-
Notifications
You must be signed in to change notification settings - Fork 1
/
Regression 5 year data
73 lines (47 loc) · 2.21 KB
/
Regression 5 year data
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
#NARROW DOWN COVRIATES TO INCLUDE IN FINAL MODEL
load("Covariates.rda")
all<-all[,c(1,3:12)]
a<-read.csv("draft_ethnicity_brexit_05.csv")
test5<-merge(a,all,all.x=TRUE)
attach(test5)
RegModel.1 <-
lm(Pct_Leave.x~rate_nonUk+age_grp18_24+age_grp25_34+age_grp35_44+age_grp45_54+age_grp55_64+
age_grp65_over+Area.Sq.Km+ Mid.2016.Population+Pct_Turnout.x+Rejected_Ballots+
People.per.Sq.Km, data=test5)
#STEPWISE SIMPLIFICATION
drop1(RegModel.1, test="F")
#At the 5 year level, drop1 proposed dropped variables are Age group 45-54
RegModel.2<-update(RegModel.1,~.-age_grp45_54)
anova(RegModel.1,RegModel.2,test="Chi") #P=0.8306
anova(RegModel.2) #Mid 2016 population seems non significant
RegModel.3<-update(RegModel.2,~.-Mid.2016.Population)
anova(RegModel.2,RegModel.3,test="Chi") #P<0.001
#Significant: rate_nonUk, all age bands except 45-54, Area.sq.km
#Mid2016 Population, Pct_Turnout.x, change10,Mid.2016.Population,Pct_Turnout.x,Rejected_Ballots,People.per.Sq.Km
#EXPAND ETHNICITY (REPLACE rate_nonUk with ethnicity specific rates)
load("Brexit05Covariates.rda")
RegModel.5<-lm(formula = Pct_Leave.y ~ age_grp18_24 + age_grp25_34 + age_grp35_44 +
age_grp55_64 + age_grp65_over + Area.Sq.Km +
Asian + Black + Mid.2016.Population + Other + Pct_Turnout.y + People.per.Sq.Km +
Rejected_Ballots +
White.British + White.Other, data = fiveall)
drop1(RegModel.5, test="F")
RegModel.6<-update(RegModel.5,~.-Black)
anova(RegModel.5,RegModel.6,test="Chi") #P=0.9657
RegModel.7<-update(RegModel.6,~.-Other)
anova(RegModel.6,RegModel.7,test="Chi") #P=0.4993
RegModel.8<-update(RegModel.7,~.-Asian)
anova(RegModel.7,RegModel.8,test="Chi") #P=0.2445
drop1(RegModel.8, test="F")
RegModel.9<-update(RegModel.8,~.-age_grp55_64)
anova(RegModel.8,RegModel.9,test="Chi") #P=0.3672
RegModel.10<-update(RegModel.9,~.-age_grp25_34)
anova(RegModel.9,RegModel.10,test="Chi") #P=0.2261
RegModel.11<-update(RegModel.10,~.-Mid.2016.Population)
anova(RegModel.10,RegModel.11,test="Chi") #P=0.1029
drop1(RegModel.11, test="F")
summary(RegModel.11)
RegModel.12<-update(RegModel.11,~.-Mid.2016.Population)
anova(RegModel.11,RegModel.12,test="Chi") #P=0.0315 #Cannot be dropped
drop1(RegModel.12, test="F")
summary(RegModel.12)