so had no other transportation) and suffered a head injury (so moving me was something bystanders were not eager to do). My grandfather, on the other hand, had an event which was later determined to be a minor stroke, and drove himself to the ER. He was also in a rural area at the time.

In my experience, getting an ambulance to an urban residence takes a lot less time than getting one to a rural one, and the return distance to the hospital is shorter as well. If the variable of interest is "getting the patient into the hospital as quickly as possible" then it may actually make sense for a rural patient to get his wife or neighbor to drive him -- although I wouldn't be quick to discount the medical expertise of the EMTs.

livejournal version

Friday, September 7, 2007

Baby face

Friedman and Zebrowitz (1992) suggest that sex-role stereotyping is tied to facial maturity. People of both genders with "babyish" faces are perceived as warmer and more loving, and women tend to have more babyish faces than men. In their experiment, they found that sex-role stereotyping was weakened or even reversed when the difference in "facial maturity" was reversed.

Zebrowitz and Montepare's (2005) review for Science points out that "such features as a round face, large eyes, small nose, high forehead, and small chin" appear to be universal, and influences viewers opinions of competence (lower) and integrity (higher). In situations of increased threat, Pettijohn and Tesser (2005) found that less neotenous (baby-like) features were preferred. Baby-faced men, according to Todorov et al (2005) tend to be slightly more intelligent and better educated than their mature-faced peers.



Female-to-male (FTM) transsexuals tend to be a little baby-faced. I knew that, and I knew why, but I didn't realize that it could be influencing our social environment this profoundly...

I've just spent 5 minutes typing and re-typing this paragraph, and I can't find a good way to put it. So, I'll warn you, the next few sentences even offend me: MTFs have it a lot harder than FTMs. This research makes me wonder if there are layers to this difference in social outcome that are associated with our facial structures. Baby-faced men get read as honest, whereas mature-faced women get read as harsh. How much of the MTF reputation for bitchiness is based on this social stereotype? How much of the FTM reputation for soft-spokenness? How much of our "socialized behavior patterns" could be based in facial observer bias?

livejournal version

Thursday, September 6, 2007

Recommendation Letters

Schmader et al (2007) analyzed 886 letters of recommendation for male (235) and female (42) applicants to tenure-track positions in Chemistry and Biochemistry. Very few gender differences reached even a p < .05 level of significance (possibly partially due to the very small female sample). In fact, more differences were found between the tone and word types of the letters between departments than between genders. Biochemistry recommendation letters, for example, contained more negative feeling words and fewer positive feeling words. However "standout adjectives" (e.g. outstanding, unique, exceptional) were more often found in letters written for male applicants (although it is unclear in the paper whether this is based on subgroup analysis, see "Underpowered", 8/24/07 for more info). No analysis was done on the gender of the letter writer.

A similar study was done by Trix and Psenka (2003) on recommendation letters for applicants to the medical faculty (71% male). The recipients of these letters (those reviewing the applications) were 96% male, and the writers were 85% male. Trix and Psenka found letters for female applicants were shorter (avg 26 words shorter), and tended to include more "doubt raisers" -- hesitant language, qualifiers, and faint praise.



I have to admit a certain amount of skepticism about most types of language analysis. Schmader et al used Pennebaker's 2001 "Linguistic Inquiry and Word Count" (LIWC) software, which was also used in gender predictions of blog text, discussed in December. Trix and Psenka used a more qualitative method involving non-blind reviewer ratings of letters (and here I admit extreme skepticism about qualitative analysis). It's extremely easy to find sexism when you're looking for it. It is not clear if there is any difference in the year in which the letters were written. Trix and Psenka's letters were written in the 1990's, while there is no date information on Schmader's sample.

Business writing does not constitute natural language use for most people. It tends to say more about the letter writer's training than about the letter subject. I was extremely disappointed by the lack of writer-analysis in Schmader's study, and a study I'd be very interested to see is an analysis of the sources of recommendation letters for male and female applicants: is there a difference in which, and how many, people they ask for letters of recommendation?

livejournal version

Wednesday, September 5, 2007

Federal Jobs (US and Canada)

An article in yesterday's National Post announces that women now outnumber men in "core federal government employment." Since just 1995, the balance in federal bureaucracy has shifted from 54.1% male to 54.2% female. Much of this shift is attributed to a 33% cut in less-skilled jobs over the past decade. Men held 61% of the executive jobs in this survey.

According to the U.S. Office of Personnel Management (2004), there have also been declines in the number of "blue-collar" federal jobs (non-postal) in this country. Women have made gains in the technical and professional "white-collar" categories, but there have been declines in the number of women employed in clerical positions. The total employment split was 56% to 44%; interestingly, only 39% of those retiring in 2004 were female. According to testimony by the Rhonda Trent, president of Federally Employed Women (2007, PDF), women made up 47.1% of the total federal workforce in 2005, but only 26.2% of the Senior Executive Service.



It seems that the frequently mentioned phenomenon of greater diversity in male status (men outnumber women on the lowest and highest tiers) also applies in federal employment. This does not strike me as a merit-based system, although it could be education-based. While women have been outpacing men in education in recent years, the average age of the federal workers in the U.S. sample was 46.7 -- a generation that would have entered college around the time I was born. Canada seems to be doing a better job of equality than the U.S., which doesn't surprise me at all. <LOLCAT> Hai, I can emigrait now plz? </LOLCAT> I'm also trying to figure out how women could be retiring less often than men. It could be that 2004 was just a fluke. Another option is that women leave without retiring -- since women tend to live longer than men, it seems unlikely that they were "dying in the saddle."

livejournal version

Tuesday, September 4, 2007

The Hunter-Gatherer Divide

Much of evolutionary psychology appears to hinge on the idea that men were hunters and women were gatherers, and that gendered traits evolved in response to the different needs of these goals. Hunter-Gatherer economies are thought to be the only mode of subsistence for humans for 2 million years, ending somewhere between 5,000-10,000 years ago.

Burton, Brudner, and White (1977) suggest that gendered division of hunter/gatherer patterns came about because "constraints of nursing and the effect of supplementary feeding of infants" restrict female participation in many activities (such as hunting). Goodman et al (1985), however, points out that the Aeta (or Agta) women in the Phillipines are able to actively participate in group and individual hunting "without detriment to normal fertility and child care." Lyn Wadley (1998) notes several examples of female hunters in Africa, as well as citing the Agta bow-women*.



I am reserving judgment, for the time being, on hunter-gatherer gender differences. It doesn't make a lot of sense to me that women would not participate in hunting (since many early hunting strategies required many hunters), and even less that men would not gather. It's difficult to find sources that are not either "superiority of men" texts from the 1950's and earlier, or "feminist response" texts from the 1970's and 1980's. It does appear that female hunters may be an exception, rather than a majority, but it's hard for me to know what to think, given the strong bias on both sides of the debate.

*Thanks to Astrogeek01 for finding the Wadley article.
livejournal version

Monday, September 3, 2007

Nintendo Wii Users @ Restaurants

According to the Nintendo Wii "Everybody Votes" channel, which allows Wii users with an internet connection to take polls, there are differences in the ways men and women use restaurants. Men seem to go to restaurants more often, but not for the same reasons as women. 28% of men, compared to 25.6% of women, answered that they ate more than 3 meals per week in a restaurant. Overall group response = 27.3% Given a choice between a movie theater and a restaurant for enjoyment, however, men chose the movies (57% to 43%) while women chose restaurants (53.6% to 46.4%). Overall group response was 54% for movies.

Given these numbers, it appears that between 68-72% of Wii poll respondents are male. However, it is possible that some women are using male characters, because that is the only way to avoid having your character wear a dress. These were both polls of U.S. residents only.



As I have mentioned before, I don't really cook. It makes a lot of sense to me that if you eat a significant portion of your meals at a restaurant, it stops being something you do "to enjoy yourself." It does annoy me that "total number of respondents" is never available from the "Everybody Votes" channel, but these polls really are for entertainment value only.

Happy Labor Day!
livejournal version