View Single Post
Old 08-14-2019, 08:07 AM
Stanislaus is offline
Join Date: Apr 2001
Location: London
Posts: 3,103
Originally Posted by Sunny Daze View Post
I read the paper and it was a crap paper. The authors themselves admitted that they didn't get enough subjects for statistical significance. They are also really reaching for the causation they claim exists, even though larger studies say different things.
It's such a crap paper. Aside from the appalling sample and aside from the fact that the theoretical basis for the study relies on small-sample results that failed replication in bigger studies, they are clearly chasing the result they want.

For those who haven't read it, the paper reports on two studies. In the first (which is as far as I got), women currently in a relationship are shown videos of two men, and told that these are dating site intro videos. One man is attractive, one is average. The women are asked to imagine what a date with each of these men might be like, based on the video. These imaginings are then coded for levels of sexual attraction. These results are then analysed according to a) whether the women were on the pill when they started their current relationship and b)whether they are now, or if not are on either the high- or low-fertility point of their menstrual cycle.

There's a lot of dodginess here, not least the utter madness of applying their statistical analysis to a subjective non-interval scale of attraction, but it's worse than that. The short intro speech the men give in the videos is utterly skewed to the point of parody and reveals *quite a lot* about the authors' hang-ups:

The attractive confederate presented himself as a "bad boy", as follows:
"Hi, my name is Haim. I see myself as a curious, dominant, and ambitious person. I'm a nightlife man; I love to hang out until the wee hours. I enjoy action and
changes, because I get easily bored of anything and therefore I'm always on the go, moving. I have a tremendous affection for extremes of any kind; the sense of danger, excitement, and adventure. I always do new things, and look for someone who is willing to contain me and my craziness."
Hello ladies! I am a slave to my id, massively unreliable, highly likely to be unfaithful and in no way any kind of prospect for a relationship, although I am kind of thinking it would be cool to have a substitute mum so I don't have to grow up.

But that's OK, you're thinking. Surely they gave the "average" looking guy the same speech, so that the only variable was physical attraction? Oh you sweet summer children:

The average-looking confederate presented himself as an "ordinary guy", as follows:
"Hi, my name is Erik, and I'm a generally happy guy. I love to travel around the world alone, with friends, or with people I meet wherever I go. I'm open to new experiences while I'm hanging out with friends or new acquaintances. I study business management and meanwhile work as a waiter at a coffee shop. I work out frequently in the gym. I think I'm a curious guy who likes to live in the moment and enjoy life."
Some of the same footloose and fancy-free attitude, but couched in different language with more emphasis on existing relationships and also a) plus more mature attitude to career and b) minus the whole "it's your job to manage my craziness" nonsense.

Would it be a surprise if women who imagined dates with Haim imagined snogging/making out/shagging? There is literally no other reason to date Haim. He's a good looking man-child who'll either leave you for the next woman who makes even brief eye contact with him, or cheat on you for the danger and excitement of it. Given that the study didn't ask: "Would you date this man?" (kind of an omission there, you might think) but "What would it be like to be on a date with this man?" then it would hardly be a surprise if women responded by saying they'd focus on the physical. That's all Haim offers. He is clearly designed to provoke a particular response so that the authors can then turn around and make claims about attractiveness. Which they do.

(The utter guff about how women might desire Haim "because of" his unreliability is weirdly reminiscent of some of the worst incel prejudices about what women look for in a man but I'm sure that's just a co-incidence. It does, however, show that they knew they were loading the dice.)

But in actual fact, most women didn't find Haim desirable. On the 1-5 scale, his highest rating among any sub-group was just under 2. Which means women were more likely than not to give him a swerve. They don't report the difference in overall "desirability" between Haim and Eric, but eyeballing the data it appears to about the same.

I wouldn't put too much into that though. For the sub-sample of people who had been on the pill, and were now at high-fertility (i.e. the group most relevant to the authors' hypothesis) the mean score for Eric's desirability was 1.00, with a standard deviation of 0.00. That is to say, by an astonishing co-incidence, all 16 (heh, "all") women rated Eric as 1 on the scale. If that seems amazing, it gets better. Because the women didn't actually rate Eric or Haim on a scale. They described a date in plain language. The scale was created by "independent judges" coding that description into a 1-5 scale. So we're asked to believe that it just so happens that the group of most interest to the researchers (women who had suppressed their hormones when forming a relationship and were now in the full grip of hormone-flooded lust) were independently coded as having the absolute minimum of desire for the average guy, which conveniently means that the difference between their score for Eric and their score for Haim just creeps over the line into statistical significance. Women on the pill now and then could imagine at least snogging Eric (1.45 desirability); women who hadn't been on the pill and were at high-fertility could imagine at least snogging Eric (1.73) but somehow women who had come off the pill and whose uterus was jonesing for a zygote couldn't imagine anything even a little bit physical.

Come the fuck on. This is junk. They loaded the study design with confounding variables and when that wasn't enough, they put their thumb on the scales and they still only just scraped into statistical significance.