Did Uber misrepresent data to promote ride safety?
MANILA, Philippines – A communication professor is taking to task car-sharing service Uber for allegedly misrepresenting data to promote the safety of Uber rides in dangerous areas of San Francisco – and then taking down the said post.
Gina Neff, an associate professor of communication at the University of Washington and the School of Public Policy at Central European University in Budapest, said on Monday, November 24, the Uber Data Blog post – now deleted – is archived with the title "How crime location knowledge is a proxy for Uber demand.”
Its original web address, however, points to another possible title: “How prostitution and alcohol make Uber better.”
In her tweets, Neff called the blog post spurious, as it showed a "correlation between prostitution and ridership.”
Spurious: Uber took down data science blog post showing correlation between prostitution and ridership. Cached here http://t.co/1PJGh68OWJ— Gina Neff (@ginasue) November 24, 2014
A VentureBeat report explained that Uber’s data team took a "lighthearted look at how Uber’s own ride data mashes up with local crime statistics," using San Francisco's Crimespotting Map to best determine when and where demand for an Uber ride is likely to occur.
Uber's findings, according to their post: "Areas of San Francisco with the most prostitution, alcohol, theft, and burglary also have the most Uber rides. Be safe, Uberites!”
Following this claim, the Uber data team went off-tangent by noticing that a rise in prostitution arrests in San Francisco and Oakland appear to happen on the second Wednesday of the month. It’s a point they reasoned out as stemming from Social Security and welfare checks arriving on the second, third, and fourth Wednesdays of the month.
Neff added in a tweet, “Which Uber mistakes offend me most: misogyny, recklessness with user data, or spurious correlations masquerading as #bigdata? #CantDecide.”
Neff said she began looking through Uber's archives after hearing about another post that reportedly went missing, called “Rides of Glory,” which appeared to track the possibility of one-night stands being done with Uber’s help.
Marketplace.org summarizes it, saying that an examination of their rider data – sorting for people who took an Uber between 10 pm and 4 am on a Friday or Saturday and then took another Uber ride 4 to 6 hours later from around the previous night’s drop-off point – pointed to potential one-night stands occurring and facilitated by Uber.
Uber has been under fire in recent weeks after reports of one of its executives suggesting Uber investigate journalists to fight back against negative press, with another Uber executive being investigated for privacy violations after reportedly accessing a journalist's personal Uber ride data without that journalist's consent. – Rappler.com