Survey ResearchChapter 9SURVEY RESEARCHSURVEY RESEARCHSurvey Research is a form of observation.Why surveys sometimes fail -- several common mistakes:  Pick the wrong sampling strategy  Low response rate  Bad questionnaire: too long, asked the wrong questions, etc.  Lots of data but no easy way to analyzeSome basic terms: primary data: Original data that you collect (e.g., through a survey, an interview, etc.)Secondary data: Data that has already been collected by someone elseCensus: (1) A count of the population and a property evaluation in early Rome; (2) A usually complete enumeration of a populationSample: (1) A representative part of a single item from a larger whole or group(2) A finite part of a statistical population whose properties are studied to gain information about the whole Heisenberg Uncertainty Principle: This implies that one cannot observe with disturbing in some way the observed. The Culture of survey research:  Conducting a survey is a form of communicative inquiry of asking questions.  Both communication and inquisitiveness are shaped by culture.  Survey research is both a science and a social activity (a dialogue), and cannot be easily exported from one place to another.Ways to Administer Surveys: This will shape your questionnaire, as well as your sample. How the Questionnaire gets to the Respondent: a. in-person survey (interview)b. mail c. telephone d. internetHow they answer the questionnaire: 1. Self-administered 2. Administered by staffSome Advantages and Difficulties with Each Approach (ordered from low to high-tech) In-person Mail Survey Telephone InternetAdvantages - higher response rates- the most personable of the approaches - lots of control over the process - cost savings: no interviewers - time savings - questionnaire can be done at the respondent's convenience - greater assurance of anonymity - standardized wording (no interview bias) - respondents can check records - accessibility: can reach anyone with a mail address - fast - lower cost than face-to-face- more anonymous than face to face - can use wide geographic area - random-digit dialing facilitates an unbiased sample - Computer assisted technology both accelerates the process and allows for complex filtering - answers already in electronic format - can allow for complex filtering questions - very quick response time - can reach anyone with access to a computer world-wide - perhaps the quickest to fill out (especially if multiple choice) Difficulties - can be costly, especially if covering awide geographic area - time-consuming - can be expensive - requires good training of interviewers - lack of flexibility: no probing, no corrections, no soothing reluctant respondents - low response rate - no control over the environment - no control over question order - no control over who fills outthe questionnaire - no control over date of response - cannot use complex questionnaire format (e.g., complicated filtering) - lots of resistance to telephone interviews - can be confused with marketing - needs to be fairly short - no good for long questions, long lists of answer categories - can take many tries to reachsomeone at home - no visual materials possible - bias introduced -- only some people have access to computers. How do you determine the population? - low response rate - hard to get long answers to open-ended questions (as compared to in-person or telephone) - a mistrust of the way the data might be used (e.g., for