I was just wondering if anybody could help me, I'm conducting a path analysis in SPSS Amos with a binary outcome. I've specified the variable as binary using Tools>Data Recode, I've then calculated the variable estimates using MCMC.

My question is: how do I know calculate overall model fit?

The model is as follows, with all variables except psychosis (binary) being continuous.

  • Esteem <--- Bullying
  • Intrusion <--- Bullying
  • Intrusion <--- Esteem
  • PsychosisRisk <--- Intrusion
  • PsychosisRisk <--- Bullying

Many thanks.

  • $\begingroup$ Can you add more details about your data & your model here, maybe paste in some output? I doubt this question is answerable at present. $\endgroup$ – gung - Reinstate Monica Jul 10 '14 at 16:11
  • $\begingroup$ Hello, Yes of course - I have 4 variables. Bullying (continuous), Self-Esteem (Continuous), Intrusive Thoughts (Cont.) and Psychosis Risk (Binary). The following model is being tested: Bullying -> Psychosis Risk, Bullying -> Esteem, Bullying -> Intrusion, Esteem ->, Intrusion->Psychosis Risk. $\endgroup$ – Evelyn Jul 10 '14 at 18:07
  • $\begingroup$ Doesn't Amos give you any fit indices like Chi², CFI, RMSEA, SRMR, etc.? I'd expect it to do that automatically when fitting any SEM... $\endgroup$ – Nick Stauner Jul 10 '14 at 20:55
  • $\begingroup$ Unfortunately it appears not to calculate those for models containing a dichotomous outcome. $\endgroup$ – Evelyn Jul 11 '14 at 11:09
  • $\begingroup$ did you run Bayesian analysis ? there is a way you can do it - let me find the link for a PPT which shows how to calculate R-square I am working on similar problem - I want to know if your outcome is $\endgroup$ – user82020 Jul 11 '15 at 4:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.