Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Fit indices after -factor-

    Dear all,
    I would like to ask you something.

    I run an exploratory factor analysis, and I would like to have a look at some fit indices (precisely: chi square, RMSEA, SRMR), but I do not find the right command to display them.

    Do you have any clues about it? Thanks a lot, best, G.

  • #2
    These statistics are not created by the -factor- command. Use -sem- to do a confirmatory factor analysis and you will get these statistics.

    Comment


    • #3
      Dear Clyde, could you suggest me how should I use sem to make this analysis? I am not familiar with this package, could you suggest me the right command please?

      Thanks a lot, best, G.

      Comment


      • #4
        Dear Clyde, I make you a more precise question.
        Those are my data:
        ----------------------- copy starting from the next line -----------------------
        Code:
        * Example generated by -dataex-. To install: ssc install dataex
        clear
        input byte(sodd_general sodd_guad sodd_secur sodd_numore sodd_tipore sodd_ambient sodd_distance sodd_interest) float(classe_dominanza male titstud3 sud)
        10 10 10 10 10 10 10 10 5 1 1 1
         8  8  8  7  9  6  8  7 2 1 2 1
        10 10 10 10  8 10 10 10 3 1 2 1
        10  4  9  3  3  9 10 10 4 1 1 1
         6  6  7  6  6  7 10  7 2 1 2 0
         8  7  7  8  6  8  7  9 5 1 2 0
         8  5  1  3  4  5  7  8 6 1 1 0
         7 10  7 10 10  8 10 10 1 1 1 1
         8  8  8  8  8  8 10  8 5 1 2 0
        10 10 10 10 10 10  6 10 2 1 2 0
        end
        label values classe_dominanza classe_dominanza
        label def classe_dominanza 1 "bor", modify
        label def classe_dominanza 2 "cmi", modify
        label def classe_dominanza 3 "pbu", modify
        label def classe_dominanza 4 "pba", modify
        label def classe_dominanza 5 "cou", modify
        label def classe_dominanza 6 "coa", modify
        label values male male
        label def male 1 "Uomo", modify
        label values titstud3 titstud3
        label def titstud3 1 "nessun titolo o media", modify
        label def titstud3 2 "superiori", modify
        ------------------ copy up to and including the previous line ------------------

        I tried to estimate a promax factor analysis with the command:

        factor sodd_general sodd_interest sodd_guad sodd_secur sodd_numore sodd_tipore sodd_ambient sodd_distance , factors(4) pf
        rotate, promax

        and the, by the sem package, with the following command:

        sem (L1 -> sodd_general sodd_interest) (L2 -> sodd_guad sodd_secur) (L3 -> sodd_numore sodd_tipore) (L4 -> sodd_ambient sodd_distance)

        the results are totally different, why? This means that factor and sem make two different things, or what?

        Thanks a lot, G.

        Comment


        • #5
          The basic factor model is inherently unidentified. (That is why it is even possible to rotate factor solutions.) -factor- and -sem- identify the model in different ways, so they produce different solutions.

          I did not mean to imply in my first post that the RMSEA etc. produced by -sem- are applicable to the output of -factor-. I was trying to say that you can't get those after -factor-, and you have to use the confirmatory factor analysis of -sem- to get such statistics at all.

          Comment

          Working...
          X