IRB

1
IRB

• Protocol
• Need to use the online form
• Consent
• Questionnaires

2
Longitudinal vs. Cross Sectional
2003 20yrs 40yrs (20yrs in
1983) 60yrs (20yrs in 1963)
Just relying on cross-sectional data alone can be
erroneous.
3
Survey Designs and Subject Recruitment
4
Selecting Research Subjects

• Opportunity samples
• Population
• Sample
• Probability sampling
• Distinction betw random sampling and random
  assignment

5
Point vs. Interval Estimate

• Point estimate
• Estimate particular characteristic of population
• Number
• in sample x population N
• Example
• Interval estimate
• How approximate the point estimate is
• Percentage interval
• Example

6
Basic Concepts in Survey Samples

• Sample unbiased
• True value of population
• example
• Stability
• Minimal spread

7
Lessons from Biased Sampling

• Literary Digest vs. Gallup Poll (quota sampling)
• Army draft

8
Sampling

• Simple random sampling
• Stratification sampling
• Area probability sampling

9
Simple random sampling

• Simple undivided population
• Random
• Option
• Replacement
• No replacement
• When to use what?
• How?
• Random dialing
• Random drawing of names on paper
• Random number

10
Stratification Sampling

• Divide population in strata
• Gender, political affiliation, ethnicity, etc.
• Statistically weighted population

11
Area Probability Sampling

• Divide into geographical areas

12
Nonresponse Bias

• how to deal with it?