If CFI < .90, you would examine modification indices to correlate error terms within the same factor (e.g., A1 with A2).
Its mathematical algorithms are considered the industry gold standard. lisrel 8.80
You might ask: “Why use software from 2005 when we have R ( lavaan ), Mplus, or JASP?” If CFI <
Comparing models across different demographics (e.g., gender, country). Developed originally by Karl Jöreskog and Dag Sörbom,
Developed originally by Karl Jöreskog and Dag Sörbom, LISREL (an acronym for near S tructural REL ations) was the first software to allow researchers to estimate models with latent variables . Unlike standard regression, which assumes all variables are measured without error, LISREL accounts for measurement error by using multiple indicators to "capture" an underlying concept. Core Capabilities of LISREL 8.80:
Check the Fit Indices . LISREL provides a comprehensive list, including: RMSEA (Root Mean Square Error of Approximation) CFI (Comparative Fit Index) SRMR (Standardized Root Mean Square Residual) Why Use LISREL 8.80 Today?