Processing Data : The Survey Example

"Clear and concise in its recommendations, this book should find its place on lists of required reading for all practical survey design and analysis courses."

--The Statistician

Although data processing is frequently thought to occur only after the data are collected, it actually occurs throughout the phase of study design, implementation, and analysis. Decisions made during the design of data collection instrument influence the kind of data and the format of the data available for analysis. Devoted entirely to data processing, this much needed volume addresses such key issues as: Should you create your own instrument for a questionnaire? How do you test a questionnaire? What are the characteristics of good data processing? How do you deal with missing data? How do you scale an evaluation and create subfiles for analysis? Processing Data begins with a discussion on the selection of the data selection techniques and how this impacts data processing and the data for later analysis. Each major section concludes with examples, and when appropriate, directs the reader to commonly available computer software that can aid in data processing.

Linda Bourque, PhD, is a Professor in the Department of Community Health Sciences and an associate director of both the Center for Public Health and Disasters and the Southern California Injury Prevention Research Center in the UCLA School of Public Health. Trained as a sociologist, she teaches courses on research design with an emphasis on the design, data processing, and data analysis of questionnaires and community-based surveys. Her research during the last twenty years has focused on community response to disasters. A public website contains all of the raw data, codebooks, questionnaires, publications and related material from surveys conducted on California earthquakes since 1971 by Leo Reeder, Ralph Turner, Dennis Mileti and Linda Bourque. Current research includes the National Survey of Disaster Experiences and Preparedness (NSDEP), and the California Survey of Earthquake Preparedness.

Funded by the Department of Homeland Security and the National Science Foundation, NSDEP examines the factors that predict disaster preparedness and risk avoidant behavior, with an emphasis on terrorism. A stratified sample of 3,300 households was selected using random digit dialing: 1,000 households were selected for interview in areas considered at high risk of terrorism (Washington, D.C., New York City, Los Angeles County), and 2,300 households were selected for interview throughout the rest of the continental United States. Respondents were asked whether they had invested in six preparedness behaviors and seven risk avoidant behaviors either because of terrorism, natural disasters, other reasons, or any combination of the three. NSDEP reports and other documentation are available here.