Social Factors and Public Opinion

1
Social Factors and Public Opinion

• Michael Milburn
• Psychology 335

2
Why would you expect differences on social
grouping variables?

• For example, age, gender, education, income, race
• Age
• Growth (maturation)
• Living a different times/having different
  experiences (cohort effect)
• Gender differential socialization (Fivush)
• Education/Income
• Different opportunities/interact with different
  people
• Different interests
• Different expectations in life
• Race
• Different opportunities/experiences/interactions
• O.J. Simpson caseNational Jury Project

3
Examining Relationships between Demographic
Variables and Public Opinion

• Create contingency table of responses
• Raw counts in each cell of the table
• Percentage across levels of the dependent
  variable
• Identify the causal model
• DV Typically the opinion variable

4
Do you agree that we should have dropped the
atomic bomb on Japan during WWII?
N 135
Yes Maybe
No
Old Young
25 20 15 20 25 30
5
Do you agree that we should have dropped the
atomic bomb on Japan during WWII?
N 135
Yes Maybe
No
Old Young
25 20 15 20 25 30
Question What is the DV?
6
Do you agree that we should have dropped the
atomic bomb on Japan during WWII?
N 135
Yes Maybe
No
Old Young
25 20 15 20 25 30
Opinion is DV, so calculate row percentages
7
Do you agree that we should have dropped the
atomic bomb on Japan during WWII?
N 135
Yes Maybe
No
Old Young
60 75
25 20 15 20 25 30
Step 1 calculate row totals
8
Do you agree that we should have dropped the
atomic bomb on Japan during WWII?
N 135
Yes Maybe
No
Old Young
60 75
25/60 20/60 15/60 20/75 25/75 30/75
Step 2 divide by row totals
9
Do you agree that we should have dropped the
atomic bomb on Japan during WWII?
N 135
Yes Maybe
No
Old Young

42 33 25 27 33 40
100 100
Step 3 calculate percents
10
Do you agree that we should have dropped the
atomic bomb on Japan during WWII?
N 135
Yes Maybe
No
Old Young

42 33 25 27 33 40
100 100
Question Are proportions the same? Do the same
percentage of old and young people agree?
11
National Election Survey (NES) Data--Examples

• Focus on the ideological (Liberal-Conservative)
  aspect of the relationships among the variables

12
Historical Relationships
13
Further examples

• Opinion questions answered on a 1 to 7 scale
• I grouped together responses (1-2), (3,4,5), and
  (6-7) into three columns

14
1988 NES Data
15
More 1988 Data
16
A final 1988 question
17
Summary

• Different domains of opinion
• Foreign policy/economic/social issues
• For the relationship of income/education
• FP no relationship
• Economic more conservative as income/education
  increase
• Social more liberal as income/education increase