>>11048818>>99% of research is survey based, asking people how they feelThere is literally nothing wrong with this.
You want to study how human's behave, what makes them happy, what makes them deranged, why they act like they do. So you design and control the stimuli you apply to a large number of human minds and ask them how it makes them feel, how they behave, or how they say they would behave. Then you do some statistics and try to make a model to describe all human minds in this situation as best you can.
If you can't fully control your stimuli (maybe infeasible, maybe unethical) then go out into the world and look for poor humans who have had whatever stimuli forced upon them and ask them how they feel, how they would behave in certain scenarios.
Maybe your model has shit predictive power, try a new one see if you can do better. Try getting more data.
Maybe it only produces good predictions in specific contrived scenarios, that's no different from specific phenomenological models in other scientific fields.
Maybe your model's predictive power is limited by the nature of what you're trying to study. For example given hypothetical scenarios people might unintentionally give dishonest responses. That doesn't mean that you didn't use the scientific method to guide your hypothesis before testing it. That doesn't stop you from trying to improve your investigation, like turning your hypothetical scenarios into real ones. It doesn't stop your model from potentially helping people with mental illnesses, or helping people overcome mental challenges.
