Generalized analysis of variance
Generalized analysis of variance (GANOVA) helps to assess extent to which a selected numeric response variable is influenced by (a) qualitative factors (as name of operator, workshift, number of production line, raw material supplier, etc.) and/or (b) quantitative numeric variables (as temperature, pressure, mixer revolutions). The response variable of interest may typically be a quality parameter or any process or experimental output. As such, this module is useful for spotting how to influence or stabilise an important output parameter, to identify and prove influence of certain factors on important output and consequently eliminate or stabilise the influential factors to stabilise or improve quality.

PDF Generalized analysis of variance - Pdf manual

Output of analysis summary

Text output Graphical output
Overall ANOVA
Source of variability
Degrees of freedom
Squares sum
Variances
F-statistic
p-value
Significance
Total variability
Explained variability
Residual variability

ANOVA for individual
Predictors
Regression parameters
Squares sum
F-statistic
p-value
Significance analysis


Table of prediction
Predictor values
Confidence interval of prediction
Y-prediction plot
Residuals plot
Predicted residuals
Hat diagonal plot
Cook distance diagnostics
Partial prediction plot
Box plot
Plot of means

Main input dialog panel of Generalized ANOVA module
AnovaN

Typical form of the input data, anova module can handle both numerical and factorial variables that explain one response variable.
AnovaN

Graphical output, regression diagnostics will indicate important information and possible problen in data and model.
AnovaN