Statistical analysis of Facebook response to Bea Arthur’s death: Gay vs. Straight
Bea Arthur was all up in my newsfeed this weekend, mostly in response to her now being dead.
Here’s a statistical analysis of this activity. Be forewarned: this is real science. Prepare yourself accordingly.
Figure 1. Number of newsfeed items referencing Bea Arthur’s death, 48 hrs following death
Twelve gay people have mentioned Bea Arthur’s death on Facebook, while only 8 straight people have done so. Of course, I have many more straight friends than gay friends, so it is necessary to describe the disproportionate gay response more descriptively.
Figure 2. Gayness ratio of mentioning Bea Arthur’s death on Facebook. The gayness ratio is a measure of overrepresentation and is calculated by dividing the percentage of items posted by a group divided by that group’s percentage of the total. A value of 1 would indicate representative distribution. The figure below shows that a gay person was 6.32 times more likely to mention Bea Arthur’s death than an’average’ person and 14.36 times likelier to mention Bea Arthur than a straight person.
Figure 3. Gayness ratio of mentioning Bea Arthur’s death, males only. Gay men were nearly ten times as likely to mention Bea Arthurs death on Facebook than an ‘average’ male, and 42 times likelier than a straight male to mention Bea Arthur.
No one should be surprised by any of these results.
Click here for more background and Facebook analysis.
h/t to Kyle for giving me the idea for this post
(No Ratings Yet)
Loading ...
I’ve never heard of Bea Arthur, so I’m soooooo straight. I like these…keep em coming
Wait a tick, don’t you mean that gay people were 6.32/.44 = 14.36 times as likely to have mentioned this Bea character whom I’ve never heard of–err, of whom I’ve never heard? Let’s note that it’s too early to rule out the possibility that I’m just having one of those degenerative 3:30 am brainfarts.
Oops, nice catch there, PSKL. I think I tripped myself up in the wording.
The Gayness Ratio calculation first assumes that Bea Arthur mentions would be distributed evenly amongst all of my Facebook friends. Of course, this isn’t true, so it compares the actual incidence to the expected representative incidence to get a measure of overrepresentation.
Another way to look at this:
There were about 0.02 Bea Arthur mentions per person overall, about 0.13 Bea Arthur mentions per gay person, and about 0.009 Bea Arthur mentions per straight person.
So yes, comparing the two groups to each other, it is a factor of 14.36.
Language in the post has been clarified.