| 
  
   Surveys
    provide information in a systematic and quantitative manner, which facilitates statistical
    analysis. Surveys may be targeted toward specific population groups defined in various
    ways. For example, surveys may target particular neighborhoods or communities to
    facilitate outreach efforts and plans for new programs and services by identifying persons
    with unmet needs. Surveys provide information that is current, and, in certain situations,
    data may be collected and analyzed during a short time period (two to four weeks). In
    addition, one major advantage of surveys is that they can collect otherwise unavailable
    data directly from potential users of the types of services and programs with which the
    agency is concerned. For example, the only way to obtain data about a persons
    attitudes, opinions, future plans, expectations, and behaviors that are not directly
    observable is to ask the person about these issues.      The key to successfully designing and conducting surveys is to
    begin with small-scale, well-focused surveys.     Obtain professional consultation whenever necessary. This is
    especially important when attempting a new survey method or when performing an aspect of
    survey work with which you have not had previous experience.     Smaller is better. If your agency does not have access to a
    statistician, keep your questionnaire simple and short.
 Questionnaire hintsSurvey questionnaires are standardized in the sense that they present a uniform stimulus
    to each respondent. The wording and ordering of all questions are determined in advance
    and all questions are asked of all respondents in exactly the same way. Except for
    questions that are contingent on the respondents answer to a previous question, the
    questions are asked of each respondent in exactly the same order. These characteristics of
    survey questionnaires provide for the collection of data that are comparable across all
    subjects.
 Question wordingThe goal of writing questions is to use words that everyone understands and that impart
    only the meaning that the researcher intends. One guide is to consider the educational
    level of the least educated respondent that will be part of the survey and write questions
    using language that will be understood by that person. Questions should be polite and
    written in a conversational tone.
 Question orderThe placement of questions is an important aspect of questionnaire design. Consideration
    should be given to making the questionnaire interesting and convenient to use with a
    logical flow of topics. The opening questions should capture the respondents
    interest and motivate him or her to continue to answer the remaining questions. Its
    a good idea to start with questions that are important and have relevance to the topics of
    the survey, and that are easy for the respondent to answer. Sensitive and threatening
    questions are generally asked later on in the questionnaire. The last questions asked are
    demographic questions.
 Closed vs. open questionsAlthough survey questions are commonly referred to as being
    closed or open, it is actually the response to a question that is
    treated in either a closed or open manner. Open questions provide respondents more freedom
    of expression and allow the researcher to obtain unanticipated responses. Closed questions
    provide a convenience to both the respondent and the researcher. They are generally quick
    and easy to answer, there is less margin for error in recording and processing responses,
    and they are less costly to process than open questions.
 Closed question: In the past 6 months, how often did you use drugs like
    marijuana, cocaine, crack or heroin before having vaginal or anal intercourse with (SEX
    PARTNER #1)? Would you say. . .(READ)      Never,     1     Sometimes,    2
 About half the time,     3
 Almost all the time, or    4
 Every time?    5
 Open question: What did you do when you realized that (SEX PARTNER #1)
    did not want to use a condom? (INTERVIEWER, PROBE OFTEN: What else
    did you do?)_______________________________________________
 _______________________________________________
 _______________________________________________
  
   |