Message/Author |
|
|
Dear Mplus team, RE: MODEL = SEQUENTIAL or COVARIANCE I am imputing under H1 (saturated) time series data using TWOLEVEL BASIC a dataset with 10 variables collected at 3 time waves - all 10 variables are categorical. Each time wave also has 3 covariates that are continuous - these covariates are used to assist in the imputation of the 10 categorical variables but are not imputed themselves. So categorical and continuous variables assist in the imputation but only the categorical variables are imputed. The data are arranged in WIDE format (ie., 13 x 3 = 39 columns) to preserve sources of measurement non-invariance. Normally I might use MODEL = COVAIANCE; however, I am unsure how best to proceed when some of the covariates (albeit not imputed) are measured on a different scale. Might MODEL = SEQUENTIAL be appropriate in this instance? Many thanks - Jonathon |
|
|
Double posting answered elsewhere. |
|
|
Um? - I only posted this question once, are you confusing this with the other question I posted earlier today about influence of estimator in MI? I couldnt find an earlier post by anyone else on this exact topic. |
|
|
My mistake. Answer follows. |
|
|
Jonathon - the difference in the scale should not be a problem at all. As you say there is no assumption of time invariance in the imputation model. The variables are treated as 9 different variables that help impute 30 categorical variables. |
|
|
Thankyou Tihomir and Bengt |
|
mboer posted on Monday, March 11, 2019 - 7:47 am
|
|
|
Dear Prof. Muthen, I want to impute data with 14 categorical variables and 1 continuous variable, and i declared variables as categorical by specifying a (c) after the respective categorical variables. So I specified this like this: USEVARIABLES = female fasiii age_n IMS m82 m83 m84 m85 ! fam supp m112_r m113_r m114_r m115_r m116_r m117_r bullied d2bullied; DATA IMPUTATION: IMPUTE = fasiii age_n (c) IMS (c) m82 (c) m83 (c) m84 (c) m85 (c) m112_r (c) m113_r (c) m114_r (c) m115_r (c) m116_r (c) m117_r (c) bullied (c) d2bullied(c); values = fasiii(0-13); In the analysis part, I specified the following: analysis: TYPE=BASIC; ESTIMATOR=MLR; My questions are (1) are the imputations now done with MLR estimation / or is bayes used by default? (2) does this specification generate H1 imputations? (3) according to the output, the 'covariance' setting was applied with the imputations, but how is this possible since the imputation involves categorical data? Thank you in advance. |
|
|
The imputations are based on Bayesian estimation using the H1 model. The H1 model estimates an unrestricted polychoric/polyserial correlation matrix. You can find more information here http://statmodel.com/download/Imputations7.pdf |
|
Back to top |