Yes, I think the benefit outweighs the costs. I am interested in the differences between what drew males v. females to this site. There are much fewer females, so I need more of them to reply.
This seems to have worked too, as the 2011 survey held 8% females, whereas this one seems to be holding steady at a 13% females. (The other option is that the percentage of females on LW has grown by over 50% in the past 4-ish months. This seems less probable)
In case it wasn’t clear, I agree with you. There are costs to regret in your approach, but I think the value of the data you might generate substantially outweighs those regrets.
daenerys concluded that the message encouraging women to reply must have worked because 5 % more women answered this poll. But this could just as well have happened if the same women answered again while less men did.
Therefore I wanted to know if more women actually answered this one.
The answer to that question is no: 92 women responded to the census.
We can’t say anything about “more women answering” or “fewer men answering” when we are comparing to a survey that has both more women and more men than this one. (If you weren’t talking about absolute numbers, then all we have is the 8% vs. 13% which only tells us that at least one of “more women” and “fewer men” is true.)
A better comparison would be to one of the other polls with a similar number of responders (i.e. one in Discussion), but I couldn’t find one that recorded gender. (a list of “polls” and of “surveys”)
The big 2011 survey is pretty much like the census of LW. It runs for about a month, posted on the main page, and is one of those “Everyone who even looks at this site, please answer this survey” type deals. As far as I know only 2 have been done in the history of LW.
I would not at all expect 1000 people to answer this poll. I did a quick search for “poll” here to see how many results other polls returned. The first three that I could easily find results for: one was in the 70s, one was in the 80s, and one was in the 110s. Each of those ran for longer than a day.
Conversely, the poll that I posted has only run for a day, and has garnered 102 responses. For this reason (because 102 responses in one day compares favorably to the other Discussion polls I saw), I believe that the high percentage of female responses (currently 19% of poll respondents are women) is due to MORE women responding, NOT due to LESS men responding.
This survey is (basically) a subset of the census, and there are ~100 female responders to that one. So, unless there is a big surge of female responses to this survey, the results you described here are likely to be more representative/reliable (since absolute numbers, not relative numbers, are important).
However, I think this post is still useful for the discussion it has generated.
Although, that said, the breakdown of the referred-by-friend category (into platonic vs. romantic) could be interesting. I predict that more women than men will be referred by a date, given at least 85% of the LW population is attracted to women, but at most 20% are attracted to men. I would lean towards more romantic then platonic referrals for both genders, but most people have more platonic friends than romantic ones (only 13% of LW is definitely polyamorous), so I’m not sure.
Yes, I think the benefit outweighs the costs. I am interested in the differences between what drew males v. females to this site. There are much fewer females, so I need more of them to reply.
This seems to have worked too, as the 2011 survey held 8% females, whereas this one seems to be holding steady at a 13% females. (The other option is that the percentage of females on LW has grown by over 50% in the past 4-ish months. This seems less probable)
In case it wasn’t clear, I agree with you. There are costs to regret in your approach, but I think the value of the data you might generate substantially outweighs those regrets.
Do we know that more women has answered this poll and not just less men?
There are almost certainly going to be fewer men who have answered this poll, given 972 people identified as male on the other one.
Is that what you were meaning?
No, that is not what I meant.
daenerys concluded that the message encouraging women to reply must have worked because 5 % more women answered this poll. But this could just as well have happened if the same women answered again while less men did.
Therefore I wanted to know if more women actually answered this one.
The answer to that question is no: 92 women responded to the census.
We can’t say anything about “more women answering” or “fewer men answering” when we are comparing to a survey that has both more women and more men than this one. (If you weren’t talking about absolute numbers, then all we have is the 8% vs. 13% which only tells us that at least one of “more women” and “fewer men” is true.)
A better comparison would be to one of the other polls with a similar number of responders (i.e. one in Discussion), but I couldn’t find one that recorded gender. (a list of “polls” and of “surveys”)
The big 2011 survey is pretty much like the census of LW. It runs for about a month, posted on the main page, and is one of those “Everyone who even looks at this site, please answer this survey” type deals. As far as I know only 2 have been done in the history of LW.
I would not at all expect 1000 people to answer this poll. I did a quick search for “poll” here to see how many results other polls returned. The first three that I could easily find results for: one was in the 70s, one was in the 80s, and one was in the 110s. Each of those ran for longer than a day.
Conversely, the poll that I posted has only run for a day, and has garnered 102 responses. For this reason (because 102 responses in one day compares favorably to the other Discussion polls I saw), I believe that the high percentage of female responses (currently 19% of poll respondents are women) is due to MORE women responding, NOT due to LESS men responding.
This survey is (basically) a subset of the census, and there are ~100 female responders to that one. So, unless there is a big surge of female responses to this survey, the results you described here are likely to be more representative/reliable (since absolute numbers, not relative numbers, are important).
However, I think this post is still useful for the discussion it has generated.
Although, that said, the breakdown of the referred-by-friend category (into platonic vs. romantic) could be interesting. I predict that more women than men will be referred by a date, given at least 85% of the LW population is attracted to women, but at most 20% are attracted to men. I would lean towards more romantic then platonic referrals for both genders, but most people have more platonic friends than romantic ones (only 13% of LW is definitely polyamorous), so I’m not sure.