Last week I linked to a study from OK Cupid about race and dating. OK Cupid released some more data, including who likes to use strap-ons and what slang daters are familiar with, and who thinks they are a genius (guess who thinks they are geniuses more often than not?) The data I found most striking however, was the data on who has considered suicide.
OK Cupid blog notes that the straight identified people are less likely than "other" to consider suicide. But the number that struck out to me was that the most likely group to consider suicide are bisexual identified women.
Thoughts?
Thanks to Dave for the link.
0 TrackBacks
Listed below are links to blogs that reference this entry: OK Cupid Study Finds Bisexual Women Most Likely To Consider Suicide. .
TrackBack URL for this entry: http://www.feministing.com/cgi-bin/movabletype/mt-tb.fcgi/16996
first thought: with such a mismatched amount of responders in each category, how accurate is it?
I think the only conclusion that the study supports is that bi/gay women are more likely to consider suicide than straight women, given the absurdly small number of gay women in the sample.
Maybe you're looking at the Canadian data. In the American numbers, there are several hundred people in each category (male/female + gay/bi/straight), which is usually considered a "large sample" in statistical tables and general research.
I don't think we can use the research to make general claims, because the people in it are probably different from the people not on OKCupid. But I think it's important to note that it is in line with general understanding of suicide--women and people who are gay or bi are more likely to consider suicide than men and straight people are.
Yeah, I think that this needs some statistical interpretation--I don't know if pure percentages are the best way to express this data. I'm not good with numbers, though. Are there any statisticians to weigh in?
Although considering the bisexual experience is important, I am wondering what it means when over half our population considers ending their lives because the one provided to them is so unbearable?
maybe we're doing something wrong...?
When I considered suicide, I think it was mostly because of my illness, but certainly bisexuality didn't exactly produce warm fuzzies most of the time.
I'm bi, and I've considered suicide... but it had nothing to do with my bisexuality. I hadn't even come out to myself yet the first three times I was really serious about it. That is weird how every single group that is bi is the highest, though...
The figures are pretty meaningless.
First, the numbers are very low, so small differences can look larger when converted to percentages. The sample size is very small, so the results are unlikely to be 'statistically significant'.
Second, the suicide rate is 20 per 100 000 adults (that is, 20 people in 100 000 commit suicide in a given year). As a percentage, that is 0.02%.
The figures for 'considered' are altogether in a different league. The two sets of figures can't therefore mean anything - there is clearly no link between 'considering' and 'committing suicide'.
On the first point, I don't think the sample size is too small, but I think for how small it is, the effect should be more pronounced. The phi-coefficient I calculated for the set of data is 0.2, showing little association between being gay/bi and considering suicide. If the proportions were more similar, the value could have been higher.
On the second point, that's an issue with the blogger who titled this post something other than what the survey was testing.
I won't try and argue that numbers from an internet dating site are the height of science. However, I don't think that the fact that it's 'considered' and not 'committing' makes it less meaningful. That's rather dismissive towards those who have had thoughts of but not attempted suicide in some way, isn't it? And simply going through the thought process is often a symptom of depression. I think it's very important to see who is considering it, and (on a personal level) why and if there is any help you can offer them.
If these numbers were more informed, one could say there was something meaningful in the fact that the largest group that considers suicide is bisexual women. Suicidal thoughts can point to depression or dissatisfaction with life and if one group were feeling that at a much higher rate and we could identify why it could lead to many meaningful changes.
Of course, this study is incomplete in a lot of data that would actually lend to such answers. A lot of conclusions could be drawn from that chart. Straight men, for instance, might be more likely to reject/lie about considering the option of suicide because it's less socially accepted while people who identify as bisexual are already going against the primary social norms and therefore may feel more open to answering the question honestly. Suicide is also often seen as a feminine act, which might point to why woman both consider it more and might be compelled to answer the question they way they did. That's on top of easy connections "don't we already have numbers that say gay teens are more likely to be depressed?" Just not enough data.
Stephanie,
Fair points. However, if the methodology is taking 'considering suicide' as a marker for depression or dissatisfaction, then it is a strange way of carrying out such a study - it wouls have been muc better to ask the direct questions! Are you depressed? Are you happy?
In general, I am rather dismissive of these polls; the questions asked are often very leading, and it is therefore hardly surprising if the answers reflect the 'hopes' (or prejudices) of the researchers.
Caveat: I do not know how the poll was conducted.
While women, and especially bi-women, may think about and consider suicide the most, it's important to remember that men actually commit suicide at a rate 4-5 times greater than women and the highest rates are by white men over 65. There is, after all, a huge difference between the thought and the act.
Why is it important to remember, when this is a discussion about bi women, not men? The point of this is not necessarily about whether or not they WILL commit suicide - it's about pointing to higher levels of depression and suicidal ideation.
OK, some actual statistics (from a non-statistician, beware):
Since we want to determine whether the differences between proportions of independent samples is significant (% American women "yes" responders, bi vs. gay), a two-proportion z-test should be appropriate. The null hypothesis states that there is no difference between the two populations. Using the stats for American bi & gay women and this calculator, i get a 2-tailed confidence level of 99.7%. This means that we are 99.7% certain that the proportions are statistically different, or the odds of the difference occurring by chance alone are 1 : 333.
In other words, the results are painfully statistically significant. I suspect the only populations for which this wouldn't hold are the Canadian bi/gay populations, but you could run the numbers for yourself to see. There are other things to criticize about the study methodology, which were pretty well-covered in the comments of the race thread. Nevertheless, statistics show that there really is a difference between bi and gay responders in attitudes about suicide. Gaussian distributions tend to hold for populations > 30. Don't just assume that your idea of what constitutes "low numbers" is the gold standard.
Correct me if i used the wrong test; it's been a while.
I wasn't sure about a t-test, because it assumes some stuff about the sample and the general population. That doesn't make you wrong. I'm not an expert, and I came to the same conclusions that you did in the end.
I'm not a statistician, just someone who's done the stats for a couple biology publications. I used a chi-square test, which compares proportions between groups. A chi-square test is also problematic in this case, because some of the percentages were so low.
Just looking at the American data, every difference was statistically significant if you use the standard p
As other people have said, this pattern may not exist in the general population, because OKCupid users are different from the general population.
Well, perhaps we should think about OKCupid's users-- our society still stigmatizes single people, and what percentage of the site's users are dissatisfied with their relationship status? (I'm not saying that there's anything wrong with using an internet dating site; I'm just suggesting that there is probably a higher percentage of people who are unhappy on these sites when contrast to the population in general). So, perhaps the high 'contemplation of suicide' statistics shouldn't be that surprising; and maybe the bisexual woman who are using this site have had more stressful relationships than the straight-identified people (perhaps?)
I've used OKC both as a single person and a person in relationships. I don't think people who use online dating sites are necessarily less happy. What I do think, though, is that people are more likely to be honest on the OKC questions than even in other surveys, and let me tell you why: answering a question dishonestly on a survey does you no harm, but lying on the OKC match questions would harm you in that it would match you to people wrongly.
Just my two cents.
I appreciate the comments above discussing whether or not these statistics are significant. I unfortunately can't contribute to that discussion. However, as a bisexual woman with a history of suicide attempts, and as a person who buys into the stress vunerability model of mental illness episodes, I don't find these statistics terribly surprising. While bisexuality in and of itself was never a primary reason in my consideration of suicide, and was something I was initially personally very comfortable with as a young person "discovering" her sexuality (we're talking 6th or 7th grade here, which was the first time I ever found myself actually sexualy attracted to anyone), I can say living in a society where significant portions of both the dominant (heteronormative) culture and the main alternative to that culture (gay/lesbian culture) don't accept a sexuality as "real" is certainly stress-inducing at the very least. When a very real part of a person's identity isn't accepted and is even considered a joke by large factions of society, the fact that bisexual people report more suicidal ideation than other groups shouldn't come as any kind of shock. Personally, I've dealt with everything from my parents allowing me to spend every weekend at my girlfriend's house during my senior year of high school because they weren't comfortable explaining to my younger sisters that I was dating a woman (maybe that wasn't so bad ;)), to one of my sisters announcing "Well, at least you're not alone in your principles," the first time I dated a man, to gay-indentified aquaintances, upon discovering I was bisexual and not a lesbian, making comments like, "Well, you must be 60 [percent lesbian]-40 [percent heterosexual]" because of how I chose to wear my hair at the time. Lesbians that refuse to date bisexual women due to prejiduce and straight couples that think bisexual women are always down for threesomes regardless of whether the woman is attracted to either partner (we're attracted to everybody!) are very, very real. My gay uncle and lesbian cousin have gay and lesbian identities that are recognized by my family, my sexuality is kept hush hush. I've gone from being a 12-year-old girl very comfortable with her bisexuality to a 26-year-old woman who finds herself occasionally trying to "pick teams" for the sake of convenience. I'm not a stastician, but on a very gut level, a statistic that shows bisexual women are considering suicide at a higher rate than any other group rings entirely true.
It's really great that something like suicide doesn't get in the way of what's important, battle of the sexes jokes.
Okay, I get that hashing the statistics is fun and all, but I'd like to address the actual issue at hand: bisexual women are more likely to contemplate suicide.
I've contemplated suicide several times in my life - while I wouldn't say that being bi (I actually consider myself queer, but the world-at-large would shoehorn me into the "bi" box) was the direct reason for it, I WOULD say that it was probably a contributing, subsconcious factor.
The fact is that bisexual folks are castigated and denigrated by straights and gays alike. We're dismissed as fence-sitters. We're demonized as disease-spreaders. We're a part of the "no's" group in personals ads. I can't even tell you how many lesbians have trotted out the tired "bisexual women will just leave you for a man" line at me. We're objectified and eroticized for the sexual use of straight men - but only if we're gender-normative (aka feminine) and conventionally pretty.
It can be pretty devastating for a newly-out bi/queer person to be rejected by the very community they thought would support them.
Which isn't to say it's all doom and gloom - most of my queer and straight friends don't cast aspersions on bi folks and personally I love being my queer, queer self. But at the very least, it can be a cause of stress in one's life.
Look at the numbers difference between bi men and bi women! WTF!
For every subgroup, the differences do appear to be statistically significant... but there is an important problem to take into consideration, one that more than one graduate stats prof. has drilled into my head.
Where'd the sample come from?
Any conclusions we draw from this study we can only apply to members of the population who use OKCupid, which limits it to people of a minimum level of computer-literacy, a certain minimum amount of affluence (that is, access to a computer), and a personal disposition toward online dating... to try anything outside of that, we would need to have a more complete understanding of how exactly these variables confound the data, and I don't think we have that.
You also have the problem of also self-reporting, which frequently has an unbalancing effect on the distribution. To give a hypothetical example, if one were asked to take a survey about thoughts of suicide, I might suspect that a person who'd considered suicide already to be more likely to take the survey in the first place.
...not to mention NONE of the categories are under 50%. That is so terribly sad.
Before anyone jumps to the conclusion that American society must be better than the Canadian equivalent:
I would guess that the larger suicide consideration rate in Canada is linked to the weather patterns. Very long daylight hours, as experienced in the North of the country have been linked to psychological stresses. See here: http://www.britastro.org/dark-skies/health.html?7O
I think all this proves is that bisexual women are more likely to admit to having considered suicide (whatever that means to them) on a question on an online dating site than other people. I use OKC, so I'm familiar with the way the questions work. Other people answer that question and then provide how their "ideal" partner would answer and how strongly they feel. I think the more interesting question is whether or not knowing you could potentially be judged by a future partner for thoughts of suicide is different for bisexual women.