Transcript Chapter 2 - People Server at UNCW

(Source: Causeweb.org)
Elementary Survey Sampling
7th Edition
By Scheaffer, Mendenhall, Ott and
Gerow
Be careful what you believe
• “4 out of 5 doctors recommend…”
• “22% of Americans doubt that the Holocaust ever
occurred…”
(Does it seem possible or does is seem impossible to you
that the Nazi extermination of the Jews never happened?)
• “Alcohol is more harmful than heroin”
(http://www.significancemagazine.org/details/webexclusiv
e/874549/Alcohol-is--more-harmful-than-heroin_-Really-Sample-before-you-speak.html)
• HOWEVER, there are good sources of information (that
attempt to sample appropriately!!), Census Bureau, Bureau
of Labor Statistics, Gallup Poll
Chapter 2: Elements of the Sampling
Problem
Technical terms
• Element – an element is an object on which a
measurement is taken
• Population – a population is a collection of
elements about which we wish to make an
inference
• Sampling units – sampling units are
nonoverlapping collections of elements from the
population that cover the entire population
Definitions continued
• Frame – a frame is a list of sampling units
• Sample – a sample is a collection of sampling
units drawn from a single frame or from
multiple frames
2.3 How to select the sample
• Error of estimation
|   ˆ |  B
• Bound is usually selected as 2s(theta-hat)
• Probability sampling versus Quota Sampling
– SRS
– Stratified random sample
– Systematic sample
2.4: Sources of Errors in Surveys
• Nonobservation (sampled elements make up only part
of the target population; examples include sampling,
coverage, nonresponse)
– Sampling error – variation in sample response because not
from population (can be estimated for probability
samples)..can be reduced by survey design and sample size
– Coverage…frames do not include everyone in target
population (unlisted numbers in telephone directories)
– Nonresponse (inability to contact sampled elements,
inability of person to answer questions correctly, refusal to
answer)
Errors of Observation
• Observation (recorded data deviate from the
truth; examples include interviewer or data
collector, instrument, sampling method)
– Interviewer can bias how questions are asked
– Respondent errors (recall bias, prestige bias, not
understanding question)
– Measurement error (measurements should be
clearly defined…for example, education)
Reducing Errors in Surveys
•
•
•
•
•
Callbacks
Rewards and incentives
Trained interviewers
Data checks
Questionnaire construction
2.5 Designing a Questionnaire
• Question ordering – the ordering of questions can
change perspectives
Example: A. Do you think the US should let Communist
newspaper reporters from other countries come in
here and send back to their papers the news as they
see it?
B. Do you think a Communist country like Russia should
let American newspaper reporters come in and send
back to America the news as they see it?
Ordering (A,B) 54.7% answered yes to A, ordering (B,A)
74.6% answered yes to A
Open versus Closed, Response Options
and Wording of Questions
• Open is good for qualitative information (and to help
design options for closed questions). Closed is good
for quantitative information (however, options should
be appropriately chosen).
• Response Options – Should there be a “Don’t know” or
“Not enough information to say” option
• Wording of questions…should try to ensure it is not
biased
For example: Do you favor the use of capital
punishment?
Do you favor or oppose the use of capital punishment?
2.6 Planning a Survey
1. Statement of Objectives
2. Target population
3. The frame
4. Sample design
5. Method of measurement (usually personal interviews, telephone
interviews, mailed questionnaire, or direct observation)
6. Measurement instrument (if questionnaire, carefully create
questions)
7. Train fieldworkers (if necessary)
8. The pretest
9. Organization of fieldwork (if necessary)
10. Organization of data management
11. Data analysis