Invariance across type
A multi sample confirmatory factor analysis (CFA) was conducted on the ratings of
normal students vs. Students with learning disability. The model was not of a good fit (2 (568)= 1321.84, CFI = .87; RMSEA=.057) when no constraints of equality were assumed. Equality constraints were then assumed on factor loadings. The results showed that the model with factor loadings invariance was marginally good (2 (588) =1454.19 CFI = .86, RMSEA=.06).The chi square and CFI indicate a poor fit however, RMSEA indicates a good fit. The other constraints of equality across the two types (normal vs. Learning disability) revealed that intercepts,structural covariances and measurement residuals are not invariant (equal).