4 ms·

I wish when people use survey statistics that they would show us the original question. For example I wanted to know if the sexual advances question was specifi

I wish when people use survey statistics that they would show us the original question. For example I wanted to know if the sexual advances question was specifically AT WORK, or if it could be at any point in their lives in any context. If they're just asking if women who happen to work in tech now have ever received unwanted sexual advances in their entire life, well, that is worthless without a non-tech statistic to compare/contrast it to. Also what is an unwanted sexual advance? Someone touching someone? Someone asking someone out once? A boss pressuring their underlings for sex/dates? It has no precise meaning. Unfortunately I spent some time studying Sociology in my youth and one of the biggest takeaways from that was "you can make people agree with any position you want by changing the wording on the questions." So precisely what you ask is super important for knowing what to do with the conclusion. We actually created surveys where we got completely opposite results from exactly the same democratic just through word-alterations (on purpose).

hwaite 11y ago
> If they're just asking if women who happen to work in tech now have ever received unwanted sexual advances Given that only 60% of women reported unwanted advances, it's safe to say that it's not "at any point in their lives in any context." That being said, you make a good point about disclosing details of study.
- UnoriginalGuy 11y ago
  > Given that only 60% of women reported unwanted advances, it's safe to say that it's not "at any point in their lives in any context." Do you have data on the level of unwanted sexual advances over the course of a woman's life? I just googled it and cannot really see too much. PS - Although boiling down a definition of an "unwanted sexual advance" is required to compare/contrast different data sets.
dlss 11y ago
> "you can make people agree with any position you want by changing the wording on the questions." So precisely what you ask is super important for knowing what to do with the conclusion. This. Also: any other questions asked which got less news-worthy results. https://en.wikipedia.org/wiki/Publication_bias https://en.wikipedia.org/wiki/Publication_bias I wish there was was something like a wikipedia page which listed every paper / survey with a conclusion on a topic out there. Kind of like examine.com but for social science results rather than nutrition.

hwaite 11y ago

> If they're just asking if women who happen to work in tech now have ever received unwanted sexual advances Given that only 60% of women reported unwanted advances, it's safe to say that it's not "at any point in their lives in any context." That being said, you make a good point about disclosing details of study.

UnoriginalGuy 11y ago
> Given that only 60% of women reported unwanted advances, it's safe to say that it's not "at any point in their lives in any context." Do you have data on the level of unwanted sexual advances over the course of a woman's life? I just googled it and cannot really see too much. PS - Although boiling down a definition of an "unwanted sexual advance" is required to compare/contrast different data sets.

dlss 11y ago

> "you can make people agree with any position you want by changing the wording on the questions." So precisely what you ask is super important for knowing what to do with the conclusion. This. Also: any other questions asked which got less news-worthy results. https://en.wikipedia.org/wiki/Publication_bias https://en.wikipedia.org/wiki/Publication_bias I wish there was was something like a wikipedia page which listed every paper / survey with a conclusion on a topic out there. Kind of like examine.com but for social science results rather than nutrition.