|
|
 |
| EFA Warning bivariate table empty cell |
 
|
| Message/Author |
|
|
|
|
Hi there,
I am a new MPLUS 7 user. I was looking in past discussions for the warning message (below) I received while attempting an EFA using 30 dichotomous variables. The warning message appeared about 58 times in my output file. Your note from October 19, 2011 indicates that "When a bivariate table has an empty cell, this implies a correlation of one. Both variables should not be used in the analysis".
Two questions:
(1) Is doing an EFA with 30 dichotomous variables not a good idea- should I not attempt this?
(2) Is the fact that this is a warning not an error why I am still seeing a full output attached?
WARNING: THE BIVARIATE TABLE OF DSF5A5N AND DSF2B5N HAS AN EMPTY CELL
Thank you,
KT |
|
|
|
|
Hi again,
I am also wondering if the below error message is related to the problem I described in the above post? I received that message for "with 3 factors", "with 4 factors" and "with 5 factors."
FACTOR DETERMINACIES WILL NOT BE COMPUTED FOR EXPLORATORY FACTOR ANALYSIS WITH 3 FACTOR(S). THE MODEL COVARIANCE MATRIX IS NOT POSITIVE DEFINITE
Thank you,
KT Tomey |
|
|
|
|
When a bivariate table has an empty cell, this implies a correlation of one between two variables. This means that the two items are not statistically distinguishable. Both items should not be used in the analysis.
We don't provide factor determinacies for categorical items so I am not sure what that message means. You can send the output and your license number to support@statmodel.com for further information.
Please keep posts to one window on Mplus Discussion. |
|
|
| Back to top |
|
|
