Transcript Chapter Seven - Texas Christian University

Chapter Seven
Survey Research
Why Conduct Survey Research
•
•
•
•
Employs questionnaires and interviews asking
people to provide information about their attitudes,
beliefs, demographics (age, gender, income, marital
status), and past or intended future behavior.
Method used to study relationships among variables
and ways that attitudes and behavior change over
time.
Survey research can also help complement
experimental research findings.
Survey research is conducted on a given sample or
population.
Sampling from a Population
Population: larger group about whom we wish to
obtain information
• E.g., high school teachers in California, readers of the
National Review, students at UT Austin, truck owners.
• In most cases studying the entire population would be
impossible.
• Select a sample from the population which we can then
test.
• With proper sampling, whatever results you find with your
sample can be applied to the population as a whole.
Confidence Intervals
• Confidence interval  the interval within
which your true/population score lies.
• This interval gives you information about
the likely amount of error in your findings.
• When using a sample population, the
obtained score is not the true score of the
population.
• sample value  your best estimate of
the population value
Sample Size
• Larger sample sizes reduce the size of the
confidence interval thereby reducing the
amount of sampling error .
• Determined by using a mathematical formula
that takes into account the size of the
confidence interval and population of interest.
• Homogeneous vs. Heterogeneous Populations.
Sampling Techniques
• Probability sampling  each member of
the population has a specifiable probability
of being chosen.
• Nonprobability sampling  we don’t
know the probability of any particular
member of the population being chosen
Probability Sampling
• Simple random sampling  every member of
the population has an equal chance of being
selected.
• Stratified random sampling  used when it is
important to ensure that subgroups within a
population are adequately represented in the
sample .
• Cluster sampling  rather than randomly
sampling from a list of individuals, the
researcher can identify “clusters” of individuals
and then sample from these clusters
Non-Probability Sampling
• Haphazard sampling  you select a
sample any way that is convenient.
• Advantage: it is easy to do
• Disadvantage: it might not be a
representative sample of the population,
and the survey results might be biased
Sources of Bias in a Sample
•
Sampling frame
•
Response rate
Constructing the Questionaire
•
•
Defining the Research Objectives
Three types of Survey Questions
– Content Items
– Factual Items
– Behavior Items
Question Wording
•
•
•
•
•
Simplicity
Double-barreled questions
Loaded questions
Negative wording
Yea-saying and Nay-saying
Response to Questions
•
•
•
•
•
Closed vs. Open ended questions
Number of Response Alternatives
Rating Scales
Semantic Differential Scale
Non-verbal scale
Finalizing & Administering
•
•
•
•
Formatting Questionnaire
Refining Questions
Forms
Ways of Administration: Survey vs.
Interview