|
|
More with our
Video examples
|
|
|
|
Return
|
|
|
|
Multivariate Data Analysis
|
|
|
|
Principal Component Anlysis
summarize the associations or correlations of several numeric variables. This
method allows to represent the original data (observations and variables) with
fewer dimensions than with the original data.
- With or without supplementary variables or observations,
- Factorial maps,
- Individual (observation) results, scores, contributions and correaltions,
- Weighted or not,
- Varimax or Quartimax rotation,
- Pearson, Kendall, Spearman coefficient, or covariances,
- Missing data estimation.
|
|
|
|
|
|
|
|
|
|
|
|
Binary Correspondence Analysis is used to study the relations between two
sets of categorical items that make up the rows and the columns of a contingency
table.
- With or without supplementary variables,
- Factorial maps,
- Weighted or not.
|
|
|
|
|
|
|
|
|
|
|
|
Multiple correspondence
analysis study the relation between several categorial variables.
-
With or without supplementary variables or observations,
- Individual (observation) results, scores, contributions and correaltions,
- Factorial maps,
- Weighted or not. |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Multiple regression
explains the variation of one variable according to one or several variables.
The variable to be explained and the explanatory variables must be numeric.
- R2, adjusted R2, regression coefficients and t tests.
- Anova,
- Observation estimations and Cook distance. |
|
|
|
|
|
|
|
Return
|
|