News for hackers. Stuff that matters.
Re:The Statistical Crisis in Science by mcguire on Saturday January 17, 2015 @01:21PM (#8905027)
"Then I took the results to a psychology prof. They loaded the data into SAS or something and it proceeded to perform HUNDREDS of independent t-tests. The results came out in a few seconds and the professor exclaimed 'Oh look! Pronouns are statistically significant! Oh and possessive nouns too!' -- I cringed."

This wasn't, by any chance, James W. Pennebaker from UT Austin? (Or somebody related to him?)

I was just reading The Secret Life of Pronouns and his chapter on lying does point to pronouns as a key factor for distinguishing lying. On the other hand, he does take some pains to note that getting to 65% correctness is reasonably easy but getting much more has been impossible so far, outside of restricted environments such as laboratory conditions.

There are no comments on this comment yet.