By

A survey is a system for collecting information from or about people to describe, compare, or explain their knowledge, attitudes, and behavior. Today, billions of dollars are spent annually collecting survey data, and surveys are everywhere. The extensive use of surveys has lead to collection of huge amount of survey data. This massive amount of data is not of much use unless and until it is analyzed and some useful knowledge is extracted from it. For decades, statistics has been used as a tool for analyzing survey data. Hence, survey data are usually collected and recorded in a format suitable for statistical analysis. Recently, data mining has emerged as an active field for analyzing data. This has led the survey data organizers to record data in a format helpful for data miners. Great care is taken while collecting and recording data, but still there remain a few problems for data miners. In this project we present a few problems faced while mining survey data for association rules and present a solution to solve that problem. The purpose of a survey can range from describing some phenomenon, to establish-ing some relationships between variables. In this project, we concentrate on bivariate associations, that is, to see whether two variables are associated. We show how crosstabulations, a bivariate analysis technique can be used to determine bivariate associations. We show that the same information can be obtained from association i mining. We then show how association mining becomes more advantageous than crosstabulations as the number of variables for bivariate analysis increase. ii Acknowledgments I would like to express my sincere thanks to my supervisor, Dr. Robert Hilderman for his advice, assistance and financial support. Without his encouragement this work would not have been accomplished. I would also like to thank the Faculty of Graduate Studies and Research, and Department of Computer Science for financial support, my friends and relatives for their moral support, and my parents for making me capable of doing this type of work. Finally, I would like to thank my brother, and my sister. ii