The correlation structure between two blocks of random variables can be summarized by their canonical correlation coefficients and corresponding canonical variates. If these two blocks of variables are the manifest variables of a LISREL model, the canonical correlations can be related to the path coefficients and the canonical variates to the latent factors of the model. If the number of manifest variables becomes large, the canonical correlations approach the path coefficients and the canonical variates come close to the latent factors.
- Latent Factor
- Path Coefficient
- Canonical Correlation
- Canonical Correlation Analysis
- Canonical Variate
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.