OKCupid revealed in an official blog post that, like Facebook, it has been manipulating some aspects of its website and then analyzing the reactions of its users as an experiment.
The experiments were made to determine the best ways to develop the website, according the blog post written by OKCupid CEO Christian Rudder.
"Guess what, everybody: if you use the Internet, you're the subject of hundreds of experiments at any given time, on every site. That's how websites work," Rudder wrote on the blog post which is entitled "We Experiment on Human Beings."
Rudder's revelations about OKCupid bring to mind Facebook's similar experiments, which raised several legal and ethical issues within the outcry of the social network's user base.
Facebook revealed in June that it carried out research for emotional contagion within the social network, which is a phenomenon that causes people around an individual to assume the same emotions being felt by that individual. The experiment replicated the results on Facebook, proving that emotional contagion can also be observed online.
However, the controversial matter was that the team of Facebook data scientists carried out the research by manipulating what the News Feed of almost 700,000 users displayed, to show either mostly negative or mostly positive status updates.
In Rudder's blog post, he reveals three experiments that were run on OKCupid. The first is "Love is Blind," which removed all the posted pictures on the website to see how users will communicate with each other despite no visual stimuli.
The second experiment involved the old rating system of OKCupid, wherein users gave ratings to others on both looks and personality. The experiments showed that users only rated others according to pictures and didn't give much thought on giving ratings according to personalities.
The third experiment involved the "matching" algorithm of OKCupid, which is the dating website's prime feature. The feature tells users which people they are most compatible with. OKCupid engineers experimented with the feature by telling some users that their bad matches are actually good matches, aiming to prove that the success of the algorithm was partly due to the power of suggestion. The experiment did just that, proving that users will respond and communicate with users that are a "good match" despite the truth that they are not.
While the similarities between the experiments carried out by Facebook and OKCupid are numerous, there are also some notable differences. One is that OKCupid informed users that had their matches experimented on were informed of the right matches after the conclusion of the experiment. Facebook, on the other hand, claimed that it did not know the specific users that were involved in their experiments.
In addition, Rudder's examples of OKCupid's experiments aimed to make OKCupid better with the results obtained. Facebook's experiments only looked to prove if the moods of users were affected by positive and negative status updates, as the company assumingly looked to disprove claims that more Facebook usage leads to sadness.
Rudder's post revived the OkTrends blog of the company, which previously shared insights on the behavior of people in online dating.
