Using AdFisher, a tool developed at CMU that uses statistical methodology to analyze Google ads, researchers created 17,370 simulated male and female user profiles programmed only to visit jobseeker websites. The experiment was meant to be part of a larger study concerning general Google Ad settings.
However, the CMU researchers discovered that 1,000 profiles out of the 17,370 specifically configured to peruse available salaried positions, more results popped up for the male profiles that featured career coaching services for executive positions that paid $200k or more than they did for the female-oriented profiles.
What did the ladies get instead? Run-of-the mill posting services (and strangely enough, an auto dealer).
"The male users were shown the high-paying job ads about 1,800 times, compared to female users who saw those ads about 300 times," explained Ph.D. student Amit Datta in a press release issued by the university.
In the same statement, associate professor of electrical and computer engineering/computer science Anupam Datta confessed that the mainspring for the algorithmic male preference is more difficult to trace than one might imagine, but that programs like AdFisher might be able to correct them.
"This just came out of the blue," said Datta. "Many important decisions about the ads we see are being made by online systems. Oversight of these 'black boxes' is necessary to make sure they don't compromise our values."
He also added that these types of patterns or mix-ups could be caused by either Google's system or an advertiser's own selected targets.
In conjunction with CMU, the research was also sponsored by the National Science Foundation.
Photo: Robert Scoble | Flickr