This is an introductory description of how to use control variables

There is an inherent assumption that the distribution of sample means is always normal. Also, we have never checked if the formula for standard error makes sense.

The MBI example in the Byrne text book on SEM with MPlus has been solved using Lavaan in [R]. The values obtained are different from those in text.

Verify by simulation that dividing by n-1 gives a better approximation of the population SD.