Correlation - Page 3

Other Correlation Coefficients

Spearman rank-order correlation coefficient -- Spearman r

Used when data are ordinal. Interpreted like a Pearson correlation.

Eta coefficient -- h

Indicates the degree of relationship between two variables even if the relationships is nonlinear.

Eta-squared coefficient -- h²

Indicates how well one variable can be predicted from another even with nonlinear relationships. Interpreted in a manner similar to r².

Biserial correlation coefficient -- r_bi

For use when one variable is continuous and the other is a dichotomous variable that reflects an underlying normal distribution.

Point biserial coefficient -- r_pb

For use when one variable is continuous and the other is a 'true' dichotomous variable.

Phi coefficient -- f

For use with two 'true' dichotomous variables.

Tetrachoric correlation coefficient -- r_tet

For use with two artificial dichotomous variables with underlying normal distributions.

Multiple correlation coefficient -- R_a.bcd

The correlation between a and the set of variables b, c, and d.

Squared multiple correlation coefficient - Coefficient of Determination -- R²_a.bcd

The squared correlation between a and the set of variables b, c, and d. It represents the proportion of variability of a that is accounted for by the combination of b, c, and d.

Partial correlation coefficient -- r_ab.c

The correlation between a and b with variable c partialed out.

Graduate School of Education & Information Studies

updated