If you ever wonder why it seems like politicians around the world appear to have decided that their political futures are best served by imposing all manner of free speech restrictions, censorship, and content controls on Web services, one might be well served by examining the extent to which Internet users feel that they've been mistreated and lied to by some services -- how their trust in those services has been undermined by abusive experiments that would not likely be tolerated in other aspects of our lives.
To be sure, all experiments are definitely not created equal. Most Web service providers run experiments of one sort or another, and the vast majority are both justifiable and harmless. Showing some customers a different version of a user interface, for example, does not risk real harm to users, and the same could be said for most experiments that are aimed at improving site performance and results.
But when sites outright lie to you about things you care about, and that you have expected those sites to provide to you honestly, that's a wholly different story, indeed -- and that applies whether or not you're paying fees for the services involved, and whether or not users are ever informed later about these shenanigans. Nor do "research use of data" clauses buried in voluminous Terms of Service text constitute informed consent or some sort of ethical exception.
You'll likely recall the recent furor over revelations about Facebook experiments -- in conjunction with outside experimenters -- that artificially distorted the feed streams of selected users in an effort to impact their emotions, e.g., show them more negative items than normal, and see if they'll become depressed.
When belated news of this experiment became known, there was widespread and much deserved criticism. Facebook and experimenters issued some half-hearted "sort of" apologies, mostly suggesting that anyone who was concerned just "didn't understand" the point of the experiment. You know the philosophy: "Users are just stupid losers!" ...
Now comes word that online dating site OkCupid has been engaging in its own campaign of lying to users in the guise of experiments.
In OkCupid's case, this revelation comes not in the form of an apology at all, but rather in a snarky, fetid posting by one of their principals, which also includes a pitch urging readers to purchase the author's book.
OkCupid apparently performed a range of experiments on users -- some of the harmless variety. But one in particular fell squarely into the Big Lie septic tank, involving lying to selected users by claiming that very low compatibility scores were actually extremely high scores. Then OkCupid sat back and gleefully watched the fun like teenagers peering through a keyhole into a bedroom.
Now of course, OkCupid had their "data based" excuse for this. By their claimed reckoning, their algorithm was basically so inept in the first place that the only way their could calibrate it was by providing some users enormously inflated results to see how they'd behave, then studying this data against control groups who got honest results from the algorithm.
Sorry boy wonders, but that story would get you kicked out of Ethics 101 with a tattoo on your forehead that reads "Never let me near a computer again, please!"
Really, this is pretty simple stuff. It doesn't take a course in comparative ethics to figure out when an experiment is harmless and when it's abusive.
Many apologists for these abusive antics are well practiced in the art of conflation -- that is, trying to confuse the issue by making invalid comparisons.
So, you'll get the "everybody does experiments" line -- which is true enough, but as noted above, the vast majority of experiments are harmless and do not involve lying to your users.
Or we'll hear "this is the same things advertisers try to do -- they're always playing with our emotions." Certainly advertisers do their utmost to influence us, but there's a big difference from the cases under discussion here. We don't usually have a pre-existing trust relationship with those advertisers of the sort we have with Web services that we use every day, and that we expect to provide us with honest results, honest answers, and honest data to the best of their ability.
And naturally there's also the refrain that "these are very small differences that are often hard to even measure, and aren't important anyway, so what's the big deal?"
But from an ethical standpoint the magnitude of effects is essentially irrelevant. The issue is your willingness to lie to your users and purposely distort data in the first place -- when your users expect you to provide the most accurate data that you can.
The saddest part though is how this all poisons the well of trust generally, and causes users to wonder when they're next being lied to or manipulated by purposely skewed or altered data.
Loss of trust in this way can have lethal consequences. Already, we've seen how a relatively small number of research ethical lapses in the medical community have triggered knee-jerk legislative efforts to restrict legitimate research access to genetic and disease data -- laws that could cost many lives as critical research is stalled and otherwise stymied. And underlying this (much as in the case of anti-Internet legislation we noted earlier) is politicians' willingness to play up to people's fears and confusion -- and their loss of trust -- in ways that ultimately may be very damaging to society at large.
Trust is a fundamental aspect of our lives, both on the Net and off. Once lost, it can be impossible to ever restore to former levels. The damage is often permanent, and can ultimately be many orders of magnitude more devastating than the events that may initially trigger a user trust crisis itself.
Perhaps something to remember, the next time you're considering lying to your users in the name of experimentation.
Trust me on this one.