Out of curiosity... "purity" graph

[jay]

I looked at the posted public scores of people on my friends-list who took that test over the past couple of days, and generated a histogram. It looks like a two-mode distribution, centered around the low-30s and mid-60s respectively... I fall into the latter clump, personally (score = 69). Compare my friends with the average of all those taking the test (currently = 66)...

Oh, well... off to work now.

Comments

[mactavish.livejournal.com]

You fall into the second-biggest clump, not the last one. :)

And they still don't really mean anything in terms of "fitting in," except as you perceive it -- which is probably the most important thing.

For many folks, the fact that you made a histogram is a much more relevant detail than anything on the test. ;)

[dancing-star.livejournal.com]

Interesting, although in reality the #'s mean very little on if your going to get along with someone. That is only one aspect of a person.

Thanks for taking the time to do the graph.

[purplecthulhu.livejournal.com]

Interesting! In fact the shape (bimodal) and peak locations are pretty much what purpletigron and I predicted in a discussion last night.

Two points to note (puts on lab report marking cap)...

1. The couting errors here actually make the signal for bimodality rather weak - the error on a count of n is sqrt(n) and the largest number in any bin is 6, so there are some pretty large errors here. So professionally I could only take these results as indicative and not as definitive.

2. There is a possible reason for the bimodality - bi-ness... If, like myself or brian1789 you're pretty much straight, there is a large chunk of score in this particular purity test that you won't get. What happens to the distribution if you only combine the scores from straight or gay (whichever the person is more experienced in)? Does this remove the bimodality?

A couple of thoughts....

[brooksmoses]

I just realized, in reading one of your later posts, that this test is biased towards bisexuals by about 20%. See, people who've had full-on sex with both genders get to check off most of the points in both the "same sex" and "opposite sex" categories, whereas people who've only had experience with one gender only get one category -- even if they've had many different partners. And there goes 1/5th of the test.

Hmm....

So, consider your spread. It appears to me that a person who's had relatively vanilla sex with people of just one gender will get around 60%. A person who's had relatively vanilla sex with people of both common genders will get around 40%. Add to this that a moderate amount of kinkiness will get another 10% to 20% off, and one expects some smearing downward from those points.

Thus, I'd conclude that the gap at 50% likely indicates that your bisexual friends tend to be kinkier than your straight (or purely homosexual) friends. And the relative weighting indicates that you probably have more bisexual people on your friends list than you do straight or homosexual people.

It occurs to me that it might be interesting to try separating out the data by orientation to see if that holds up, and also to try separating it out by gender to see if there are gender differences.

[who-is-she.livejournal.com]

I scored 65.. so you're just slightly more pure than me.

:)
seriously.. I was glad to see someone else scored in my general range. All the folks I've seen taking the test have scored WAY lower numbers than me.....

I was beginning to feel a little freaky... in a very un-freaky way.
:)