[napsterbr]

A/B testing is for commercial purposes, it's wildly different when you are testing the user's psychological state.

[onewaystreet]

But a lot of A/B testing does test a user's psychological state. One could argue that doing it for commercial purposes makes it worse.

[kevingadd]

That's absolutely true. The main difference here is that the experiment's negative impact was unabstracted yet they still went ahead with the experiment. The hypothesis was more or less 'we can make people depressed', without the slightest hint of defensible motives or objectives ('we want to make our ads more effective', 'we want to find the best way to convince customers that they want our product'.)

[yazinsai]

+1. It's all about the end goal.