Transcript Administering Surveys (con ` t)

Chapter 7:
Asking People About Themselves
Why Conduct Surveys?
 Survey methodology
• Nonexperimental
- Questionnaires
- Interviews
Why Conduct Surveys? (con’t)
 Why are surveys used?
• Allows people to tell researchers about themselves
• A method for studying relationships among variables
and ways that attitudes and behaviors change over
time
• Provides useful information for making public policy
decisions
• An important complement to experimental findings
Why Conduct Surveys? (con’t)

Some important issues
• Assume people are willing and able to provide
truthful and accurate answers
- Response sets
1. Social desirability response set
•
When are respondents most likely to lie?
- Lack of trust in the researcher(s)
Constructing Questions to Ask
 Defining the research objectives
 Three general types of survey questions
• Attitudes and beliefs
• Facts and demographics
• Behaviors
Constructing Questions to Ask (con’t)
 Question wording
•
Potential problems that stem from a difficulty
understanding the question
1. Unfamiliar technical terms
2. Vague or imprecise terms
3. Ungrammatical sentence structure
4. Phrasing that overloads working memory
5. Embedding the question with misleading
information
Constructing Questions to Ask (con’t)
 Important considerations when writing questions
• Simplicity
• Double-barreled questions
• Loaded questions
• Negative wording
• Yea-saying and nay-saving
Responses to Questions
 Closed-ended questions
• Limited number of response alternatives are given
• More structured approach
• Easier to code
• Response alternatives are the same for everyone
• Useful when the dimensions of the variable are well
defined
Responses to Questions (con’t)
 Open-ended questions
• Respondents are free to answer in any way they like
• Requires time to code responses; costly
• Some responses cannot be categorized
• Useful to find out what people are thinking and how
people naturally view the world
Responses to Questions (con’t)
 Closed-ended and open-ended approaches can
sometimes lead to different conclusions (Schwartz,
1999)
Responses to Questions (con’t)
 Points to consider when asking closed-ended questions
 Number of response alternatives
 Rating scales
 Labeling response alternatives
Responses to Questions (con’t)
 Formats of rating scales
• Simplest and most direct scale is fie to seven
response alternatives with the end-points on the scale
labeled to define the extremes.
Strongly agree _____ _____ _____ _____ _____ Strongly disagree
Responses to Questions (con’t)
 Graphic rating scale
• Requires a mark along a continuous 100- millimeter
line that is anchored with descriptions at each end
Not very enjoyable
Very enjoyable
A ruler is used to measure the score on
a scale that ranges from 0 to 100.
Responses to Questions (con’t)
 Semantic differential scale
• Measures the meaning of concepts
• Respondents rate any concept on a series of bipolar
adjectives using 7-point scales
Good _____: _____: _____: _____: _____: _____: _____ Bad
Strong _____: _____: _____: _____: _____: _____: _____ Weak
Responses to Questions (con’t)
 Nonverbal scale for children
• Children may not understand other types of scales
Responses to Questions (con’t)
 Labeling response alternatives
 Sometimes needed to clearly define the meaning of
each alternative
__________ __________ __________ __________ __________
Strongly
Agree
Agree
Undecided
Disagree
Strongly
Disagree
Finalizing the Questionnaire
 Formatting the questionnaire
• Attractive and professional looking
• Free of spelling errors and neatly typed
• Questions and response alternatives should easy to
identify
• Don’t change from 5- to 4- to 7-point scales
• Ask the most interesting and important questions first
• Ask demographic questions last
Finalizing the Questionnaire (con’t)
 Refine the questions
• Before administering the survey, pilot test the
questions to a small group of people
- Use the “think aloud” procedure
Administering Surveys
 Two methods to administer surveys
• Written questionnaire
• Interview
Administering Surveys (con’t)
 Positive features of questionnaires
• Less costly than interviews
• Allows respondents to be completely anonymous
• Can be administered in person to groups or
individuals
• Can be administered through the mail, on the Internet,
or with other technologies
Administering Surveys (con’t)
 Potential negative features of questionnaires
• Understanding of the questions
• Motivation
• Response rates (mail surveys)
• People may misrepresent themselves (Internet
•
surveys)
• Unanswered questions
Administering Surveys (con’t)
 Positive features of interviews
• Involves an interaction between people
- Response rates higher than questionnaires
• People answer most, if not all, the questions
• Interviewer can clarify questions
• Interviewer can ask follow-up questions
Administering Surveys (con’t)
 Potential negative feature of interviews
• Interviewer bias
- Show approval or disapproval of certain answers
- If there are several interviewers, each could
possess different characteristics (e.g.,
attractiveness, age, ethnicity, etc.)
- Expectations could lead interviewers to “see what
they are looking for” in the answers
Administering Surveys (con’t)

Three methods of conducting interviews
1. Face-to-face interviews
2. Telephone interviews
3. Focus group interviews
Survey Designs to Study Changes Over Time
 Test hypotheses and compare survey responses
across time
 Panel study
• “Two-wave” panel study
• “Three-wave” panel study
Sampling From Populations
 Sampling
• Population is composed of all individuals of interest
to the researcher
• A sample is a subgroup of the population
Sampling From Populations (con’t)
 With proper sampling, we can use sample information to
precisely estimate characteristics of the population (aka
statistical inference)
• Confidence intervals
- Percentage of confidence that the true population
value lies within a interval
- Sampling error (margin of error)
Sampling From Populations (con’t)
 Sample size
• Reduces or increases the size of the confidence
interval
• Computation of sample size needed
Sampling Techniques
 Two basic techniques for sampling
• Probability sampling
- Each member of the population has a specifiable
probability of being chosen
• Nonprobability sampling
- Unknown probability of any member being chosen
Sampling Techniques (con’t)
 Probability sampling
• Simple random sampling – equal probability
• Stratified random sampling – divided into subgroups
(strata) and random samples are taken from each
strata
• Cluster sampling – identify clusters and sample from
these clusters
Sampling Techniques (con’t)
 Nonprobability sampling
• Haphazard sampling – convenience sampling
• Quota sampling – sample reflects the numerical
composition of various subgroups in the population
Evaluating Samples
 Representativeness of the population
• Sampling frame
• Response rate
• Reasons for using convenience samples
The End