NewScientist (23 January 2016) echoes recent, better known controversies in their One Per Cent column:
A study from Carnegie Mellon University in Pittsburgh, finds that websites show men ads for higher paying jobs than those women see.
Very puzzling, so I tracked down a rather more detailed publication, the CMU News. Sadly, they’re bollixed as well:
The study of Google ads, using a CMU-developed tool called AdFisher that runs experiments with simulated user profiles, established that the gender discrimination was real, said Anupam Datta, associate professor of computer science and of electrical and computer engineering. Still unknown, he emphasized, is who or what is responsible. Was it the preference of advertisers? Or was it the unintended consequence of machine learning algorithms that drive online recommendation engines?
An interesting detail:
To study the impact of gender, researchers used AdFisher to create 1,000 simulated users — half designated male, half female — and had them visit 100 top employment sites. When AdFisher then reviewed the ads that were shown to the simulated users, the site most strongly associated with the male profiles was a career coaching service for executive positions paying more than $200,000.
“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,” said Amit Datta, a Ph.D. student in electrical and computer engineering. By comparison, the ads most associated with female profiles were for a generic job posting service and an auto dealer.
Notice they aren’t saying the same jobs are being shown at a lower salary, they simply didn’t show the ads as often. If we are willing to associate frequency with probability of attracting the attention of qualified candidates, then this all just seems like madness: why cut out the smarter half of the species?
And then there’s the larger context:
“This just came out of the blue,” Datta said of the gender discrimination finding, which was part of a larger study of the operation of Google’s Ad Settings Web page, formerly known as Ad Preferences. The finding underscores the importance of using tools such as AdFisher to monitor the online ad ecosystem.
I find something quite amusing about subjecting a commercial operation’s secret algorithms to study, almost as if it were a new, natural species. It makes you wonder if a bunch of computer scientists, even with Ph.D.s, are really the proper researchers. Perhaps some biologists should be doing the study.
Finally, returning to NewScientist’s One Per Cent column, was this additional tidbit:
In 2013, Harvard researchers noticed that searches for African American names prompted more ads offering to check for arrest records.
