5 ms·

Does this objection require that there be markers that predispose someone towards being a survivor of sexual assault? There probably are, but I'm sure that's no

Does this objection require that there be markers that predispose someone towards being a survivor of sexual assault? There probably are, but I'm sure that's not politic to say.

derbOac 5y ago
So I'm being downvoted, which I expected because of the subject matter, but I've done research in this area and many of the correlates they talk about are associated with familial and other ecological factors that tend to cluster very locally. I don't even personally believe that sexual abuse is not associated with the things they claim, but statistically speaking, you have many of these things clustering together (sexual abuse, pubertal acceleration, psychiatric challenges), and controlling for things in the way they did is grossly insufficient. A lot of this stuff is very broadband. Let's say you have some stressors from the environment (impoverishment, heavy metals, pesticide exposure, whatever it is). Controlling for demographic and socioeconomic factors will help address this, but the only way to do it convincingly is to look at these variables within-family, preferably controlling for genetic relatedness as well. Otherwise something like, say, chemical exposure is likely to affect children as well as the adults around them in a way that might be specific to a certain household. Not being very precise in your designs does as much damage to isolating confounding environmental causes and effects of sexual abuse per se as anything else.
- vivekd 5y ago
  I would agree with you. They did control for some factors, and that seemed to be the point of the study, but my first reaction to it was it seems like the controls they used seemed insufficent. It is a shame you were being downvoted - I expect that from other places but not from hacker news.
- srcreigh 5y ago
  The great scandal of science is the great importance of many things which cannot be controlled for in an experiment.

Does this objection require that there be markers that predispose someone towards being a survivor of sexual assault? There probably are, but I'm sure that's not politic to say.

derbOac 5y ago

So I'm being downvoted, which I expected because of the subject matter, but I've done research in this area and many of the correlates they talk about are associated with familial and other ecological factors that tend to cluster very locally. I don't even personally believe that sexual abuse is not associated with the things they claim, but statistically speaking, you have many of these things clustering together (sexual abuse, pubertal acceleration, psychiatric challenges), and controlling for things in the way they did is grossly insufficient. A lot of this stuff is very broadband. Let's say you have some stressors from the environment (impoverishment, heavy metals, pesticide exposure, whatever it is). Controlling for demographic and socioeconomic factors will help address this, but the only way to do it convincingly is to look at these variables within-family, preferably controlling for genetic relatedness as well. Otherwise something like, say, chemical exposure is likely to affect children as well as the adults around them in a way that might be specific to a certain household. Not being very precise in your designs does as much damage to isolating confounding environmental causes and effects of sexual abuse per se as anything else.

vivekd 5y ago
I would agree with you. They did control for some factors, and that seemed to be the point of the study, but my first reaction to it was it seems like the controls they used seemed insufficent. It is a shame you were being downvoted - I expect that from other places but not from hacker news.
srcreigh 5y ago
The great scandal of science is the great importance of many things which cannot be controlled for in an experiment.