I am new to MPLUS but comfortable with HLM and SEM. I am preparing to do a two-level CFA for some scales. (The levels are classrooms nested in teachers.) Theory suggests that there should be three factors at the within teacher level but none at the between teacher level. Can I model this by simply letting the intercept of each indicator in the within model be a latent variable in the between model without specifying that they are indicators of a between level factor? Something like:
Analysis: TYPE = TWOLEVEL MODEL: %WITHIN% PeerInv BY m_a66 m_b157 m_a65; PeerTeas BY m_b9 m_b24
Hi I just read the above message. I am also doing multi-level, except I have 30 different year observations for each company (also around 30 companies). I am conducting a CFA analysis in order to figure out whether 3 measures load together. Then I will run this variable as an IV to test the relationship between IV and firm performance in a panel setting. These 3 measures change within the company over time. Does that mean I am interested in the within variable? Can I use the same command that question above used? Do I need to specify cluster command? I have time variable (0 2 4...), how do I include this into the command? Many thanks!
Analysis: TYPE = TWOLEVEL MODEL: %WITHIN% PeerInv BY m_a66 m_b157 m_a65;