Strange Social Phenomenon

[Alexander]

I'm sorry, are you suggesting that it is scientific to have a desired outcome and to only test people who are likely to give you your desired outcome? Because I hope not. You test a hypothesis by testing to both expected strengths and expected failure.

It's scientific to have a desired set of controls and conditions. The outcome isn't intended, the context in which to observe the outcome is. (fucking duh)

Doing what you suggested would fail to be analogous to the original experiment and would only confound results.