Analyses

Correspondence analysis

Map the rows and columns of a contingency table into a shared space.

Edit on GitHub

When to use it

Use correspondence analysis when a crosstab is too large to read cell by cell. It places row categories and column categories on the same axes so categories that occur together sit near each other.

Assumptions

Both columns are categorical. The table cannot be empty. Two components are requested by default. A 2×2 table has only one dimension of association, so the second singular value can be zero.

Running it in Tensr

Survey techniques → Correspondence Analysis. In chat: “Correspondence analysis of brand by attribute.”
Row column and column column are the two categorical variables.

Options

Prop

Type

Reading the output

Teaching method by satisfaction rating. The association is mild, so the map should not collapse onto a single perfect diagonal.

Analysis complete.

Reporting (APA 7)

Report the estimate in the table. This procedure is not summarised by one p-value.

Coming from SPSS

SPSS has correspondence analysis under Dimension Reduction. Tensr lists it under Survey techniques.