Standardized residuals, (observed - expected) / sqrt(V)
, where V
is the residual cell variance (Agresti, 2007, section 2.4.5). Special care is taken
for multiple-response variables which are in effect a series of separate tables where
‘not selected’ cells for each item are are hidden.
# S4 method for CrunchCube
rstandard(model)
A CrunchCube representing a contingency table
an array of standardized residuals or Z-scores from the hypothesis being tested. The default method is that the joint distributions of (weighted) counts are equal to the marginal distributions of the table.
Agresti, A. (2007) An Introduction to Categorical Data Analysis, 2nd ed., New York: John Wiley & Sons. Page 38.