A high Cronbach’s alpha does not prove a scale is unidimensional.

Evidence

We propose a different method to deal with heterogeneity of CN that combines exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). EFA employs a semi-supervised statistic learning method to detect possible latent factor structures, while CFA employs a supervised method. CFA also allows scholars to make testable statistic decisions based on statistical significance testing (Likelihood-Ratio test) to choose a modeling structure that best fits the data structure. — Ruiqian Li and Paul Froese, The Duality of American Christian…, 93 (#builds-on, via linked claim)

Categorization

• XREF: connects to: Combining EFA and CFA tests both exploratory structure and confirmatory fit of CN dimensions. — Ruiqian Li and Paul Froese, The Duality of American Christian…, 93

Related

• Combining EFA and CFA tests both exploratory structure and confirmatory fit of CN dimensions. (similarity 0.93)
• Scales with high Cronbach-α may still conceal several distinct latent factors. (similarity 0.83)