Rachel posted on Tuesday, November 01, 2016 - 11:39 am
Dear Dr. Muthen,
When I try to model an interaction between 2 latent variables, rather than appear as a dot between the 2 & my endogenous variable, it appears on the path diagram as a 3rd latent variable, in line with the 2 variables used to create it. 1) Is this reflecting that the model is doing something different than what I intend? 2) If so, how should I change my code (below) to get the interaction to appear as a dot between the 2 exogenous variables?
ANALYSIS: TYPE = RANDOM;
ALGORITHM = INTEGRATION; MODEL:
EXUBER54 by f55aq9 f55aq37 f55aq44 f55aq59 f55aq64 ; F55AQ44 WITH F55AQ37; F55AQ59 WITH F55AQ44 ;
EFFCONT by f55a11r f55aq29 f55a35r f55a40r f55aq54 f55aq72 f55a15r f55a19r f55aq47 f55a66r f55a78r; F55AQ54 WITH F55A11R ; F55A78R WITH F55A66R ; F55A11R WITH F55AQ47 ; F55A40R WITH F55A35R ; F55AQ72 WITH F55AQ29 ; F55AQ29 WITH F55A19R ;
SUM_WJR ON EXUBER54 EFFCONT; EXUxEFTC | EXUBER54 xwith EFFCONT; SUM_WJR ON EXUxEFTC;
I've tried to model it based on the code in page 71 of the user guide, but instead of my interaction appearing as a dot between the 2 variables and the endogenous, it's appearing as a 3rd latent variable also predicting it.
Rachel posted on Tuesday, November 01, 2016 - 1:46 pm