3 ms·

> The researchers obtained the self-reported information from Self-reported is generally unreliable. Whether its sex or food people are generally not accurate

> The researchers obtained the self-reported information from Self-reported is generally unreliable. Whether its sex or food people are generally not accurate when asked how much X or Y. So a better question is: Why would they under-report? Or are the current acceptable social norms such that it's okay to be honest, but in previous eras there was accepted (?) exaggeration? <sidebar> It would be interesting to know previously what typically "pairs" with sex. For example, alcohol. If alcohol consumption is down in certain tribes then it follows that sex might drop as well. </sidebar>

hn_throwaway_99 4y ago
If you have 2 self-reported studies 5/10 years apart, it's still valid to compare them. The article even discusses the possibility of changes in self-reporting behavior, e.g. that young people today are more comfortable labelling themselves as asexual than they were 15 years ago.
- chiefalchemist 4y ago
  How so? Social norms change. White-lie X 10 years ago might be replaced with White-lie Y today. That doesn't make comparing X to Y valid. As mentioned, before you do the comparison, you *must* make an effort to understand the changing social norms and how that might impact a *known* flawed data source (i.e., self reported). Then once you come up with a theory about the impact of social norms, you can try to test your theory... Condom sales? Up or down? Other birth control prescriptions, up or down? STDs, up or down? Preg-test sales, up or down? What about associated / related Google search data*? The bottom line...accepting self-reported data at face value is a true fool's errand. The Science is clear on this. Let's just leave it at that for today. * "Everybody Lies: Big Data, New Data, and what the internet can tell us about who we really are" by Seth Stephens-Davidowitz (a Harvard-trained economist, former Google data scientist) http://sethsd.com/everybodylies/ http://sethsd.com/everybodylies/

hn_throwaway_99 4y ago

If you have 2 self-reported studies 5/10 years apart, it's still valid to compare them. The article even discusses the possibility of changes in self-reporting behavior, e.g. that young people today are more comfortable labelling themselves as asexual than they were 15 years ago.

chiefalchemist 4y ago
How so? Social norms change. White-lie X 10 years ago might be replaced with White-lie Y today. That doesn't make comparing X to Y valid. As mentioned, before you do the comparison, you *must* make an effort to understand the changing social norms and how that might impact a *known* flawed data source (i.e., self reported). Then once you come up with a theory about the impact of social norms, you can try to test your theory... Condom sales? Up or down? Other birth control prescriptions, up or down? STDs, up or down? Preg-test sales, up or down? What about associated / related Google search data*? The bottom line...accepting self-reported data at face value is a true fool's errand. The Science is clear on this. Let's just leave it at that for today. * "Everybody Lies: Big Data, New Data, and what the internet can tell us about who we really are" by Seth Stephens-Davidowitz (a Harvard-trained economist, former Google data scientist) http://sethsd.com/everybodylies/ http://sethsd.com/everybodylies/