Basic and Applied Statistical Inference and Measurement

George Washington University, Washington DC

Investigators

Abstract

The project will develop statistical procedures for a variety of problems that arise in the use of statistics in law, public policy, and in basic social science research. The project will examine problems in legal, social science, and public policy contexts and develop appropriate probability models and statistical methodology for their solution. The objective is to ensure that sound methodology is used in these important areas. This project, for example, will study some techniques that have been used in actual legal cases to assess their statistical properties and create alternative methods that are more efficient. Because protecting the privacy of study participants and guaranteeing the confidentiality of the information they provide is essential for retaining public cooperation, the project also will investigate methods using grouped rather than individual data. The legal system decides many important issues in our society. It therefore is important that the statistical evidence submitted in actual cases involving, for example, equal employment or environmental safety, is reliable and that sound conclusions are drawn from it. One goal of the project is to develop improved methods for interpreting statistical studies and analyzing statistical data arising in these areas. The intent of the project is to study problems relating to issues that are basic to a scientific discipline. Improving methods used to assess whether a test accurately predicts job performance or who will do well in school is important in educational research and psychology, as well as in discrimination cases. Similarly, methods for obtaining proper conclusions from data that is in grouped form to protect the anonymity of individual respondents can be used to analyze survey data used in economic, sociology, and public health.