How to Get a Good Sample

1
Chapter 4

• How to Get a Good Sample

2
Thought Questions
2, page 48
Suppose a properly chosen sample of 1600 people
across the United States was asked if they
regularly watch a certain television program, and
24 said yes. How close do you think that is to
the percentage of the entire country who watch
the show? Within 30? 10? 5? 1? Exactly the
same?
3
Thought Questions
3, page 48
Many television stations conduct polls by asking
viewers to call one phone number if they feel one
way about an issue and a different phone number
if they feel the opposite. Do you think the
results of such a poll represent the feelings of
the community? Do you think they represent the
feelings of all those watching the TV station at
the time or of some other group? Explain.
4
Thought Questions
4, page 48
Suppose you had a telephone directory listing all
of the businesses in a city, alphabetized by type
of business. If you wanted to phone 100 of them
in order to get a representative sampling of
opinion on some issue, how would you select which
100 to phone? Why would it not be a good idea to
simply use the first 100 businesses listed?
5
Common Language

• Population
• Unit
• Sampling Frame
• Sample
• Sample Survey
• Census

6
Population
7
Sampling Frame
List of Units 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
16 17
8
Sample Survey
9
Census
List of Units 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
16 17
Census
10
Margin of Error

• The amount by which the proportion obtained from
  the sample will differ from the true population
  proportion rarely exceeds the margin of error.
• Typical margin of error 1/sqrt(n)
• The sample proportion differs from the population
  proportion by more than the margin of error in
  less than 5 (1 in 20) of surveys.

11
Case Study
Satisfaction with Way Things Are Going in U.S.
46 are satisfied up 12 points from February
and representing a record high 47 are
dissatisfied a record low
MarketSearch SC Poll, September
1996 http//www.msearch.com/pr100996.html
12
Case Study
How the Poll was Conducted The MarketSearch
Poll of South Carolina is a semi-annual telephone
survey of 800 consumers statewide. The Poll from
which these findings are taken was conducted in
late August and early September, 1996. The survey
has a sampling error of 3.5 percent.
13
Probability Sampling Plans

• Simple random sampling (SRS)
• Stratified random sampling
• Cluster sampling
• Systematic sampling
• Random digit dialing
• Multistage sampling

14
Difficulties and Disasters in Sampling

• Difficulties
• Using the wrong sampling frame
• Not reaching the individuals selected
• Getting no response or getting a volunteer
  response
• Disasters
• Getting a volunteer sample
• Using a convenience or haphazard sample

15
Using the Wrong Sampling Frame
Including some units not in the population.
Undercoverage Excluding some units in the
population.
16
Difficulty

• To prepare for her book Women and Love, Shere
  Hite sent questionnaires to 100,000 women asking
  about love, sex, and relationships.
• 4.5 responded
• Hite used those responses to write her book
• Moore (Statistics Concepts and Controversies,
  1997) noted
• respondents were fed up with men and eager to
  fight them
• the anger became the theme of the book
• but angry women are more likely to respond

17
Disaster

• Brooks Shoes study Utts Case Study 2.1, p. 26
• convenience sample
• 71 recognized a Brooks shoe
• Properly selected sample by Suave Shoe Corp.
• 2.7 recognized a Brooks shoe

18
Key Concepts

• Populations and Samples
• Samples can be used to infer something about a
  population
• Margin of error
• Sample surveys
• difficulties and disasters