Chance correction and map evaluation

The author investigates the possibility of predicting the quality (i.e., accuracy) of taxonomic maps (defined as maps displaying the spatial distribution of a classification scheme) prior to their compilation. The procedure involves the calculation of a “confusion matrix,” based on the concept of a 1:1 equality between map unit (class) and field unit (class), which can be used as benchmark against which departures from that ideal condition can be assessed probabilistically.

Göksel Türk

Papers

Predicting the quality of taxonomic maps