tcarey

Forum Replies Created

Viewing 1 post (of 1 total)
  • Author
    Posts
  • in reply to: *Please Help* Days Nested Within Students #2088

    tcarey
    Participant

    Thank you Dr. Mehta, your suggestion worked!

    Although the model is now running, I am still struggling to figure out how to define my lambda matrices (within-level and cross-level). I have three latent factors at both level 1 and level 2. Each of these latent factors are connected to 18 observed indicators, with 6 indicators presumed to load on factor 1, 6 on factor 2, and 6 on factor 3. I am trying to test a CFA to confirm this factor structure.

    When I specify the lambda matrices I do so as follows:

    ##Level 1 lambda loadings
    Lev1_pat <- matrix(c(0,1,1,1,1,1,0,1,1,1,1,1,0,1,1,1,1,1),18,3)
    Lev1_pat
    Lev1_val <- matrix(c(1,.4,.9,.8,.4,.6, 1,.8,.5,.8,.4,.6, 1,.3,.7,.5,.9,.9),18,3)
    Lev1_val

    ##Cross level lambda loadings
    ly12_pat <- matrix(c(0,1,1,1,1,1,0,1,1,1,1,1,0,1,1,1,1,1),18,3)
    ly12_pat
    ly12_val <- matrix(c(1,.4,.9,.8,.4,.6, 1,.8,.5,.8,.4,.6, 1,.3,.7,.5,.9,.9),18,3)
    ly12_val

    In each case, the resultant output estimates 48 lambdas when I only want 18 estimates (6 estimates per latent variable). Any advice you have about how I might go about altering my matrix specification would be much appreciated. All of my other matrices appear to be fine (nu, alpha, theta, psi) …once I get the lambdas sorted out, I think it will be ready to go!

    Thank you so much for all of your help!

Viewing 1 post (of 1 total)