library(poliscidata)
## Registered S3 method overwritten by 'gdata':
##   method         from  
##   reorder.factor gplots

Q1

Q1A

Decrease, Increase

Q1B

Lower codes on abort_rank3, higher codes on gun_rank3

Q1C

Negative sign, positive sign

Q1D

xtabC(~abort_rank3+cook_index3, statesD)
##             cook_index3
## abort_rank3  More Rep   Even More Dem
##   More restr    58.82  38.89     0.00
##   Mid           35.29  38.89    26.67
##   Less restr     5.88  22.22    73.33
##   Total        100.00 100.00   100.00
##   Count         17.00  18.00    15.00
xtabC(~gun_rank3+cook_index3,statesD)
##             cook_index3
## gun_rank3    More Rep   Even More Dem
##   More restr     0.00  16.67       80
##   Mid           17.65  44.44       20
##   Less restr    82.35  38.89        0
##   Total        100.00 100.00      100
##   Count         17.00  18.00       15
svychisqC(~abort_rank3+cook_index3, statesD)
## 
##  Pearson's X^2: Rao & Scott adjustment
## 
## data:  survey::svychisq(formula, design, statistic = "Chisq", na.rm = TRUE)
## X-squared = 20.85, df = 4, p-value = 0.0003307
svychisqC(~gun_rank3+cook_index3,statesD)
## 
##  Pearson's X^2: Rao & Scott adjustment
## 
## data:  survey::svychisq(formula, design, statistic = "Chisq", na.rm = TRUE)
## X-squared = 34.331, df = 4, p-value = 6.336e-08
somersD(~cook_index3+abort_rank3, statesD)
## Somers'dyx 
##  0.5487365
xtabC(~gun_rank3+cook_index3, statesD)
##             cook_index3
## gun_rank3    More Rep   Even More Dem
##   More restr     0.00  16.67       80
##   Mid           17.65  44.44       20
##   Less restr    82.35  38.89        0
##   Total        100.00 100.00      100
##   Count         17.00  18.00       15
svychisqC(~gun_rank3+cook_index3, statesD)
## 
##  Pearson's X^2: Rao & Scott adjustment
## 
## data:  survey::svychisq(formula, design, statistic = "Chisq", na.rm = TRUE)
## X-squared = 34.331, df = 4, p-value = 6.336e-08
somersD(~cook_index3+gun_rank3, statesD)
## Somers'dyx 
## -0.6979543

## Q1E

… knowing how states rate on the Cook Index increases our ability to explain/predict their gun control ranking by nearly 70%.

Q1F

… there is a 0.0003 probability of observing the relationship between the Cook Index values and the abortion law ranking by chance.  Therefore, you should reject the null hypothesis.

Q1G

Correct.  Legislators in both parties don’t seem to adopt a consistent stance on individual freedom of choice.  States with more Democratic representation tend to decrease restrictions on abortions while increasing restrictions on gun ownership.  State with more Republican representation, in contrast, tend to increase restrictions on abortions while decreasing restrictions on gun ownership.

Q2

Q2A

In a comparison of individuals, women will be more likely than men to think that abortion should be allowed.
In a comparison of individuals, women and men will not differ in the opinions about female roles outside the home.

Q2B

xtabC(~abhlth + sex, gssD)
##        sex
## abhlth    Male Female
##   YES    85.50  87.73
##   NO     14.50  12.27
##   Total 100.00 100.00
##   Count 577.15 656.01
svychisqC(~abhlth + sex, gssD)
## 
##  Pearson's X^2: Rao & Scott adjustment
## 
## data:  survey::svychisq(formula, design, statistic = "Chisq", na.rm = TRUE)
## X-squared = 2.1319, df = 1, p-value = 0.3543
CramersV(2.1319, 2, 2, 1233.16)
## [1] 0.04157897
xtabC(~femrole2 + sex, gssD)
##              sex
## femrole2        Male Female
##   Traditional  55.40  37.64
##   NonTrad      44.60  62.36
##   Total       100.00 100.00
##   Count       592.03 685.81
svychisqC(~femrole2 + sex, gssD)
## 
##  Pearson's X^2: Rao & Scott adjustment
## 
## data:  survey::svychisq(formula, design, statistic = "Chisq", na.rm = TRUE)
## X-squared = 62.392, df = 1, p-value = 9.945e-08
CramersV(62.392, 2, 2, 592.0 + 686.0)
## [1] 0.2209526

Q2C

Pedantic Pontificator’s hypothesis about the femrole2-sex relationship is not supported by the analysis.
Under the assumption that the null hypothesis is correct, the abhlth-sex relationship could have occurred by chance more frequently than 5 times out of 100.
A higher percentage of males than females think that women belong in non-traditional roles.

Q2D

… a relationship as strong as that observed between opinion on allowing abortion for the health of the mother and gender occurs with 35% probability.

Q3

Q3A

gssD.led = subset(gssD, educ_2=="0-12 yrs")
gssD.hed = subset(gssD, educ_2=="13+ yrs")
xtabC(~partyid_3 + egalit_scale3, gssD.led)
##          egalit_scale3
## partyid_3 Less egal Middle More egal
##     Dem       15.21  36.10     38.45
##     Ind       44.05  51.44     48.76
##     Rep       40.74  12.46     12.79
##     Total    100.00 100.00    100.00
##     Count    140.07 157.80    210.10
svychisqC(~partyid_3 + egalit_scale3, gssD.led)
## 
##  Pearson's X^2: Rao & Scott adjustment
## 
## data:  survey::svychisq(formula, design, statistic = "Chisq", na.rm = TRUE)
## X-squared = 95.08, df = 4, p-value = 4.132e-09
somersD(~partyid_3 + egalit_scale3, gssD.led)
## Somers'dyx 
## -0.2431895
xtabC(~partyid_3 + egalit_scale3, gssD.hed)
##          egalit_scale3
## partyid_3 Less egal Middle More egal
##     Dem       19.62  41.11     61.17
##     Ind       32.58  40.78     34.40
##     Rep       47.80  18.10      4.43
##     Total    100.00 100.00    100.00
##     Count    286.65 238.42    197.47
svychisqC(~partyid_3 + egalit_scale3, gssD.hed)
## 
##  Pearson's X^2: Rao & Scott adjustment
## 
## data:  survey::svychisq(formula, design, statistic = "Chisq", na.rm = TRUE)
## X-squared = 230.36, df = 4, p-value < 2.2e-16
somersD(~partyid_3 + egalit_scale3, gssD.hed)
## Somers'dyx 
## -0.3927059

Q3B

All three statements are correct.
At both levels of education, people with stronger egalitarian beliefs are more likely to be Democrats than are people with weaker egalitarian beliefs.
For the less-educated group, random sampling error would produce the observed relationship between egalitarianism and partisanship less frequently than 5 times out of 100.
The partisanship-egalitarianism relationship is stronger for the more-educated group than for the less-educated group.

Q3C

… knowing an individual’s beliefs with respect to egalitarianism increases our ability to explain/predict whether they identify as Democratic by almost 40%.

Q3D

The partisanship-egalitarianism-education relationships are a set of interaction relationships.

The effect of egalitarian beliefs on Democratic identification depends on level of education. Among those with 12 or fewer years of education, stronger belief in egalitarianism corresponds to a 23.6% increase in Democratic identification (the difference between more and less egalitarian beliefs).  Among those with 13 years or more education, the difference is greater, 41.5%.  While varying belief in egalitarianism explains 24% of variation in Democratic identification among those with less education, it explains 39% of the variation among those with more education.