Firstly:
I want to run a CFA with 7 latent variables and 39 observed variables.
For a guide of how I constructed my CFA, I used this Youtube Video:
http://www.youtube.com/watch?v=JkZGWUUjdLg
It looks to me to be exactly correct according to the process explained in the above mentioned clip. However the CFA will not run, and the output reads this message: "The model is probably unidentified. In order to achieve identifiability, it will probably be necessary to impose 15 additional constraints".
My first question is, why is the CFA not running, and why does it say I need to add 15 more constraints? To my understanding this CFA is correct.
Secondly:
I am running a SEM with 9 latents and 49 observed variables. This model will also not run, and when I check the output it reads: "Iteration limit reached, the results that follow are therefore incorrect". I do not understand what this means.
Does anyone know what is going wrong with this SEM?
A calculation of the sample size needed suggests I need 289 participants. I have 325 participants, although they are composed of a smaller data set multiplied by 3 or 4 times to reach sufficient sample size (could this be an issue?).
Best Answer
The last question can be peeled off. If I have understood you correctly, you have just cloned the data to get a big enough sample size. That makes no sense at all. Even if you can fool the software, it is still statistical nonsense. Think of it this way: if it is valid, why isn't it prominent in every statistical text as a simple to understand, low technology way of fixing many data problems, such as large standard errors or disappointing P-values?
The earlier questions are outside my area of expertise. Unfortunately, much of your post boils down to your thinking that you asked something valid, but the software doesn't agree, so why is that? Saying that you watched a video and did something similar is not informative. People using the same software will almost certainly want to know exactly what instructions you gave the software.