Types of Surveys
Survey data are collected by using self-administered questionnaires, face-to-face interviews, and/or telephone interviews. A questionnaire is a printed research instrument containing a series of items to which subjects respond. Items are often in the form of statements with which the respondent is asked to "agree" or "disagree" Questionnaires may be administered by interviewers in face-to-face encounters or by telephone, but the most commonly used technique is the self-ad1l/ministered questionnaire.
The questionnaires are typically mailed or delivered to the respondents' homes; however, they may also be administered to groups of respondents gathered at the same place at the same time. Sociologist Kevin E. Early (1992), for example, conducted a survey regarding the lower rates of suicide among African Americans than whites in the United. States. Early collected and analyzed survey data to test his hypothesis that "the black church's influence is an essential factor in ameliorating and buffering social forces that otherwise would lead to suicide." A self administered questionnaire was completed by congregation members in conjunction with services at six black churches considered representative of the thirtyseven black churches in Gainesville, Florida. Self-administered questionnaires have certain strengths. They are relatively simple and inexpensive to administer, they allow for rapid data collection and analysis, and they permit respondents to remain anonymous (an important consideration when the questions are of a personal nature).
A major disadvantage is the low response rate. Mailed surveys sometimes have a response as low as t0 percent-and a 50 percent response rate is considered by some to be minimally adequate (Babble, 2004). Thee response rate is usually somewhat higher if the survey is handed out to a group that is asked to till it out on the spot. Moreover, for surveys involving ethnically diverse or international respondents, the questionnaire must be available in languages other than English (Fink, 1995).
Survey data may also be collected by interviews. An interview is a data-collection encounter in which an interviewer asks the respondent questions and records the answers. Survey research often uses structured interviewer, in which the interviewer asks questions from a standardized questionnaire. Structured interviews tend to produce uniform or explicable data that can be elicited time after time by different interviews.
For example, in addition to surveying congregation members, Early (1992) conducted interviews with pastors of African American churches, using a series of open-ended questions. Next, he read four vignettes (stories about people) relating to suicide to the pastors and then asked questions designed to determine the pastors' opinions and attitudes concerning the behavior displayed in the vignettes. His goal was to learn the extent to which the African American church reinforces attitudes, values, beliefs, and norms that discourage suicide. Unlike the open-ended questions used in Early's more qualitative approach, dosed-ended questions may be used when researchers want to have alarge number of respondents and to generate standardized answers to questions. For example, in a study involving forty-nine hospital accident and emergency departments and psychiatric services in the United Kingdom, researchers developed dosed-ended questions to be asked of patients who had attempted suicide by poisoning. The purpose of the study was to determine if media representation of suicide and deliberate selfharm encouraged suicidal behavior in vulnerable individuals. Specifically, the researchers wanted to know if the rate of self-poisoning and the choice of overdose drugs were influenced by a television drama, Casualty, which portrayed an acetaminophen overdose in
one of its episodes. Questionnaires were completed by more than 1,000 self-poisoning patients during the three-week periods before and after the program was broadcast. Was there a direct link between viewing the episode and the person's decision to take an overdose, choice of drug, and speed with which he or she arrived at the hospital? According to the researchers, there was a 17 percent increase in the number of hospital patients who reported that they engaged in self-poisoning in the week after the broadcast. and a 9 percent Increase In the second week. Moreover, the rate of poisonings by acetaminophen Increased more than that of uny other drug. In fact, 20 percent of the patients interviewed indicated that the program had influenced their decision to overdose, and 17 percent said it had influenced their drug choice (Hawton, Simkin, Deeks, et al., 1999).
Interviews have specific advantages. TIley are usually more effective in dealing with complicated issues and provide an opportunity for face-to-face communication between the interviewer and the respondent. When open-ended questions are used, the researcher may gain new perspectives. TI'e pastors interviewed in Early's study distinguished between suicide (which is "unthinkable for black people" because it is a
"white thing, not a black thing") and alcohol abuse, drug addiction, and homicide (which are wrong and "sinful" but are "an understandable response to the socioeconomic and political conditions of blacks in the United States") (Early. 1992: 79). When closed ended questions are used, it is easier for interviewers to code responses and for researchers to compare individuals.' responses across categories of interest. For
example, researchers in the overdose study were able to compare such variables as sex. age, choice of overdose drug, history of taking overdoses, and whether the choice of substance was influenced by television programs. Based on their findings, the researchers argued that media portrayals of self-poisoning or self injury on popular television shows may contribute to self-harming behavior and choice of method used, and thus should be of concern to the general public and to media producers as well (Hawton, Simkin. Deeks, et al., 1999). As this and other research studies show, interviews provide a wide variety of useful information; however, a major disadvantage is the cost and time involved in conducting the interviews and analyzing the results. Also. one weakness of interviews is that in responding to the questions asked, people may bc influenced by the interviewer's race, age, sex, size, or other attributes. A quicker method of administering questionnaires is the telepholle survC)\ which is becoming an increasingly popular way to correct data. Telephone surveys save time and money as compared to self administered questionnaires or face-toface interviews. Some respondents may be more honest than when they are facing an interviewer. Telephone surveys also give greater control over data collection and provide greater personal safety for respondents and researchers than do personal encounters. In computer-assisted telephone interviewing (sometimes called CAT£),the interviewer uses a pewter to dial random telephone numbers, reads the questions shown on the video monitor to the respondent,and then types the responses into the computer terminal. The answers are immediately stored in the central computer, which automatically prepares them for data analysis. Although use of the CATI system overcomes the problem of unlisted telephone numbers by randomly dialing numbers, it is limited by people's widespread use of answering machines, voice mail, and caller ID to filter their incoming telephone calls.