In a multivariate predictive model, what is the name of the error term's sum of squares and cross-products matrix, representing unsystematic variance?

In a multivariate predictive model, what is the name of the error term's sum of squares and cross-products matrix, representing unsystematic variance?

• A. Eta squared (η2)
• B. Exp(B)
• C. Error SSCP (E)
• D. F-statistic

Answer: (C)

In multivariate predictive models, total variability across the dependent variables is captured by a sum-of-squares-and-cross-products (SSCP) matrix, which can be split into what the model explains and what remains unexplained. The matrix that represents this unexplained, unsystematic variance is the error SSCP, typically denoted E. This matrix summarizes the residual variability after fitting the predictors and is central to multivariate tests because test statistics compare the model’s explained variance to this unexplained portion (for example, through Wilks’ lambda and related measures). Eta squared is an effect-size measure of how much variance the model explains, Exp(B) is the exponentiated coefficient used in logistic contexts, and the F-statistic is a test statistic comparing mean squares. The error SSCP is the one that specifically encodes the unexplained variance in the multivariate setting.

In multivariate predictive models, total variability across the dependent variables is captured by a sum-of-squares-and-cross-products (SSCP) matrix, which can be split into what the model explains and what remains unexplained. The matrix that represents this unexplained, unsystematic variance is the error SSCP, typically denoted E. This matrix summarizes the residual variability after fitting the predictors and is central to multivariate tests because test statistics compare the model’s explained variance to this unexplained portion (for example, through Wilks’ lambda and related measures).

Eta squared is an effect-size measure of how much variance the model explains, Exp(B) is the exponentiated coefficient used in logistic contexts, and the F-statistic is a test statistic comparing mean squares. The error SSCP is the one that specifically encodes the unexplained variance in the multivariate setting.