Survey Research

Survey Research

1. Uses questionnaires and interviews to ask people to provide information about themselves

   • Response set
     • Tendency to respond to all questions from a particular perspective
     • Social desirability response
       • Most common - answers in the most socially acceptable way

2. Constructing questions

   • Types of questions
     • Attitudes and beliefs
     • Demographic questions
     • Behaviours
   • Question wording
     • No unfamiliar terms
       • Vague terms
         • Embedding the sentence with misleading information
     • Unnecessary complexity
     • Double barrelled questions
       • Asks two things at once
     • Loaded questions
       • Written to lead people a certain way
     • Negative wording
     • Yea saying and nay saying
       • Word the questions so that consistent agreement is unlikely
       • Do not want people to get used to one answer

3. What kind of data are you seeking?

   • Closed ended questions
     • Limited number of responses
     • Easier to code
     • Rating scales
       • Characteristics
         • Ask people to provide "how much" judgments
         • Fully labelled scales are more reliable
         • 5 and 7 point scales are more common
       • Types of scales
         • High frequency scale
           • Alternatives assume a high frequency
         • Low frequency scale
           • Alternatives assume a low frequency
         • Graphic rating scale
           • Requires a mark along a continuous line that is anchored with descriptions at each end
         • Semantic differential scale
           • Measures the meaning people ascribe to concepts
             • Put a checkmark on different lines
         • Non verbal scale
   • Open ended questions
     • Respondents can answer any way they like
     • Harder to code

4. Administering surveys

   • Questionnaires
     • Personal administration
     • Mail surveys
       • Low response rates
     • Internet surveys
   • Interviews
     • Face to face interviews
       • Time consuming and expensive
     • Telephone interviews
     • Focus group interviews
       • Interview with 6-10 people for 2-3 hours.
   • Panel study
     • People are surveyed at multiple points in time

5. Sampling from a population

   • Confidence intervals
     • Range of plausible values for the population
   • Sample error
     • Error that comes from not measuring the entire population
   • Sampling frame
     • The actual population from which a random sample is drawn
   • Response rate
     • Percentage of people who actually completed the survey
   • Sampling techniques
     • Probability sampling
       • Simple random sampling
         • Every member has an equal chance of being selected
       • Stratified random sampling
         • Population is divided into subgroups, simple random sampling is used after
       • Cluster sampling
         • Identifying clusters of individuals and sampling from there
     • Non probability sampling
       • Convenience sampling
       • Purposive sampling
         • Obtain a sample of people that meet a predetermined criterion
       • Quota sampling
         • Choose a sample that reflects the numerical composition of subgroups