Some of you may remember the outrage that emerged last month after Facebook (finally) came clean about the casual human experiments it had been doing, without users’ permission, or notice. Their emotional manipulation study, which controlled the kind of news an individual saw on their feed (negative or positive) to see whether their mood would be affected, had telling results, but was also slated as unethical and morally wrong by many outraged social-media users.
Since then, the popular dating site, OK Cupid, has announced to its users that they shouldn’t be upset about human experimentation, since the website uses it all of the time. The founder of the website, Christian Rudder, announced that experimentation was part of the reality that came with using the internet, stating that ‘That’s how websites work’. Rudder went on to explain that he and the engineers at OK Cupid sometimes indulged in a bit of fun with random user’s accounts so that they could research the best ways of further building the site. In the past, for example, tweaking has included nonchalantly deleting patches of profile text and informing pairs that may have made a great match that they actually weren’t right for each other.
Isn’t that…the opposite of what these sites are supposed to do?
Some of the Previous Experiments (and what they suggest):
Love is (not) blind:
On January 15th of last year, the site decided to temporarily take down all user photos to see if people would still want to interact without knowing what a person looked like. Since human-beings are inherently shallow, the site traffic dropped significantly, but those that did interact were found to respond and exchange information more quickly…Until photographs were restored. The women who did try out a blind date usually reported having a pretty good time, especially if the date was with a less attractive man, surprisingly. Yet, despite this, once photos were restored onto the site, the women went right back to judging people according to their looks.
It’s (not) important to have a good personality:
The second experiment, once again based on looks – because human-beings are nothing if not aesthetically minded – attempted to determine whether a correlation existed between how people were scored for their personality, and looks. Basically, the website wanted to know whether someone could get a good rating for their personality, even if their appearance was rated negatively. Sadly, once again the results made us look bad, by finding that how a person looked seemed more important than any other factor. People would positively rate a person’s profile if they were attractive, even if they had no idea what their personality was like whatsoever. According to the site’s founder, Rudder, where a picture is worth a thousand words, actual, real words are worth ‘almost nothing’.
Who needs words when you have a pretty face?
People will do what they’re told is right for them (even if it isn’t):
The final experiment that has been revealed by OK Cupid looked into just how powerful suggestion could be in the matchmaking progress. The engineers deliberately altered the match algorithm that is typically used to tell users who they were compatible with, so that instead they got results directing to people they were completely incompatible with. The study found that, even though these people were likely not to get along, because the site told them they were, they were much more likely to interact and send messages to one another. According to Rudder, the suggestion of compatibility is just as effective as the real thing.
“The myth of compatibility works as well as the truth.”
Surprisingly, despite the negative response that Facebook got for doing (pretty much) the same thing, the comments that emerged on OK cupid’s blog suggested that people weren’t as weirded-out as you might have expected by the human experimentation. In fact, rather than being upset after finding out what had actually been going on with their profiles, users said they found the data to be either funny, or pretty ‘depressing’.
Share your views in the comment box
