In my survey, I added (females) a question about 'are you on hormonal birth control', yes/no. If I check the correlation with the question 'You're more sexually attracted to people who appear visually [masculine vs feminine]' (7-point scale), will there be a correlation?
Anything stronger than r=0.12 will result in resolving this 'yes'.
I have the data, I haven't looked at it yet so I don't know the answer yet.
*edit: Will only look at females who are either neutral or slightly into penises (as opposed to into vaginas). if there's any other obvious things to control for i forgot about i will add them in too
Update 2025-08-11 (PST) (AI summary of creator comment): - Direction doesn't matter: Any positive or negative correlation counts; will resolve YES if the magnitude |r| > 0.12.
I asked GPT-5 what else you might want to control for and it suggested age, because:
Women's preferences for masculinity in male faces are highest during reproductive age range and lower around puberty and post-menopause https://pubmed.ncbi.nlm.nih.gov/20060226/
HBC usage clearly varies by age, so age can mechanically induce a spurious HBC-masc link.
It suggested you control for this by only comparing women within a single age window (e.g. 20-34).
@jessald hey, if you keep removing confounding variables then the odds are gonna go down
it's a market asking if there's correlation, not causation
Relevant Wikipedia extract:
Several studies suggest that there is a link between hormone levels and partner selection among humans. In a study measuring female attraction to males with varying levels of masculinity, it was established that women had a general masculinity preference for men's voices, and that the preference for masculinity was greater in the fertile phase of the menstrual cycle than in the non-fertile phase.[36] There is further evidence from the same study that in fertile stages of the menstrual cycle, women also had a preference for other masculine traits such as body size, facial shape, and dominant behavior, which are indicators of both fertility and health.[36] This study did not exclude males with feminine traits from being selected, however, as feminine traits in men indicate a higher probability of long-term relationship commitment,[36] and may be one of several evolutionary strategies.[40]
https://en.m.wikipedia.org/wiki/Sexual_selection_in_humans
Source paper:
Feinberg, D. R.; Jones, B. C.; Law Smith, M. J.; Moore, F. R.; DeBruine, L. M.; Cornwell, R. E.; Hillier, S. G.; Perrett, D. I. (1 February 2006). "Menstrual cycle, trait estrogen level, and masculinity preferences in the human voice". Hormones and Behavior. 49 (2): 215–222. doi:10.1016/j.yhbeh.2005.07.004. PMID 16055126. S2CID 14884832.
However, it is possible that Aella might have better data and methodology that overturns this finding.
@TheAllMemeingEye worth noting the paper linked had a sample size of n=26 and measured (not self reported) voice preference against hormone cycles rather than a dummy variable for pill taking
another paper with n=6482 internet survey responses measuring self-reported facial masculinity preferences with the pill as a independent variable - much closer to aella's data and 250x larger sample, and from 2019 rather than 2006 - found nothing (link):
more generally this seems like the kind of small-n research from the 2000s that got replication crisis'd but the narrative persists because it plays into manosphere sexual anxieties (pill that lets women sleep around is contributing to decline of masculinity!). i'm confident if there is an effect at all it's small. order up at 55%!
@Aella I kinda want the next one of these to be positive/neutral/negative instead of y/n. I suspect that gets lower engagement, but I'm still curious...
@Aella I don't remember from the survey, sorry. It's been a long time since I did it. Are there options for being more into penises than "neutral" or "slightly".
If for example there are females who are "highly" or "strongly" "moderately" etc. into penises but they're not included here that may influence predictions.
@euclaise a lot of women take hormonal birth control for reasons other than pregnancy prevention though, so it would be interesting to know if that potential correlative effect exists in attraction to masc/femme lesbians too. Alternatively, you could potentially isolate out lesbians and run those tests in a separate sample.