simsem / semTools

Useful tools for structural equation modeling
75 stars 36 forks source link

First order factors being treated as second order factors in a model with only one second order factor #84

Closed ktmanis closed 3 years ago

ktmanis commented 3 years ago

########CFA MODEL FOR SECOND-ORDER FACTOR########

SPECIFY MODEL

CFA.JAIS.20 <-

  • '
  • SPECIFY FIRST-ORDER FACTORS

  • SV =~ SV2 + SV3 + SV4
  • CG =~ CG2 + CG3 + CG4
  • RI =~ RI2 + RI3 + RI4
  • TRU =~ TRU1 + TRU2 + TRU5
  • COM =~ COM3 + COM4 + COM5
  • KI =~ KI1 + KI2 + KI4
  • KEX =~ KEX3 + KEX4 + KEX6
  • CK =~ CK2 + CK3 + CK5
  • SDC =~ EK2 + EK4 + EK5
  • CR =~ CR1 + CR2 + CR3
  • OP =~ OP1 + OP2 + OP3
  • MP =~ MP2 + MP3 + MP4
  • SPECIFY SECOND-ORDER FACTORS

  • IC=~1KI+1KEX+1*CK
  • IC~~IC
  • RESIDUALS - THESE ARE DETERMINED THROUGH MODIFICATION INDICES - modindices()

  • MP2 ~~ MP4
  • CG2 ~~ CG4
  • EK4 ~~ EK5
  • KI1 ~~ KI4
  • RI2 ~~ RI4
  • OP1 ~~ OP2
  • KEX3 ~~ KEX4
  • ' reliability(fit.CFA.20) Higher-order factors were ignored.
          KI       KEX        CK

alpha 0.6849464 0.8872550 0.7818604 omega 0.8152728 0.8071967 0.8004426 omega2 0.8152728 0.8071967 0.8004426 omega3 0.8422551 0.7122995 0.8631846 avevar 0.5096021 0.6484720 0.5728214

reliabilityL2(fit.CFA.20, "IC") omegaL1 omegaL2 partialOmegaL1 0.7461249 0.8237391 0.8849996 reliabilityL2(fit.CFA.20, "OP") ##first-order factor being treated as second-order omegaL1 omegaL2 partialOmegaL1 0.6768584 0.6768584 0.6768584

TDJorgensen commented 3 years ago

You need to put backticks around your syntax if you want to format your syntax example. Bug now fixed in development version.