Transcript Survey Methodology

Survey Methodology
EPID-626
Lecture 1
What is a survey?
• A survey is a system for collecting
information to describe, compare, or
explain knowledge, attitudes, and
behavior.
(Fink, 1995)
1/12/00
What can surveys reasonably
assess?
• Behaviors associated with disease
• Personal attributes that affect disease
risk
• Knowledge/attitudes that influence
health behaviors
• Use of health services
• Self-reported disease occurrence
(Rothman, 1998)
1/12/00
What may be unreasonable
for survey?
• Very personal information
• Information about socially undesirable
behavior
• Intricate details of past events
• Complex clinical information
1/12/00
Options for topics that may be
unreasonable for survey
• Don’t use a survey
– Example: examine medical records instead
• Use survey but use other methods to
verify some information
– Example: use a survey to measure
possible cancer risk factors and ask about
history of cancer. Examine medical records
to verify reported information about history
of cancer
1/12/00
• Compensate for potential problems in
the survey design
– Example: make the survey anonymous if
asking for very personal information or
about socially undesirable behavior
• Use a survey, but recognize the
limitations of the results
– Example: present prevalence rate, but
comment that it is likely to underestimate
the true prevalence
1/12/00
“Sponsoring a special-purpose survey
data collection is a rather expensive
solution to an information problem.
Before launching such an effort, one
should explore thoroughly the potential
for gathering the same information from
existing records or from other sources.”
(Fowler, 1995)
1/12/00
Alternatives to surveys
• May be necessary
– biologic specimens, medical testing
• May be more feasible
– existing data sources
• Multiple methods may be used together
1/12/00
Advantages of surveys
• You can:
– obtain standardized measurements
– use probability sampling
– ensure that all the data needed for a
given analysis can be obtained
1/12/00
Your research question
• Is the question
– relevant?
– important?
– timely?
• Always begin with a literature review
• Formulate a hypothesis
• Can you test the hypothesis with a
survey?
1/12/00
Feasibility
• Can you obtain the necessary
information to answer the research
question?
– Can you get the information?
– Will the data be accurate?
– How much will it cost (staff, other
resources)?
– How long will it take?
– Is your study ethical?
1/12/00
Ready to start? Not even
close!!
• Think about your
research endpoint
• What kind of data do
you need?
• How will you collect
the data?
• Who will you collect
data from?
1/12/00
Wanted: Project Coordinator
• Assisting the department director in carrying out
applied survey research, your duties will include:
formulating study designs and sampling frames;
..developing instruments; supervising data
collection, entry, and manipulation; application
of descriptive and inferential statistics;
interpreting results and preparing reports.
(Fink, 1995)
1/12/00
Survey To-Do List
• Set objectives for information collection
• Design research
• Prepare a reliable and valid data
collection instrument
• Analyze data
• Report the results
(Fink, 1995)
1/12/00
Checklist for a great survey
•
•
•
•
•
•
•
•
Specific objectives
Straightforward questions
Sound research design
Sound choice of population or sample
Reliable and valid survey instrument
Appropriate analysis
Accurate reporting of survey results
Reasonable resources
1/12/00
(Fink, 1995)