|
 |
Problems minimizing the function |
 
|
Message/Author |
|
Roger Brown posted on Wednesday, December 06, 2006 - 5:19 pm
|
|
|
Dear Bengt and Linda, I have been doing some psychometric work on some scale, starting with EFA's and encountered the following error: PROBLEM OCCURRED IN EXPLORATORY ANALYSIS WITH 3 FACTOR(S). THE PROGRAM HAD PROBLEMS MINIMIZING THE FUNCTION. TOO MANY FACTORS HAVE PROBABLY BEEN USED. My analysis: type = efa 2 4; Sample size 215 with 14 categorical variables. I get a 2 factor and a 4 factor solution, but no 3 factor solution. I have tried to change start values, but I guess the EFA routine will not allow this. Any suggestions? Thanks. Roger Brown UW-Madison |
|
|
This can happen. It is most likely a Heywood case. The only option is to change estimators to see if that helps. |
|
Roger Brown posted on Thursday, December 07, 2006 - 8:54 am
|
|
|
Thanks, changed to a WLS estimtor and was able to obtain a 3 dimensional solution that makes sense. Roger |
|
Back to top |
|
|