Bivariate Regression

1
Bivariate Regression
Vote 08
Vote 04
2
Bivariate Regression
Liberalism
Vote 08
3
Trivariate Regression
Liberalism
Vote 08
Vote 04
4
Crosstab Liberalism and Presidential Vote
1.00 Extremely conservative 2.00 Conservative 3.00 Slightly conservative 4.00 Moderate 5.00 Slightly liberal 6.00 Liberal 7.00 Extremely liberal
1. John McCain 95.4 74.1 63.0 41.2 17.2 15.0 1.2
2. Barack Obama 4.6 25.9 37.0 58.8 82.8 85.0 98.8
Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0
5
Correlation Liberalism and Presidential Vote
6
Bivariate Regression Liberalism and Presidential
Vote
7
Crosstab Presidential Vote (2004) and
Presidential Vote (2008)
Bush Kerry
1. John McCain 80.9 7.5
2. Barack Obama 19.1 92.5
Total 100.0 100.0
8
Correlation Presidential Vote (2004) and
Presidential Vote (2008)
9
Bivariate Regression Vote 04 and Presidential
Vote
10
Crosstab Presidential Vote (2004) and
Presidential Vote (2008), controlling for
liberalism
Liberalism Bush Kerry
Conservative 1. John McCain 89.4 16.2
Conservative 2. Barack Obama 10.6 83.8
Conservative Total 100.0 100.0

Moderate 1. John McCain 59.3 15.6
Moderate 2. Barack Obama 40.7 84.4
Moderate Total 100.0 100.0

Liberal 1. John McCain 51.7 3.6
Liberal 2. Barack Obama 48.3 96.4
Liberal Total 100.0 100.0
11
Correlation Presidential Vote (2004) and
Presidential Vote (2008), controlling for
liberalism
12
Crosstab Presidential Vote (2008) and
Liberalism, controlling for Presidential Vote
(2004)
1.00 Conservative 2.00 Moderate 3.00 Liberal
1. John McCain 89.4 59.3 51.7
2. Barack Obama 10.6 40.7 48.3
Total 100.0 100.0 100.0
 
1. John McCain 16.2 15.6 3.6
2. Barack Obama 83.8 84.4 96.4
Total 100.0 100.0 100.0 
13
Correlation Presidential Vote (2008) and
Liberalism, controlling for Presidential Vote
(2004)
14
Correlations
15
Explaining Venn Diagrams
Variance Vote 08
16
Explaining Venn Diagrams
Variance Vote 04
17
Explaining Venn Diagrams
Variance Liberalism
18
Plot showing positive covariance
Mean urban
Mean female literacy
19
Plot showing negative covariance
Mean calorie intake
Mean infant mortality
20
Plot showing no covariance
Mean GDP
Mean crop production
21
Pearson r .62
Vote 04
Liberalism
Vote 08
Pearson r .73
Vote 04
Vote 08
Pearson r .61
Liberalism
22
Vote 08
Vote 04
Liberalism
23
Multivariate Regression Presidential Vote
(2008), controlling for Liberalism and
Presidential Vote (2004)
24
Appendix Syntax
The following set of commands changes a nominal
variable into an ordinal variable. Prespref_2004
had six categories for their presidential
preference, including Nader and others. I needed
a variable that just included Bush and Kerry.
if prespref_2004 1 presvote_2004 1. if
prespref_2004 2 presvote_2004 2. execute. I
wanted a variable that went in the same direction
as presvote_2004 and presvote so I recoded
ideology to make liberalism coded high. recode
ideology (71) (62) (53) (44) (35) (26)
(17) into liberalism. execute. Here I am adding
value labels for liberalism. add value labels
liberalism 1 'Extremely conservative' 2
'Conservative' 3 'Slightly conservative' 4
'Moderate' 5 'Slightly liberal' 6 'Liberal' 7
'Extremely liberal' . Here I am creating a three
category variable for libealism so the crosstab
is easier to understand. recode liberalism (11)
(21) (31) (42) (53) (63) (73) into
liberalismc. add value labels liberalismc 1
'Conservative' 2 'Moderate' 3 'Liberal'
. Here I am running a crosstab with presidential
vote as the dependent variable and liberalism as
the independent variable. crosstabs tables
presvote by liberalism/cells col /stats
corr. Here I am running a crosstab with
presidential vote as the dependent variable and
2004 vote as the independent variable. crosstabs
tables presvote by presvote_2004/cells col
count/stats corr. Here I am running a
correlation between the three variables. corr
presvote presvote_2004 liberalism. Here I am
running a crosstab with presidential vote as the
dependent variable and 2004 vote as the
independent variable, controlling for liberalism.
crosstabs tables presvote by presvote_2004 by
liberalismc/cells col /stats corr. Here I am
running a crosstab with presidential vote as the
dependent variable and liberalism as the
independent variable, controlling for 2004 vote.
crosstabs tables presvote by liberalismc by
presvote_2004/cells col /stats corr. Here I
am running a regression with presidential vote as
the dependent variable and liberalism and vote 04
as the independent variables. regr vars
presvote presvote_2004 liberalism / depe
presvote/enter.