The thing of it is, there is no evidence that the phenomenon is repeatedly predictable until the factors controlling the outcome have been identified through causality or at the very least a strong correlation. Again, that requires more than a few experiments by advocates.
That's why I used the eye color example. If the study is ostensibly designed to demonstrate that eye color is an important predictor of outcome, a careless or zealous researcher may put too much weigh on eye color as the determinant and ignore other factors which may have played a bigger role. For example, if the experiment is on one classroom or one school, there may be other factors at play such as teacher skills, parental involvement, parental educational attainment, income and stability of home life, earlier educational interventions, etc.
Unless those potential sources of bias are identified and examined as part of the analysis, the conclusion that eye color segregation produces better outcomes may be totally wrong. That's why having advocates handle research is a prescription for disaster. The advocate thinks that the eye color segregation worked and promotes it as proven when in fact the data do not support that conclusion.
The thing of it is, there is no evidence that the phenomenon is repeatedly predictable until the factors controlling the outcome have been identified through causality or at the very least a strong correlation. Again, that requires more than a few experiments by advocates.
= new reply since forum marked as read