Sample size with observed variables PreviousNext
Mplus Discussion > Structural Equation Modeling >
 Nick posted on Monday, April 25, 2016 - 7:41 pm
I am running a mediation analysis with all observed variables. A saturated model, as is usually the case. The sample size is 144, however, with 34 free parameters it doesn't quite fit the "rule of thumb" or 5:1 ratio of subjects to parameters often thrown around. I understand this rule of thumb guideline is slightly arbitrary, but I cannot find much sample size information regarding observed variables. Does N=144 seem ok for 34 parameters with only observed variables? The only thing of interest is essentially the bootstrapped indirect effects.

Thank you.
 Bengt O. Muthen posted on Tuesday, April 26, 2016 - 5:59 am
I would guess that using bootstrapping there is a good chance 144-34 is ok.

You could do a Monte Carlo simulation study to check, using bootstrapping.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message