The only way I see of alleviating some of the problems you mention is by controlling for them and doing further experiments. My experiment was not intended to

The only way I see of alleviating some of the problems you mention is by controlling for them and doing further experiments. My experiment was not intended to measure evilness but rather to give an example of a general approach to the problem, and was for measuring malicious intent. Regardless, given time I'm sure we could come up with much better experiments for studying both. You're right though at what you are getting at, and I think the underlying idea here is that there is a continuous scale from bad (0) to good (1) for any given action/policy, where given additional context our internal moral barometer can move across this domain as well as an independent societal human flourishing barometer which measures the broader repercussions of the action/policy. Honestly, I do not know enough about ethics or research in this general area to have a truly insightful conversation on this topic. I'd just like to add that 1) I think good and evil are so broad in their scope that we'd be virtually forced to break them into separate definitions and measurements and study these constituent parts if we wanted to understand the general concepts and 2) we'd need to make a clear distinction between what humans perceive as good/evil in each context vs. what has good/evil repercussions at different levels (individual/familial/societal) based on measurements of brain states, pain/suffering, and human flourishing.