laughing dog
Contributor
t
You are basically holding this type of research to an impossible standard in order to dismiss it results.
In social science research, it is usually impossible to explicitly control for all possible influences even when conducting an experiment, let alone trying to tease out statistical significance from data collected in the past. Good research (which does exist) does acknowledge its shortcomings. While I am not expert or very familiar with discrimination research, your glib dismissal does not accurately portray it.No. As you say, control for all possible influences. With discrimination "research" there's almost always no attempt to control at all. It's just documenting the difference and pretending that proves something.Technically, LP is correct - disparate outcomes do not PROVE anything. Disparate outcomes provide evidence not proof.
Of course, disparate outcomes are used all the time as evidence to rebut any theory in any field. It is the basis for the scientific method: set up an experiment that controls for all possible influences but one and look for disparate outcomes. Taking LP's claim literally, he is denying the validity of the scientific method.
You are basically holding this type of research to an impossible standard in order to dismiss it results.
I asked you first.
Where the heck am I supposed to have said it isn't remotely common at all? I haven't offered an opinion on its frequency; and I'm not sure why I should be required to have one in order to comment on whether high frequency is a legitimate requirement before action is taken. The demand to show high frequency appears to be rhetoric intended to transfer burden-of-proof; burden-of-proof is something that shouldn't be transferred without showing cause.