Developing the Sampling Plan

1
Developing the Sampling Plan
Chapter 15
2
Review Stages in Research Process
Formulate Problem
Determine Research Design
Determine Data Collection Method
Design Data Collection Forms
Design Sample Collect Data
Analyze and Interpret Data
Prepare Written/Oral Report
3
Preparatory Questions
What is your Population?
4
Preparatory Questions
Census Vs. Sample
5
Preparatory Questions
How are a Parameter, a Statistic and Sampling
Error related?
6
Preparatory Questions
What is a Sampling Frame?
7
Case 5-1
Newt vs. Toade
8
Case Discussion
1. What is the target population?
The second paragraph reveals the target
population is all registered voters in the U.S.

9
Case Discussion

• What population parameter is
• being estimated and what
• statistic is being calculated?

The overall job approval rating of Senator
Newton.
10
Case Discussion

• What sampling frames were
• used in each of the two studies?

Toade A comprehensive national list of people
for each of the 5 income categories.
Newton A random list of 250 subscribers from the
newspaper in each state with the largest number
of subscribers.
11
Case Discussion
4. What type of sampling procedure was
employed in the study conducted by Toade
and Associates?
12
Toade and Assoc.
What is the difference between a Nonprobability
Sample and a Probability Sample?
13
Probability vs. NonProbability
If personal judgment is used in figuring out who
will be sampled, its not a Probability Sample
vs.
14
Toade and Assoc.
What is the difference between a Quota Sample
and a Stratified Sample?
15
Quota vs. Stratified
We need 100 samples 15 under 25K 20 gt25 and
lt50K 50 gt50K and lt100K 15 over 100K 100 or
100 samples
We need 25 random samples from each 1. under
25K 2. gt25 and lt50K 3. gt50K and lt100K 4.
over 100K
16
Quota vs. Stratified
Polling Democrats and Republicans
17
Toade and Assoc.
What is the difference between a Stratified
Sample and a Cluster Sample?
18
Stratified vs. Cluster
Randomly choose 2 of 1-4, then sample 50 from
each 1. under 25K 2. gt25 and lt50K 3. gt50K
and lt100K 4. over 100K
We need 25 random samples from each 1. under
25K 2. gt25 and lt50K 3. gt50K and lt100K 4.
over 100K
19
Stratified vs. Cluster
This is an Area Cluster
20
Toade and Assoc.
What is the difference between a Simple Random
Sample and a Stratified Sample?
21
Stratified vs. Simple Random
Randomly generate 100 names from the population
We need 25 random samples from each 1. under
25K 2. gt25 and lt50K 3. gt50K and lt100K 4.
over 100K
22
Case Discussion
5. Do you believe the sampling procedure
employed in the Toade study was
appropriate given the objective of
measuring the Senators overall job
approval rating?
23
Case Discussion
6. What type of sampling procedure was
employed in the study conducted by the
research firm on behalf of Senator
Newton?
24
Discussion Prompter
Is this a Probability Sample or a
Nonprobability Sample?
25
Systematic Sample
The case says that the names were generated
randomly, how could they have generated the names
systematically (a systematic sample?)
26
Systematic Sample
Need 100 have list of 1000.
Calculate Interval
Randomly choose 1-10 to begin
Calculate any additional names needed to get to
100 respondents
27
Systematic Sample
Need 100 have list of 1000.
Calculate Interval
i 1000/100 10
If only 85 are expected to answer the phone
100/.85118
i 1000/118 8
28
Nonprobability Sample
Is this is a Convenience Sample?
29
Discussion Prompter
Is this a Quota Sample?
30
NonProbability Sample
Is this a Snowball Sample?
31
Snowball Sample
32
Discussion Prompter
Is this is a Judgment Sample?
33
Case Discussion
7. Do you believe the sampling procedure
employed in the study conducted for
Senator Newton was appropriate given the
objective of measuring the Senator's
overall job approval rating?
34
Case Discussion
8. Which of the two studies do you
believe produced the numbers that are
most trustworthy?
35
Case Discussion
9. Design a sampling plan that could be
used in a new study that you believe
would provide the fairest way to resolve
this case. Describe in detail your
reasoning for proposing this method of
sampling.
36
How Would Conduct the Study?
1. Define Target Population 2. Define the
Sampling Frame 3. Select a Sampling Procedure
Probability Sample
Nonprobability Sample

• Convenience
• Judgment
• Quota

• Simple Random
• Systematic
• Stratified
• Cluster
• Area

37
Exam Review Chapter 9

• Goal Understand Behavior
• Observation vs. Communication
• Types of Primary Data that can be collected using
  communication to understand behavior

38
Exam Review Chapter 12

• 1. Four Scales
• Understand meaning
• How to analyze
• Be able to identify
• 2. Understand validity and error

39
Exam Review Chapter 13

• Independent Ratings Scales
• Itemized-Ratings Scale Options
• General Itemized-Ratings Scales
• Summated-Ratings (Likert) Scale
• Semantic-Differential Scale
• Graphic-Ratings Scale
• Comparative Ratings Scales
• Constant Sum Scale
• Magnitude Scale

When, why, scale Identify
40
Exam Review Chapter 14
Questionnaire Design - Understand steps 1-10 Step
3 Determine Content of Individual Questions
Understand Issues and how to handle them Step 4
Determine Form of Response to Each Question
Understand Open vs. Closed and order bias Step 5
Determine Wording Each Question Understand
problems with wording Step 6 Determine Question
Sequence understand factors that make
successful sequence Step 7 Determine Physical
Characteristics of Questionnaire understand
factors that make successful physical
characteristics