A survey is a highly structured series of written questions that is administered to a large number of persons. Surveys require data analysis using some statistical methods. Surveys result in numbers, percentages, averages and other mathematical devices that show how a population looks, behaves, thinks, or reacts.

Surveys and questionnaires are some of the most often used methods of doing research. Certain curricula come with standard questionnaires to be used. Surveys are often borrowed from another agency and adapted. Surveys can be filled out by the client alone or filled out by an interviewer asking the questions of the client (good for low-literacy or disabled clients).

Surveys begin with a written questionnaire and are conducted in a systematic manner. For example, the same survey is conducted with many different persons. Most questions in a survey are close-ended, that is, they require a simple answer such as yes/no, often/never, or age/sex/ethnicity. Often, at the end of a survey, time is left for open-ended questions that may elicit longer responses.

While surveys are the most common tool for evaluation, they are also expensive, time-intensive, and practically useless if not done well. The number one rule of conducting surveys is: don’t collect what you can’t use. This will not only ease the burden on participants, it will also help with staff efficiency. Agencies should have a clear idea of what exactly they want to measure and whether or not they have realistic expectations from their clients and program.

If you do conduct surveys, have a plan and resources to analyze and disseminate the results. Agencies should have data analysis expertise or have the funds to hire an outside analyst. Agencies also sometimes secure in-kind donations of analysis from local universities or research firms.

NOTE: If you do a good job developing and using a survey, you ought to find out something you didn’t know.

How are they used?

Surveys can help:

•    Describe populations (demographics such as age, race, education, etc).

•    Show prevalence of behaviors. How many clients are having sex under the influence of alcohol or drugs? How many regularly attend public sex/cruising spots?

•   Determine level of knowledge of clients/populations. Do they understand about the HIV testing window period? Do women know that the pill does not protect against HIV?

Surveys should never be developed without the input of the target audience. Focus groups or in-depth interviews can help come up with questions and issues that are important to your clients. In addition, pilot-testing the survey will tell you if the questions are understandable and if any questions should be added or deleted.Often agencies get “stuck” in the traditional questions of behavior change measures: condom use, drug or alcohol use, number of partners, sexual activity while using drugs or alcohol, needle sharing, etc. But there are many questions that can be asked that go beyond the “standard” questions and explore important underlying issues in staying safe.

Good Questions, Better Answers --  � 1998 California Department of Health Services and
Northern California Grantmakers AIDS Task Force  -- http://www.goodquestions.com