Surveys: One common quantitative public relations research method is the survey or questionnaire. The survey sample frame defines the particular population under study (for example, all non-management employees at XYZ Company who have worked there for less than five years). In order for a survey to have external validity (defined by Broom & Dozier, 1990, as “the degree to which a measure is valid for other settings and populations,” p. 397), researchers must draw a probability sample. Probability samples are drawn in such a way that the researcher knows the probability of selecting any particular sampling element (Broom & Dozier). In random sampling, this means that every person who falls within the population defined by the sample frame has an equal possibility of being surveyed. The size of the sample depends on budget, and the margin of error and degree of uncertainty with which the researcher can tolerate. Questions should also be tested to ensure items not are not ambiguous, irrelevant, confusing, or biased (Gronstedt, 1997). Care must also be taken with the wording of questions, since the questionnaire must attempt to operationalize difficult concepts, such as employee satisfaction, or community satisfaction with the organization. Operationalizing “is the process of connecting an abstract concept to observable phenomena in the real world” (Broom & Dozier, 1990, p. 163). The questions must be appropriately designed to measure those concepts. For that reason, Broom & Dozier (1990) recommend using existing item sets from previous research; see the Handbook of Research Design and Social Measurement (Miller & Salkind, 2002) for examples.
Surveys can be conducted face to face, by telephone, by mail, and via the Internet. Face-to-face surveys have the highest response rates. They also have the advantage of enabling the interviewer to show respondents materials to respond. Face-to-face surveys are also the most expensive and time-consuming to administer. Also, the presence of the interviewer may cause respondents to skew their responses to appear more favorable to interviewer. Telephone surveys are less expensive than face-to-face surveys, and can be completed more quickly. Virtually every home in the United States has a telephone, and the development of random-digit dialing programs (which eliminate the problem of relying on telephone directories, since 30% of the phones nationally are unlisted) theoretically provides access to most Americans. However, telephone surveys have higher refusal rates than face-to-face surveys. Many people use caller ID and answering machines to screen calls, making it harder to reach participants (Dillman, 2000). Although survey calls are exempt from National Do Not Call Registry provisions (as are those by charities, politicians and those with whom people already have established business relationships) (www.donotcall.gov, 2004), the public seems more resistant to responding to calls from strangers. In telephone research, the choices must be easily understood over the telephone. Telephone interviews are usually shorter than face-to-face interviews, generally lasting 10 minutes or less, therefore limiting your ability to analyze difficult concepts or issues (Broom & Dozier, 1990; Stacks 2000).
Mail surveys cost less than telephone surveys and are easier to administer. Mailing lists may be available to target specific publics. However, because mail surveys are easy to ignore, they also have a lower response rate. To increase response rates, Dillman (2000) recommends these five
steps:
- Develop a respondent-friendly questionnaire that is easy to read and understand, and has an attractive layout.
- Plan to contact the recipient up to five times, with a prenotice letter, the questionnaire mailing itself, a follow-up thank-you postcard, a replacement questionnaire, and a final contact by telephone or express mail.
- Use first-class stamps on return envelopes instead of business-reply envelopes.
- Personalize correspondence, using real stationary, the respondent’s name on the letter, and an actual signature.
- Include a token financial incentive ($1 to $5) with the survey. (However, the promise of incentives after a survey is completed, or the chance of winning a prize will not increase response rates.)
Internet surveys are gaining in popularity because they can reach large numbers of respondents without increasing costs. Once the data collection system is developed, costs are similar whether the sample is 100 or 10,000. Many Web-survey systems also provide analyses. Internet surveys “provide survey capabilities far beyond those available for any other type of self-administered questionnaire” (Dillman, 2000, p. 354). However, there are several drawbacks to Internet surveys. Web-based surveys may not be compatible with all browsers, which means some people who want to respond to the survey may not be able to do so. Also, although use of the Internet is growing steadily, many people do not have access to the Internet. Many, also, are not familiar enough with the Internet to navigate through a Web-based survey. Web-based surveys often do not provide researchers with control over their sample. Researchers may know about who is responding to the survey, but not about those who are not. Web surveys, however, may be good for populations with high rates of computer use, such as employees in a high-tech firm (Dillman, 2000).
0 Comments