On gender and me being ethnocentric

[tko_ak]

This isn't LGBT related, but it does relate to gender roles, religion, and violence, so I thought I would share.

First, we see that a 75-year-old Saudi Arabian woman is to receive 40 lashes and four months in prison, for having the audacity to mingle "with two young men who were reportedly bringing her bread." Here's another story for the

Comments

[randomguy3]

It's intruiging that nearly twice as many men thought that it was OK ("in some circumstances but not others", whatever that means) to hit or slap their wives or girlfriends for wearing sexy or revealing clothes in public as for cheating on them. I wonder why.

The other thing is that while the responses from women saying that it was acceptable (again, "in some circumstances but not others") were consistently lower than those for men, the differences generally weren't significant. The differences in responses between genders were only statistically significant for two out of the five questions.

Personally, I wouldn't trust the survey. A poll with so many unexplained anomolies, and with a sample size of less than a thousand, isn't really useful for anything other than pointing the way to where further research needs to be done.

From the introduction to the results:

Please note that this research is not part of the Home Office National Statistics programme but is exploratory work. Results should be seen as indicative and directional rather than absolute. It is not recommended that exact percentages are used outside the context of this document

[randomguy3]

Oh, and I recommend reading the rest of the survey, even if I wouldn't trust any of the numerical figures. Particularly interesting are the entries that are zero (which are not rounded - values strictly between zero and a half are represented by a *).

[tko_ak]

I think we always have to be skeptical of any poll results. The article doesn't get into many specifics, about its margin of error. But it did say that the sample size was 1,065, which while not massive is larger than most U.S. presidential polls (which are typically pretty accurate).

[randomguy3]

Huh. The article says 1065, but the PDF poll results says 915 people were questioned.

Anyway, a good rule of thumb is that a sample size of less than 2000 shouldn't be trusted, because the error is too great.

I suspect they pull some tricks to get accurate results in voting polls with smaller sample sizes. The result is dependent on the whims of the swing voters, whereas most voters will always vote the same way.

[tko_ak]

I never heard that, in either of my research methods clsses where we went over polls. And most presidential polls have under a thousand responses, and they're typically pretty accurate. If the margin of error is below 5%, they're seen as accurate.

[randomguy3]

Eh, well, my old physics teacher always said that you want a margin of error of √2% or less. But that's scientists for you.