

Wald test for mean equality 

Message/Author 

John Dziak posted on Tuesday, March 30, 2010  3:52 pm



I wanted to ask about your technical report, "Wald Test of Mean Equality for Potential Latent Class Predictors in Mixture Modeling." In general I found it very helpful, but I had a question about the the third equation from the top of the second page. This equation gives the correct formula for the "sample variance" (squared standard deviation) but it seemed from the context that what was really desired was the "sampling variance of the mean" (squared standard error) which I think would entail squaring the n in the denominator. It seemed to me that the test would not have very much power otherwise. Maybe I missed something though. Perhaps you could have a look at it sometime? Thanks very much, John Dziak 


You are definitely correct. Thank you for pointing this out. It will be corrected in the next version. 


Just to clarify, it will be corrected in the next version of the writeup. It is correctly coded in the program. 

John Dziak posted on Monday, April 12, 2010  4:23 pm



Thanks very much! 

Back to top 

