Statistics Glossary → Supplementary (passive) points
Supplementary (passive) pointsconcept
Row or column categories in correspondence analysis that do not take part in determining the solution (the axes) but whose profiles are placed on the map afterwards for comparison with the other categories.
Correspondence analysis derives axes from the row and column profiles of a contingency table and places the categories on a map made of these axes. Active categories contribute to computing the axes. Supplementary (passive) categories are left out of the computation; once the axes have been found, their profiles are projected into the same space. Their positions on the map can then be interpreted, but they do not change the solution.
Supplementary points are useful for showing an external benchmark (a national average, data from another year) or categories that rest on few observations and could otherwise sway the solution disproportionately. Because passive categories contribute nothing to the axes, their positions should not be used to define what the axes mean; and supplementary points poorly represented by the map can look misleadingly close to the centre.
An everyday example
A researcher maps newspaper reading habits in seven provinces with correspondence analysis. She would like to see the national distribution on the map too, but worries that including it in the solution would let the 'average' rather than the provinces pull the axes. Adding the national figures as a supplementary row means the map is built from the seven provinces, with the national average appearing on it as a reference point.
In MerQur
In Correspondence Analysis, categories named in the 'Supplementary row categories' and 'Supplementary column categories' fields do not determine the solution; their profiles are projected onto the map afterwards.
Analyses where this concept appears
Correspondence Analysis (CA / MCA)