Peter V. Marsden is Edith and Benjamin Geisinger professor of sociology and dean of social science at Harvard University.
In addition to social science methodology, his research interests lie in the areas of social organization, especially social networks and organizations, and the sociology of medicine.
James D. Wright is provost distinguished research professor in the department of sociology at the University of Central Florida.
He has published 21 books on various topics including guns, poverty, drugs, natural disasters, American politics, and research methods. His current research interests include the "divorce reform" movement, health aspects of homelessness, urban poverty and inequality, and violence. He has been editor of Elsevier's journal Social Science Research since 1978.
The Handbook of Survey Research is a comprehensive reference work on the most widely used method in the social sciences, exploring developments over the past half century and examining new usages. It updates the first edition, published in 1983.
Chapters include: sampling; measurement; questionnaire construction and question writing; survey implementation and management; non-response and missing data; special types of surveys; and integrating surveys with other data collection methods. New issues covered include ethical and human subjects issues, measurement models, the role of cognitive psychology, Internet surveys, archiving and dissemination, and cross-national/cross-cultural surveys.
What is the rationale behind the handbook, and what differentiates it from other books on survey research methods?
The handbook fills a need for a comprehensive overview of survey research methods that is nearly encyclopedic in coverage, with attention to fundamentals as well as some advanced topics. It is suitable both as a text for advanced undergraduates and graduate students, and as a useful, convenient reference tool for survey practitioners.
The previous version was published in 1983. Why wait for over 25 years for a new edition?
Elsevier, the publisher of the first edition, made periodic enquiries over the years about an updated version. The timing for such an ambitious project was never quite right until we took on the updating task in the early 2000s. By then, myriad changes in the survey industry made it essential to rewrite the chapters on many topics, and add new ones to cover what had emerged since the 1980s.
What audience(s) are you aiming at?
Our main targets are graduate-level students in courses on survey design, along with professionals in survey research and social science.
Would you agree that handbooks in general tend to concentrate on reviewing existing research and projecting future trends, rather than publishing original empirical research? (See "How to... write a chapter for an edited volume".)
Yes. We hope that by summarizing and organizing past advances, the handbook provides a guide to current best practice, and serves as a foundation for new research that will advance and refine best practice.
With 28 chapters and 44 authors, putting together the handbook must have been a major undertaking. How did you manage it?
From initial agreement to books in hand was a five-year process, during which time we both carried on other projects and discharged other responsibilities. With tools including e-mail and word processing systems that support collaboration (via inserting comments, tracking changes), the entire process became vastly more efficient than in the 1980s when the first edition was produced. Managing the enterprise was certainly facilitated by an excellent, committed group of chapter authors and a pair of editors who saw eye-to-eye on almost every issue that arose.
How did you go about organizing the handbook, and selecting topics and authors? (For example, why did you not include data analysis?)
The outline of chapters was drafted, reviewed, and redrafted numerous times during the process. Likewise the list of prospective authors. Some topics were obviously essential to a new edition (e.g. a chapter on Internet surveys, a chapter on mixed-mode surveys); other topics were added along the way when emerging developments in the survey industry seemed to demand coverage. Four chapters from the original handbook dealing with data analysis issues were omitted because many other texts now provide ample coverage of generic data analysis techniques that were not well covered when the first edition was published. We did include chapters on survey-specific analysis topics including complex sample designs, missing data methods, and analysing repeated surveys.
What were the qualities you looked for in the chapters?
We sought thorough, detailed, balanced, comprehensive and accessible coverage of key issues in survey research by the leading figures in the industry.
I believe you had eight reviews of the original outline for the book. How did you go about reviewing individual chapters?
In most cases, we each independently reviewed each chapter as it was submitted. One or the other of us then summarized our suggestions and communicated them to authors. In a few cases, we sought outside reviews of specific chapters. Every chapter was extensively revised based on the internal and external reviews.
In what ways would you consider the survey to be a better, and more accurate instrument than when the first edition was published in 1983?
Countless advances in survey methodology – sampling innovations, new modes of data collection, an outpouring of methodological research on question and questionnaire design, heightened sensitivity to human subjects issues and confidentiality protection, modern methods for managing missing data, mixed-mode designs, new models for measurement – are chronicled in the handbook. These developments make surveys better than they were 25 or 30 years ago. Likewise a handbook that incorporates them is a better handbook. Survey data are now being integrated with numerous other sources of information – administrative data, geographic data – and many researchers now experiment within surveys. New problems are emerging – in particular, falling response rates are a significant threat – but improved understanding of them leads to new measures to counteract such problems.
How long do you think it will be before a third edition of the handbook is needed? Can you speculate as to what topics it will contain?
We think this edition will serve its target readership well for some time. Of course, editors can always identify additional issues that bear attention; we'd like to add material on methods for sampling rare or difficult to reach populations, surveys of businesses or other organizations, and coding/data preparation practices, among other subjects.
New developments at the cutting edge of the survey industry, e.g. advances in surveying via cell phones, new strategies to improve sampling for Internet surveys – are reported each month. A third edition would surely expand attention to these and other issues. Whether a third edition will be developed sometime depends heavily on the market response to this edition, publisher interest and enthusiasm, continuing commitments on the part of leading survey researchers to contribute to the enterprise, and the availability of time and energy from the current – or other – editors.
The survey is the quintessential quantitative method. The journal Social Science Research publishes quantitative research. Do you consider quantitative research to be superior to qualitative for the social sciences?
Both of us have devoted our careers to careful quantitative understandings of social and behavioural phenomena. Neither of us would presume to assert that quantitative research is inherently "superior", though. We view the best social science research as question rather than method driven. Survey research rates high on representativeness and generalizability, but for some research questions – especially those that explore a phenomenon in depth and prize naturalness – qualitative research designs may be optimal. Armed and Considered Dangerous: A Survey of Felons and their Firearms (Wright and Rossi, 1986) answers certain kinds of questions about violence in the inner city. Elijah Anderson's Code of the Streets answers other kinds of questions about the same phenomenon drawing on qualitative material.
The qualitative-quantitative contrast can be overdrawn, too: mixed-method designs are becoming common in social science now. A good deal of what survey researchers have learned about cognitive processes in surveys and interviewer behaviour draws on qualitative data.
Anderson, E. (2000), Code of the Street: Decency, Violence, and the Moral Life of the Inner City, W.W. Norton & Company, NY.
Wright, J. and Rossi, P.H. (1986), Armed and Considered Dangerous: A Survey of Felons and their Firearms, Aldine de Gruyter, NY.
Margaret Adolphus interviewed Peter V. Marsden and James D. Wright in May 2011.
Visit the information page for: Handbook of Survey Research