I am attempting to test three different factor structures for 34 items.
When I specify a bifactor model (one general and one specific factor) at Level 1, I can find the same model at Level 2.
When I specify a seven-factor model at Level 1, the same model or the simpler bifactor model will not converge at Level 2. Error messages indicate that this is usually because there is no more Level 2 variance to be explained for one or more items.
Could the L1 factor structure influencing the amount of L2 variation for the items? Or, is this more likely a consequence of misspecification.