Message/Author |
|
Isok Kim posted on Monday, June 03, 2013 - 3:39 pm
|
|
|
Hi, I modeled a path analysis involving multiple mediators with continuous outcome var. I requested STDYX output, but it only shows standardized estimates without the S.E., est/S.E., or p-value columns. Below is my syntax: ANALYSIS: BOOTSTRAP = 5000 ; MODEL: rde ON cea ; rc2pe ON rde age ; rc2ee ON rde age ; rc2pd ON rde ; rc2ed ON rde ; cesd ON rde rc2pe rc2ee rc2pd rc2ed ; rc2pe WITH rc2ee rc2pd rc2ed ; rc2ee WITH rc2pd rc2ed ; rc2pd WITH rc2ed ; MODEL INDIRECT: cesd IND rc2pe rde ; cesd IND rc2ee rde ; cesd IND rc2pd rde ; cesd IND rc2ed rde ; OUTPUT: SAMPSTAT STANDARDIZED(STDYX) CINTERVAL(BCBOOTSTRAP) ; Am I missing something? thanks, isok |
|
|
We don't give standard errors for standardized coefficients when the BOOTSTRAP option is used. |
|
Isok Kim posted on Tuesday, June 04, 2013 - 9:54 am
|
|
|
Thanks Linda, I was confused because output for indirect effects had SE, est/se, and p-value columns shown, but not for the direct effects. Running the syntax without the bootstrap option DOES give me all of the columns! Thanks, isok |
|
|
I am using the model indirect command and my output only includes total indirect and specific indirect effects - no direct effects or total effects are provided in the output. How can I obtain the direct effects and total effects? |
|
|
See the last part of the MODEL INDIRECT description on page 691 of the Version 7 user's guide. |
|
|
Great, I found the answer there. I ran it without specifying the mediator (e.g., y3 IND x1) and got the full output I was looking for. Thanks so much. |
|
Back to top |