Wednesday, May 20, 2015

Today's Research Popularization Fail


Several summaries of a study about online dating attracted my beady eyes recently.  The study, supposedly telling us that more groomed dating photographs  affect heterosexual men and women differently, looked useful grist for my mill.  For instance:

Researchers at the University of Connecticut conducted an experiment to determine how people judged each other based on their online dating profile photos. They presented 671 volunteers with a single photo that was either casual or enhanced and of a man or a woman. Researchers reported that men were less likely to trust women who posted an "enhanced" photo with good angles, good lighting and make-up.
But that didn't stop the men wanting to date those women anyway, said lead study author Rory McGloin, a communications professor at UConn.
"They thought she was more attractive, they wanted to go on a date with her ... but they didn't trust her," McGloin said.

On the flip side, women found men with enhanced photos to be more trustworthy, according to their findings, which is set to be presented at the International Communication Association annual conference later this month.
Bolds are mine, and they are very important bolds.

Suppose that I tell you my research shows that hats created from aluminum foil really do keep Fox News from corrupting your brain.  Suppose that I tell you I tested this with hundreds of individuals.  But nope, you CANNOT see the research paper, even though all you have is my word about the results!  After all, I'm going to read it aloud to some colleagues in a few weeks!

I'm not accusing the researchers of making anything up.  But to popularize a paper which nobody can get hold of is extremely bad manners.  Even unethical, because of this:

Let's say that some paper popularized this way turns out to be utter rubbish (such as my tinfoil study).  When the criticisms come in, the popularizers are no longer at all interested in correcting their earlier messages, so readers are left believing in lies.  Here's one example of a retracted study which was widely popularized, but its retraction got no publicity.  More examples here.

For the sake of completeness I should mention that I asked Mr. McGloin for the paper by e-mail.  That was ten days ago, but if I get his answer I will write more about online dating and mating.