"Her study is meaningless. She cherry-picked posts, and then she based her percentages on the number of posts she picked, as if that somehow translates to overall percentages. "

Did you not read the paper? http://dspace.rowan.edu/bitstream/handle/10927/869/johnsonl-t.pdf?sequence=1 ( See "Chapter 2. Method. )

She picked a month at random and reviewed all 1712 posts ( noting that two were removed by the Mods. ) While it might not be meaningful given the flux here ( particularly the rise of new posters in summer, ) it hardly qualifies as cherry-picking.

And no, she doesn't publish 1712 posts in their entirety in her paper. She doesn't even post all of the ones conforming to the supposed "cognitive distortions." So the few quotes are meant to be characteristic of those containing "cognitive distortions."

A look at Table 1 shows that even the most common distortion "children are sexy" is found in only 36% of the 1712 posts. Clearly we are failing to groom each other properly here. ;p

TOC does a pretty good job doing a critique of the flaws in the paper.

Dante