Spell Check

An article on Gizmodo titled "Study: People Who Point Out Typos Are Jerks" caught the eye of EVERYDATA for several reasons.  First, because we all know people who love to point out typos.  But, second, because it had us wondering about the statistics behind the claims.

The study is based on an experimental design in which 83 people were randomly provided potentially different written examples of a housemate ad--some which contained certain types of grammatical or spelling errors, and others that didn't.  Here is an example from the study:

Hey! My name is Pat and I’m interested in sharing a house with other students who are serious abuot (about) there (their) schoolwork but who also know how to relax and have fun. I like to play tennis and love old school rap. If your (you’re)someone who likes that kind of thing too, maybe we would mkae (make) good housemates.

 Participants also were surveyed to determine their personality characteristics.  The authors tested the influence of many different personality traits and the reactions of the respondents to the various types of grammar errors.   The main finding in the article, as I read it, is that although a number of things were tested, the main result is that more agreeable people react less to grammar errors.

As a consumer of everydata, it is interesting to note four things: (1) this is an experimental study, meaning the researchers created the stimuli for the study; (2) the participants opted in and were paid $1 dollar to participate, and we always want to be aware of that type of sample selection, (3) the paper tests many outcomes and possibilities, but it seems the small sample size really doesn't allow for very precise estimation of all of the interactions they study, and (4) it is interesting to contrast the somewhat measured tone of the paper with the more "intense" tone of the article in Gizmodo. 

The original research is published in the journal PLOS ONE.