Simple Correspondence Analysisoverview how to example data see also
Simple Correspondence AnalysisStat > Multivariate > Simple Correspondence Analysis
Simple correspondence analysis helps you to explore relationships in a two-way classification. Simple correspondence analysis can also operate on three-way and four-way tables because they can be collapsed into two-way tables. This procedure decomposes a contingency table in a manner similar to how principal components analysis decomposes multivariate continuous data. An eigen analysis of the data is performed, and the variability is broken down into underlying dimensions and associated with rows and/or columns.
Input Data
Categorical variables: Choose to enter the data as categorical variables. If you do not use the Combine subdialog box, enter two worksheet columns. The first is for the row categories; the second is for the column categories. Minitab then forms a contingency table from the input data.
Columns of a contingency table: Choose to enter the data as columns of a contingency table. Each worksheet column you enter will be used as one column of the contingency table. All values in the contingency columns must be positive integers or zero.
Row names: Enter a column that contains names for the rows of the contingency table. The name column must be a text column whose length matches the number of rows in the contingency table. Minitab prints the first 8 characters of the names in tables, but prints the full name on graphs. If you do not enter names here, the rows will be named Row1, Row2, etc.
Column names: Enter a column that contains names for the columns of the contingency table. The name column must be a text column whose length matches the number of columns in the contingency table. Minitab prints the first 8 characters of the names in tables, but prints the full name on graphs. If you do not enter names here, the columns will be named Column1, Column2, etc.
Number of components: Enter the number of components to calculate. The minimum number of components is one. The maximum number of components for a contingency table with r rows and c columns is the smaller of (r-1) or (c-1), which is equivalent to the dimension of the subspace onto which you project the profiles. The default number of components is 2.