An Unscientific Ad

A male version of Dove’s latest ad campaign:

Virginia Postrel criticizes Dove’s “social experiment”:

Gil Zamora, the forensic artist, is indeed a well-respected professional who worked for many years doing composite sketches for the San Jose, California, Police Department. But, unlike his interview subjects, he knew the point of the exercise going in.

The “experiment” wasn’t double-blind. Zamora’s knowledge matters because a verbal description of any given feature allows a lot of interpretative leeway, and he could have unconsciously biased his drawings. “This is a social experiment,” he said in an interview, repeating the Dove mantra. “We were trying to show how women are their own worst beauty critics and how they see themselves and how others see them.” Right.

Exacerbating the potential for bias, Zamora deviated from his standard procedure, which includes giving the witness a chance to review the sketch and correct any misinterpretations. There are two possible sources of error in a composite sketch: the witness’s memory of what the person looked like and how that memory gets translated from impression to words to drawing. One witness’s “round face” is fat but oval, another’s is round and not especially fleshy, while another’s is what an artist would call square. Proportions mean different things to different people. And some people simply have limited vocabularies.