Assignment 9: Reordering Factors

Voter Data

voterData <- readr::read_csv("~\\Major\\data science\\DATA_333\\Abbreviated Dataset Labeled(October Only)(1).csv")

head(voterData)

Using factor() to reorder variables

voterData <- voterData %>% 
  mutate(PartyIdentification = factor(PartyIdentification, levels = c("Democrat",
                                                                      "Republican",
                                                                      "Independent",
                                                                      "Other",
                                                                      "Not Sure")),
         education = factor(education, levels = c("No High School",
                                                  "High School Graduate",
                                                  "Some College",
                                                  "2-year",
                                                  "4-year",
                                                  "Post Grad")),
         region = factor(region, levels = c("Northwest",
                                            "Midwest",
                                            "South",
                                            "West",
                                            "Not in the US")),
         urbancity = factor(urbancity, levels = c("City",
                                                  "Suburb",
                                                  "Town",
                                                  "Rural Area",
                                                  "Other")),
         GayMarriage = factor(GayMarriage, levels = c("Favor",
                                                      "Oppose",
                                                      "Not Sure")))
# Note: Abortion is already correctly ordered

Cross Tabs

(variables: education, PartyIdentification, region, urbancity, Abortion, GayMarriage)

Does level of education affect party identification?

No, across all levels of education, the majority of respondents identify as democrat and the fewest respondents identify as not sure or other. However, respondents with no high school or no college education are more likely to be identify as republican than independent.

kable(voterData %>% 
  group_by(education, PartyIdentification) %>% 
  summarize(n=n()) %>% 
  mutate(percent = n/sum(n))) %>% 
  kable_styling() %>% 
  scroll_box(height = "300px")

education	PartyIdentification	n	percent
2-year	Democrat	299	0.3813776
2-year	Independent	243	0.3099490
2-year	Not Sure	10	0.0127551
2-year	Other	9	0.0114796
2-year	Republican	223	0.2844388
4-year	Democrat	710	0.3600406
4-year	Independent	663	0.3362069
4-year	Not Sure	11	0.0055781
4-year	Other	31	0.0157201
4-year	Republican	557	0.2824544
High School Graduate	Democrat	743	0.3758220
High School Graduate	Independent	512	0.2589782
High School Graduate	Not Sure	49	0.0247850
High School Graduate	Other	25	0.0126454
High School Graduate	Republican	648	0.3277693
No High School	Democrat	62	0.3734940
No High School	Independent	38	0.2289157
No High School	Not Sure	3	0.0180723
No High School	Other	3	0.0180723
No High School	Republican	60	0.3614458
Post Grad	Democrat	452	0.3735537
Post Grad	Independent	413	0.3413223
Post Grad	Not Sure	6	0.0049587
Post Grad	Other	13	0.0107438
Post Grad	Republican	326	0.2694215
Some College	Democrat	733	0.3911419
Some College	Independent	593	0.3164354
Some College	Not Sure	27	0.0144077
Some College	Other	27	0.0144077
Some College	Republican	494	0.2636073
NA	Democrat	5	0.2941176
NA	Independent	9	0.5294118
NA	Republican	3	0.1764706

voterData %>% 
  group_by(education, PartyIdentification) %>% 
  summarize(n=n()) %>% 
  mutate(percent = n/sum(n)) %>% 
  ggplot() + geom_col(aes(x=education, y=percent, fill=PartyIdentification)) + theme(axis.text.x = element_text(angle = 90))

Does level of education vary across regions of the United States?

Yes, the majority of respondents in the Midwest, Northwest, and not in the U.S. have the highest education level of a high school graduate, while the majority of respondents in the South and in the West have the highest education level of a 4-year college. But, across all regions, the fewest number of respondents have no high school education.

kable(voterData %>% 
  group_by(region, education) %>% 
  summarize(n=n()) %>% 
  mutate(percent = n/sum(n))) %>% 
  kable_styling() %>% 
  scroll_box(height = "300px")

region	education	n	percent
Midwest	2-year	181	0.0998345
Midwest	4-year	410	0.2261445
Midwest	High School Graduate	539	0.2972973
Midwest	No High School	33	0.0182019
Midwest	Post Grad	242	0.1334804
Midwest	Some College	405	0.2233867
Midwest	NA	3	0.0016547
Northwest	2-year	117	0.0788941
Northwest	4-year	346	0.2333109
Northwest	High School Graduate	438	0.2953473
Northwest	No High School	22	0.0148348
Northwest	Post Grad	267	0.1800405
Northwest	Some College	291	0.1962239
Northwest	NA	2	0.0013486
Not in the US	2-year	6	0.1111111
Not in the US	4-year	11	0.2037037
Not in the US	High School Graduate	17	0.3148148
Not in the US	No High School	1	0.0185185
Not in the US	Post Grad	9	0.1666667
Not in the US	Some College	10	0.1851852
South	2-year	269	0.1002983
South	4-year	663	0.2472036
South	High School Graduate	651	0.2427293
South	No High School	70	0.0260999
South	Post Grad	367	0.1368382
South	Some College	655	0.2442207
South	NA	7	0.0026100
West	2-year	206	0.1071243
West	4-year	535	0.2782111
West	High School Graduate	322	0.1674467
West	No High School	38	0.0197608
West	Post Grad	314	0.1632865
West	Some College	503	0.2615705
West	NA	5	0.0026001
NA	2-year	5	0.1111111
NA	4-year	7	0.1555556
NA	High School Graduate	10	0.2222222
NA	No High School	2	0.0444444
NA	Post Grad	11	0.2444444
NA	Some College	10	0.2222222

voterData %>% 
  group_by(region, education) %>% 
  summarize(n=n()) %>% 
  mutate(percent = n/sum(n)) %>% 
  ggplot() + geom_col(aes(x=region, y=percent, fill=education))

Do urban and rural areas vary by level of education?

Yes, the majority of repsondents in cities and suburbs have a highest education level of 4-years of college, while the majority of respondents in rural areas, towns, and other places have graduate degrees.

kable(voterData %>% 
  group_by(urbancity, education) %>% 
  summarize(n=n()) %>% 
  mutate(percent = n/sum(n))) %>% 
  kable_styling() %>% 
  scroll_box(height = "300px")

urbancity	education	n	percent
City	2-year	209	0.0925188
City	4-year	584	0.2585215
City	High School Graduate	496	0.2195662
City	No High School	49	0.0216910
City	Post Grad	366	0.1620186
City	Some College	552	0.2443559
City	NA	3	0.0013280
Other	2-year	8	0.1568627
Other	4-year	9	0.1764706
Other	High School Graduate	18	0.3529412
Other	No High School	1	0.0196078
Other	Post Grad	8	0.1568627
Other	Some College	7	0.1372549
Rural Area	2-year	149	0.1004720
Rural Area	4-year	252	0.1699258
Rural Area	High School Graduate	533	0.3594066
Rural Area	No High School	48	0.0323668
Rural Area	Post Grad	145	0.0977748
Rural Area	Some College	350	0.2360081
Rural Area	NA	6	0.0040459
Suburb	2-year	310	0.1032645
Suburb	4-year	838	0.2791472
Suburb	High School Graduate	603	0.2008661
Suburb	No High School	37	0.0123251
Suburb	Post Grad	532	0.1772152
Suburb	Some College	679	0.2261825
Suburb	NA	3	0.0009993
Town	2-year	100	0.0876424
Town	4-year	272	0.2383874
Town	High School Graduate	311	0.2725679
Town	No High School	30	0.0262927
Town	Post Grad	151	0.1323401
Town	Some College	273	0.2392638
Town	NA	4	0.0035057
NA	2-year	8	0.1250000
NA	4-year	17	0.2656250
NA	High School Graduate	16	0.2500000
NA	No High School	1	0.0156250
NA	Post Grad	8	0.1250000
NA	Some College	13	0.2031250
NA	NA	1	0.0156250

voterData %>% 
  group_by(urbancity, education) %>% 
  summarize(n=n()) %>% 
  mutate(percent = n/sum(n)) %>% 
  ggplot() + geom_col(aes(x=urbancity, y=percent, fill=education))

NA
NA

Does level of education affect people’s support of abortion?

No, across all levels of education, the majority of respondents believe that abortion should be legal in some cases and illegal in others, and the fewest respondents are unsure.

kable(voterData %>% 
  group_by(education, Abortion) %>% 
  summarize(n=n()) %>% 
  mutate(percent = n/sum(n))) %>% 
  kable_styling() %>% 
  scroll_box(height = "300px")

education	Abortion	n	percent
2-year	Illegal in all cases	107	0.1364796
2-year	Legal in all cases	254	0.3239796
2-year	Legal in some cases and Illegal in others	376	0.4795918
2-year	Not sure	40	0.0510204
2-year	NA	7	0.0089286
4-year	Illegal in all cases	238	0.1206897
4-year	Legal in all cases	771	0.3909736
4-year	Legal in some cases and Illegal in others	851	0.4315416
4-year	Not sure	93	0.0471602
4-year	NA	19	0.0096349
High School Graduate	Illegal in all cases	335	0.1694487
High School Graduate	Legal in all cases	550	0.2781993
High School Graduate	Legal in some cases and Illegal in others	929	0.4699039
High School Graduate	Not sure	148	0.0748609
High School Graduate	NA	15	0.0075873
No High School	Illegal in all cases	40	0.2409639
No High School	Legal in all cases	31	0.1867470
No High School	Legal in some cases and Illegal in others	77	0.4638554
No High School	Not sure	16	0.0963855
No High School	NA	2	0.0120482
Post Grad	Illegal in all cases	139	0.1148760
Post Grad	Legal in all cases	505	0.4173554
Post Grad	Legal in some cases and Illegal in others	515	0.4256198
Post Grad	Not sure	45	0.0371901
Post Grad	NA	6	0.0049587
Some College	Illegal in all cases	224	0.1195304
Some College	Legal in all cases	678	0.3617930
Some College	Legal in some cases and Illegal in others	835	0.4455710
Some College	Not sure	128	0.0683031
Some College	NA	9	0.0048026
NA	Illegal in all cases	4	0.2352941
NA	Legal in all cases	4	0.2352941
NA	Legal in some cases and Illegal in others	6	0.3529412
NA	Not sure	3	0.1764706

voterData %>% 
  group_by(education, Abortion) %>% 
  summarize(n=n()) %>% 
  mutate(percent = n/sum(n)) %>% 
  ggplot() + geom_col(aes(x=education, y=percent, fill=Abortion)) +theme(axis.text.x = element_text(angle = 90))

Does level of education affect people’s support of gay marriage?

Yes, the majority of respondents with a highest education levels of some college, a 4-years college degree, and post-graduate degrees favor gay marriage, while the majority of respondents with a highest education levels of no high school, high school, or a 2-years college degree oppose gay marriage.

kable(voterData %>% 
  group_by(education, GayMarriage) %>% 
  summarize(n=n()) %>% 
  mutate(percent = n/sum(n))) %>% 
  kable_styling() %>% 
  scroll_box(height = "300px")

education	GayMarriage	n	percent
2-year	Favor	332	0.4234694
2-year	Not sure	102	0.1301020
2-year	Oppose	346	0.4413265
2-year	NA	4	0.0051020
4-year	Favor	1018	0.5162272
4-year	Not sure	211	0.1069980
4-year	Oppose	735	0.3727181
4-year	NA	8	0.0040568
High School Graduate	Favor	685	0.3464846
High School Graduate	Not sure	276	0.1396055
High School Graduate	Oppose	1009	0.5103692
High School Graduate	NA	7	0.0035407
No High School	Favor	46	0.2771084
No High School	Not sure	21	0.1265060
No High School	Oppose	98	0.5903614
No High School	NA	1	0.0060241
Post Grad	Favor	657	0.5429752
Post Grad	Not sure	132	0.1090909
Post Grad	Oppose	416	0.3438017
Post Grad	NA	5	0.0041322
Some College	Favor	849	0.4530416
Some College	Not sure	211	0.1125934
Some College	Oppose	811	0.4327641
Some College	NA	3	0.0016009
NA	Favor	6	0.3529412
NA	Not sure	1	0.0588235
NA	Oppose	9	0.5294118
NA	NA	1	0.0588235

voterData %>% 
  group_by(education, GayMarriage) %>% 
  summarize(n=n()) %>% 
  mutate(percent = n/sum(n)) %>% 
  ggplot() + geom_col(aes(x=education, y=percent, fill=GayMarriage)) + theme(axis.text.x = element_text(angle = 90))