nuisance parameter
in statistical hypothesis testing, a population parameter of secondary interest that must be accounted for in order to obtain an estimated value for a parameter of primary interest. For example, assume a researcher wishes to determine whether male and female schoolchildren differ in their mathematics ability. He or she might administer a math test to a sample of children at a particular school and use that data to estimate male and female means for all children. If, however, the researcher needed to calculate the variance of the scores before attempting to determine the means, the former would be a nuisance parameter.