Publications Search - Abstract View
||The potential effect of statistical dependence in the analysis of data in jury discrimination cases: Moultire v. Martin reconsidered
||Gastwirth JL, Miao W
||This article describes a statistical analysis that incorporates the dependence between consecutive grand juries in jurisdictions that allow jurors to serve a second year. This holdover juror system is used in South Carolina and is permitted in California. The appropriate modifications to the usual standard deviation analysis are presented and applied to data from Moultrie v. Martin, 690 F.2d 1078 (4th Cir. 1982). The analysis shows that ignoring the dependence made the statistical evidence of discrimination appear stronger than it was. Furthermore, the standard method, which ignores the dependence, also exaggerates the power of the hypothesis test to determine whether there was a potentially discriminatory selection process.