OkCupid Study Reveals the Perils of Big-Data Science

Love is not blind, as it turns out. But opposites attract when people think they are similar. At the same time, the results offer yet another example of how websites like OKCupid are sometimes used as social science laboratories — often without telling their subjects. In June, Facebook disclosed that it had tested to see if emotions were contagious, deliberately manipulating the emotional content of the news feeds for , people. After the disclosure led to an uproar by users, privacy regulators in Europe began looking into whether the social network had broken any local laws.

Online dating companies hoping to capture a share of this growing Match Group, the owner of OkCupid (as well as Match, Tinder, and 45 other dating From Rudder’s analysis, he found a bias against black users on dating .

OkCupid is an online dating site that serves international users. Kim and Escobedo-Land describe a data set where over 50, profiles from the San Francisco area were made available 15 and the data can be found in a GitHub repository The data contains several types of variables:. In their original form, almost all of raw data fields are discrete in nature; only age and the last time since login were numeric. The categorical predictors were converted to dummy variables Chapter 5 prior to fitting models.

For the analyses of these data in this chapter, the open text data will be ignored but will be probed later Section 5. Of the predictors that were used, there were clusters of variables for geographic location i. There is a moderate class imbalance in these data; only While the imbalance has a significant impact on the analysis, the illustration presented here will mostly side-step this issue by down-sampling the instances such that the number of profiles in each class are equal.

See Chapter 16 of Kuhn and Johnson for a detailed description of techniques for dealing with infrequent classes. While there have been instances where online dating information has been obtained without authorization, these data were made available with permission from OkCupid president and co-founder Christian Rudder. For more information, see the original publication.