Scribd
Upload
88 views

SEM Fit

This document discusses fit statistics commonly reported for confirmatory factor analyses and structural equation models. It provides a table listing several popular fit statistics used to assess model fit, including chi-square, GFI, NFI, CFI, RMSEA, RMR, and AVE. It recommends cut-off values for each measure that indicate good model fit. The document also notes that Kline suggests reporting chi-square, RMSEA, CFI, and SRMR values at a minimum.

Uploaded by

Jan Winczorek
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
88 views

SEM Fit

This document discusses fit statistics commonly reported for confirmatory factor analyses and structural equation models. It provides a table listing several popular fit statistics used to assess model fit, including chi-square, GFI, NFI, CFI, RMSEA, RMR, and AVE. It recommends cut-off values for each measure that indicate good model fit. The document also notes that Kline suggests reporting chi-square, RMSEA, CFI, and SRMR values at a minimum.

Uploaded by

Jan Winczorek
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Fit

Statistics commonly reported for CFA and SEM



There are more than a dozen different fit statistics researchers use to assess their confirmatory
factor analyses and structural equation models. Here we have assembled a list of the most
popular fit statistics used and recommended cut-offs that indicate a good fit.

Measure Name Description Cut-off for
good fit
Χ2 Model Chi- Assess overall fit and the discrepancy between p-value>
Square the sample and fitted covariance matrices. 0.05
Sensitive to sample size.
H0: The model fits perfectly.
(A)GFI (Adjusted) GFI is the proportion of variance accounted GFI ≥ 0.95
Goodness of Fit for by the estimated population covariance. AGFI ≥0.90
Analogous to R2. AGFI favors parsimony.
(N)NFI (Non) Normed- An NFI of .95, indicates the model of interest NFI ≥ 0.95
TLI Fit Index improves the fit by 95% relative to the null NNFI ≥ 0.95
model. NNFI is preferable for smaller samples.
Tucker Lewis Sometimes the NNFI is called the Tucker Lewis
index index (TLI)
CFI Comparative Fit A revised form of NFI. Not very sensitive to CFI ≥.90
Index sample size. Compares the fit of a target
model to the fit of an independent, or null,
model.
RMSEA Root Mean A parsimony-adjusted index. Values closer to RMSEA <
Square Error of 0 represent a good fit. 0.08
Approximation
(S)RMR (Standardized) The square-root of the difference between SRMR <0.08
Root Mean the residuals of the sample covariance matrix
Square Residual and the hypothesized model. If items vary in
range (i.e. some items are 1-5, others 1-7)
then RMR is hard to interpret, better to use
SRMR.
AVE Average Value The average of the R2s for items within a AVE >.5
(CFA only) Explained factor

Stephen Parry
Kline suggests that at a minimum the following indices should be reported:
1) The model chi-square
2) RMSEA
3) CFI
4) SRMR


How to estimate these fit indices:
• In R, use the FitMeasures function from the lavaan package.
• In SAS’s Proc Calis, specify the fitindex option with the particular indices you want.
• In Stata, after executing a CFA or SEM, use the command: estat gof, stats(all)


References:
Principles and Practice of Structural Equation Modeling. Rex B. Kline. 2005.
Structural Equation Modelling: Guidelines for Determining Model Fit. Daire Hooper, et al. 2008.

Stephen Parry

You might also like