|
|
Matching Mplus Output to APA Reportin... |
|
Message/Author |
|
|
Hi there, trying to report the results of my multilevel analysis according to APA standards (p. 147f. in the 2009 publication manual), I would like to know where to find the variance-covariance estimates (reported at the bottom half of the table in the manual). Under the heading "Random parameters / Level 2", they list * Intercept/intercept (sigmasq{a0}) * Age/age [one of the predictors] (sigmasq{a1}) * Age/intercept (sigmasq{a10}) ... Then, under the heading "Random parameters/Level1": * Intercept/intercept (w{0}) * and the -2LL [at least, I found that one; this corresponds to the "deviance," if I am not mistaken?] Maybe it's just a problem of notation. Sorry if I am overlooking something completely obvious. Thanks, Tanya |
|
|
I mean, where can I find it _in the Mplus output_. Thanks. |
|
|
It sounds like you have a model with a random intercept and random slope similar to Example 9.2. MODEL: #WITHIN% s | y ON age; %BETWEEN% s y; s WITH y; where on within you will get an estimate of the residual variance of s and on between you will get the variance of the random slope s and the random intercept y and their covariance. |
|
|
Hi Linda, thanks for your help. Actually (and this is where I got stuck), the APA manual mentions an intercept/intercept covariance at level 2 (sigmasq_a0) as well as at level 2 (w_0) from the null model on, and I am not sure why they term it a random factor. In addition, they do not seem to use the standard indices (they use an a instead of ... u? r?), which makes it a little confusing. If you do not have the manual at hand, I could send you the table in question. Sorry if my problem is probably completely straightforward, and I'm just confused by the lack of explanation in the APA table note. Thanks, Tanja |
|
|
I would look in the Raudenbush and Bryk book and also at a published article that uses the APA standards. |
|
Back to top |
|
|