Confirmatory Factor Analysis

1
Confirmatory Factor Analysis

• INCO 601
• Scott Titsworth

2
Differences Between EFA and CFA

• Exploratory
• Determines the number of factors
• Determines whether factors are correlated or not
• Variables free to load wherever
• Theory Generating

• Confirmatory
• Number of factors set a priori
• Factor intercorrelations set a-priori
• Variable loadings fixed a-priori
• Theory Testing

3
(No Transcript)
4
The Advantage of CFA

• The Measurement Model includes error
• Random Error
• Error from Unreliability
• When used in Structural Equation Modeling the
  entire measurement model can be included

5
Assessing ModelsAbsolute Fit Indices

• Chi Square An insignificant chi square is
  indicative of good fit.
• The chi square is sensitive to large sample size
  (Ngt100) and may be a poor estimation of fit if
  significant.
• X2/DF Values less than 5 indicate good fit.
• This test is equally susceptible to problems of
  larger sample sizes.

6
Assessing ModelsIncremental Fit Indices

• SRMR Target of .09 or less.
• Tucker-Lewis Index Target of .95 or greater.
• RMSEA Target of .06 or less.
• When comparing two models that have adequate fit,
  more subjective standards like explanatory power
  and parsimony.

7
Assumptions Behind SEM

• Sample Size A conservative estimate is 15 cases
  per observed variable. A liberal estimate is 5
  cases.
• Adequate Identification SEM requires some
  specification of values to provide sensible
  output.
• Outcome Variables Continuous and normally
  distributed.
• Missing Data Addressed Listwise deletion is less
  than 5 of cases.
• Model Theoretically Justified The SEM must have
  theoretical justification.