Saturday, January 24, 2009

NYT strikes out again

The NYT reports on positive spillovers from Obama on black test-takers.

Note the equation by the study authors of "not statistically different from zero" with "equals zero". Of course, the NYT does not give us the point estimate so that the reader can make up his or her own mind.

The NYT does not tell us how the sampling was done either. What was the population? What was the response rate? Did the response rate vary by test cohort or by race within test cohorts? These are not hard questions if you actually are interested in seriously assessing the results rather than just providing false uplift to your readers.

Thought question: is it worse to report on results that have not yet been peer reviewed when you know what you are doing - the NYT even notes in the article that the study has not been peer reviewed - or when you are ignorant of the correct behavior, as a middle school blogger might be?

I would praise the NYT for citing some skeptical scholars but I think they are only doing it becasue they know they should not be reporting on this study at all, both because it is not peer reviewed yet and because it does not really pass the smell test, and so they are providing themselves cover and a false appearance of balance.

I could not find the study posted on the web pages of any of the authors, here, here and here, but it may be that norms about posting working papers are different in psychology than in economics.

On the other side of the ledger, the Smoking Gun reports some negative Obama spillovers.

Hat tip: Dan Black (on the NYT piece)