Rating: 
Summary: Misnamed, Lacking Scientific Methodology
Review: Well, this may be "the book" when it comes to surveys, but it still lacks a lot. First, I found the chapters on writing questions and organizing a questionnaire to be very useful. However, it looks like the author only added a couple chapters about Internet surveys, instead of reviewing his decades-old book in terms of the impact of the Internet. For example, I don't find it particularly useful to discuss how to fold a survey, or how to distribute it by mail, when it's being administered on-line!Furthermore, this book is lacking any real scientific methodology. I suspect this is a result of the nature of the field, but survey designers should at least try to employ some good experimental design approaches. For example, this book does not help me at all to ensure that the survey actually gives me information that I need. While he does give information on writing interpretable questions, he has no recommendations on how to determine the goals of the survey, how to design questions that will address those goals, how to arrange questions in the survey to ensure good data that addresses those goals, etc. And what about statistical accuracy, and how certain types of questions are easier to measure? Any suggestions on how to evaluate free-response questions? Why isn't there an entire chapter on "How to avoid bias and inaccuracy in responses", instead of having suggestions scattered around the text? This book has a lot of useful sociological hints on how to increase the response rate from a population. However, this isn't the most important aspect of surveys---it sorely lacks the scientific basis for the design and evaluation of the "experiment" that is the survey. Without this, it doesn't matter how many people respond, because the data will be useless.
| 
