There's no doubt that companies can benefit from workplace surveys and questionnaires. Good surveys accurately home in on the problems the company wants information about. They are designed so that as many people as possible actually respond. And good survey design ensures that the spectrum of responses is unbiased. In this article, the author, a former research scientist at the University of Michigan and currently the president of a survey design firm, explores some glaring failures of survey design and provides 16 guidelines to improve workplace assessment tools. Applied judiciously, these rules will not only make a tangible difference in the quality and usefulness of the data obtained but will also produce an increased response rate. The guidelines--and the problems they address--fall into five areas: content, format, language, measurement, and administration. Here are a few examples: Survey questions should require people to assess observable behavior rather than make inferences; each section should contain a similar number of items and each item should have a similar number of words; words with strong associations to gender, race, or ethnicity should be avoided; the wording in one-third of the questions should be changed so that the desirable answer is a negative one; and response scales should provide a "don't know" or "not applicable" option. Following the guidelines in this article will help you get unbiased, representative, and useful information from your workplace survey.