one of the seven variables in my path model is a binary one, with yes/no answers, and this variable is an exogenous one. what estimation method should I use in Mplus? Is Maximum Likelihood good to go with?
Thanks for clearing it out for me.
The choice of estimator is more critical when endogenous variables are non-normal, so you should be fine with ML, although I would consider alternatives providing robust chi-square test statistics, like MLR, MLM, MLMV.
If you want to use MLM (Satorra-Bentler) you will have to exclude cases with missing data in any of the variables in your model. This means you would be assuming the missing data mechanism is Missing completely at random (MCAR).
Therefore, I would suggest you use Robust Maximum Likelihood (MLR in Mplus). This will handle the missing data problem assuming Missing at random (MAR), and you will not have to exclude cases with missing data.
Greetings,
Roberto
Roberto Melipillan thank you. The information you provided is so much helpful.
I have understood that MLR is recommended for small to medium sample sizes, while mine is 665, and compared to this, the number of cases containing missing data is negligible. Considering these very few cases, I have previously deleted them in SPSS. Is this kind of deletion acceptable? and is MLR still preferable to MLM in this case?
thank you for your time,
Hi Saeed,
MLR can be used with any sample size, and because MLR can handle missing data (assuming MAR), I would recommend you not to delete cases with missing data and use MLR rather than MLM.
Best,
Roberto
I would second Roberto's advice that deleting cases with missing data is practically never a helpful approach.
See, e.g.:
Article Missing Data
Hi Roberto Melipillan,
If I understand correctly, cases with missing data will not be excluded. Any idea of how MLR handles the missing value?
Thanks.
Hi Chee-Seng Tan,
MLR implements a ML estimation approach (e.g., EM) using all available data. Therefore, MLR doesn't:
1) delete incomplete cases (e.g., "listwise deletion"),
2) impute data (e.g., "multiple imputation").
Roberto
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis