Cramer's V is a chi-square based method of determining correlation in tables with more than 2x2 rows and columns, according to ChangingMinds.org. It is used to determine the strength of a correlation that has already been deemed significant by a chi-square test.
Cramer's V was named after mathematician Harold Cramer. The method returns a measure between zero and one, with variables closer to one indicating a stronger association, notes the Virginia Commonwealth University. It defines a perfect relationship as being predictive and a null relationship as being statistically independent, and the measure can only reach one when both variables have equal margins.