Quote:
Originally Posted by patriotpa
In engineering, one runs what is called a monte carlo analysis. You caculate all tolerances and look for best and worst case, then design for the center of that range. That way, if tolerances stack up against you, you still fall within acceptable range.
|
We also run those in social-behavioral research to generate structural equations models that do the "best job" at explaining or fitting the data. But, have you noticed that very slight differences in the data that MCA is using can lead to totally different outcomes. And, I think you were spot on in your last sentence.