load("C:\\Users\\Kevin\\Documents\\R\\immigration\\gap_full.RData")
objects()
dim(gap)
names(gap)
summary(gap)

gap2 <- read.delim(file="gap2.txt")


newgap <- gap[ , c("country", "q14h", "q14b", "q23a", "q23f", "q24e", "q24f", "q107","q118arg","q118ban","q118bol","q118bra","q118bul","q118can","q118chil","q118chin","q118cze","q118egy","q118eth","q118fra","q118gbr","q118ger","q118gha","q118inda","q118indo","q118isr","q118ita","q118ivo","q118jap","q118jor","q118ken","q118kuw","q118leb","q118mala","q118mali","q118mex","q118mor","q118nig","q118pak","q118pal","q118per","q118pol","q118rus","q118safr","q118sen","q118skor","q118slo","q118spa","q118swe","q118tan","q118tur","q118uga","q118ukr","q118us","q118ven","q120","q120spa","q120us","q120inda","q134usr","q134ur","q134bul","q134inda","q134jap","q134ken","q134mex","q134skor","q134safr", "q20")]


newgap$country <- factor(newgap$country, levels=47:1, labels=names(attributes(newgap$country)$value.labels))
table(newgap$country)
summary(newgap)

##################################
###Attitudes towards immigrants###
##################################

table(newgap$country, newgap$q23f)
migr.x <- numeric()
migr.x[newgap$q23f==1] <- 0.25
migr.x[newgap$q23f==2] <- 0.50
migr.x[newgap$q23f==3] <- 0.75
migr.x[newgap$q23f==4] <- 1
migr.x[newgap$q23f==8] <- 0
migr.x[newgap$q23f==9] <- NA
table(newgap$country, migr.x)

migr1 <- tapply(migr.x, newgap$country, mean,na.rm=TRUE)
dotchart(migr1[order(migr1)], color=c("red","blue","darkgreen"),  xlab = "Proportions Agreeing",cex=0.8)
plot(density(migr1, na.rm=TRUE))


###################################
###Attitudes towards immigration###
###################################

table(newgap$country, newgap$q14h)
immi.x <- numeric()
immi.x[newgap$q14h==1] <- 1
immi.x[newgap$q14h==2] <- 0.75
immi.x[newgap$q14h==3] <- 0.50
immi.x[newgap$q14h==4] <- 0.25
immi.x[newgap$q14h==8] <- 0
immi.x[newgap$q14h==9] <- NA

table(newgap$country, immi.x)

immi1 <- tapply(immi.x, newgap$country, mean,na.rm=TRUE)
dotchart(immi1[order(immi1)], color=c("red","blue","darkgreen"),  xlab = "Proportions  agreeing",cex=0.8, )
plot(density(immi1, na.rm=TRUE))


################################################
###Attitudes towards immigration restrictions###
################################################

table(newgap$country, newgap$q24f)
rest.x <- numeric()
rest.x[newgap$q24f==1] <- 1
rest.x[newgap$q24f==2] <- 0.75
rest.x[newgap$q24f==3] <- 0.50
rest.x[newgap$q24f==4] <- 0.25
rest.x[newgap$q24f==8] <- 0
rest.x[newgap$q24f==9] <- NA

table(newgap$country, rest.x)
rest1 <- tapply(rest.x, newgap$country, mean,na.rm=TRUE)
dotchart(rest1[order(rest1)], color=c("red","blue","darkgreen"), xlab = "Proportions agreeing ",cex=0.8)
plot(density(rest1, na.rm=TRUE))



#########################################################
###Attitudes towards influence from foreign protection###
#########################################################

table(newgap$country, newgap$q24e)
prot.x <- numeric()
prot.x[newgap$q24e==1] <- 1
prot.x[newgap$q24e==2] <- 0.75
prot.x[newgap$q24e==3] <- 0.50
prot.x[newgap$q24e==4] <- 0.25
prot.x[newgap$q24e==8] <- 0
prot.x[newgap$q24e==9] <- NA

table(newgap$country, prot.x)
prot1 <- tapply(prot.x, newgap$country, mean,na.rm=TRUE)
prot1
dotchart(prot1[order(prot1)], color=c("red","blue","darkgreen"),  xlab = "Proportions agreeing",cex=0.8)
plot(density(prot1, na.rm=TRUE))

prot <- tapply(prot.x, newgap$code, mean,na.rm=TRUE)


############################################
###Attitudes towards domestic way of life###
############################################

table(newgap$country, newgap$q20)

life.x <- numeric()
life.x[newgap$q20==1] <- 1
life.x[newgap$q20==2] <- 0
life.x[newgap$q20==8] <- NA
life.x[newgap$q20==9] <- NA

table(newgap$country, life.x)
life <- tapply(life.x, newgap$country, mean,na.rm=TRUE)
dotchart(life[order(life)], color=c("red","blue","darkgreen"),  xlab = "Proportions agreeing",cex=0.8)
mean(life)
plot(density(life, na.rm=TRUE))

##########################################################################################
###Attitudes towards conflict between ethnic/racial/nationality/religious/tribal groups###
##########################################################################################

table(newgap$country, newgap$q14b)
newgap$q14b
conf.x <- numeric()
conf.x[newgap$q14b==1] <- 1
conf.x[newgap$q14b==2] <- 0.75
conf.x[newgap$q14b==3] <- 0.50
conf.x[newgap$q14b==4] <- 0.25
conf.x[newgap$q14b==8] <- 0
conf.x[newgap$q14b==9] <- NA

table(newgap$country, conf.x)

conf1 <- tapply(conf.x, newgap$country, mean,na.rm=TRUE)
conf1
dotchart(conf1[order(conf1)], color=c("red","blue","darkgreen"), xlab = "Proportions agreeing", cex=0.8)
plot(density(conf1, na.rm=TRUE))



###Average immigration measures###
avrg <- (immi1+migr1+prot1+rest1+conf1+life)/6
avrg
dotchart(avrg[order(avrg)], color=c("red","blue","darkgreen"),  xlab="Average proportions", cex=0.8)





newgap2 <- data.frame(migr1, immi1, rest1, prot1, life, conf1)
newgap3 <- newgap2[c(1:39, 41:47), ]

cor(newgap3[ , c("migr1", "immi1", "rest1", "prot1", "life", "conf1")])

cor(newgap3[ , c("prot1", "life", "conf1")])

###Correlations

par(mfrow=c(1,2))
plot(immi1 ~ migr1, type="n", xlab = "Immigrants are a bad influence in this country", ylab = "Immigration is a problem in this country")
text(migr1, immi1, labels=gap2$code)
abline(lm(immi1 ~ migr1), col="red")

plot(immi1 ~ rest1, type="n", xlab ="Immigration is a problem in this country", ylab ="This country needs more restrictions and Control of Immigrants" )
text(rest1, immi1, labels=gap2$code)
abline(lm(immi1 ~ rest1), col="red")
dev.off()

par(mfrow=c(1,2))
###Protection from foreign interference vs way of life being affected###
plot(rest1 ~ prot1, type="n", xlab="This country needs more restrictions and Control of Immigrants", ylab="This country needs protection against foreign influence")
text(prot1, rest1, labels=gap2$code)
abline(lm(rest1 ~ prot1), col="red")

###Way of life vs Protection from foreign influence###
plot(life~prot1, type="n", xlab="Traditional way of life remains strong", ylab="This country needs protection against foreign influence")
text(prot1,life, labels=gap2$code)
abline(lm(life ~ prot1), col="red")
dev.off()

par(mfrow=c(1,2))

plot(migr1~conf1, type="n", xlab="Immigration is a problem in this country", ylab="Conflict is a problem in this country")
text(conf1, migr1, labels=gap2$code)
abline(lm(migr1 ~ conf1, data=newgap), col="red")

plot(immi1~conf1, type="n", xlab="Immigrants are a bad influence in this country", ylab="Conflict is a problem in this country")
text(conf1, immi1, labels=gap2$code)
abline(lm(immi1 ~ conf1, data=newgap), col="red")

dev.0ff()
dev.off()
###Chapter 4###

sa.only <- subset(newgap, newgap$country=="South Africa")

###Immmigration###
table(sa.only$q14h)
immi2.x <- numeric()
immi2.x[sa.only$q14h==1] <- 1
immi2.x[sa.only$q14h==2] <- 1
immi2.x[sa.only$q14h==3] <- 0
immi2.x[sa.only$q14h==4] <- 0
immi2.x[sa.only$q14h==8] <- NA
immi2.x[sa.only$q14h==9] <- NA

table(immi2.x)
immi2.labs <- c("Problem", "Not a problem")
immi2 <- factor(immi2.x, levels=c(1,0), labels=immi2.labs)
table(immi2)

immi2.1 <- table(immi2)
immi.tab <- prop.table(immi2.1)
round(immi.tab, digits=2)



###Immigrants###
table(sa.only$q23f)
migr2.x <- numeric()
migr2.x[sa.only$q23f==1] <- 1
migr2.x[sa.only$q23f==2] <- 1
migr2.x[sa.only$q23f==3] <- 0
migr2.x[sa.only$q23f==4] <- 0
migr2.x[sa.only$q23f==8] <- NA
migr2.x[sa.only$q23f==9] <- NA

table(migr2.x)
migr2.labs <- c("Bad", "Good")
migr2 <- factor(migr2.x, levels=c(0,1), labels=migr2.labs)
migr2.1 <- table(migr2)
migr2.tab <- prop.table(migr2.1)
round(migr2.tab, digits=2)



###Protection from foreign influence###
table(sa.only$q24e)
prot2.x <- numeric()
prot2.x[sa.only$q24e==1] <- 1
prot2.x[sa.only$q24e==2] <- 1
prot2.x[sa.only$q24e==3] <- 0
prot2.x[sa.only$q24e==4] <- 0
prot2.x[sa.only$q24e==8] <- NA
prot2.x[sa.only$q24e==9] <- NA

table(prot2.x)
prot2.labs <- c("Agree", "Disagree")
prot2 <- factor(prot2.x, levels=c(1,0), labels=prot2.labs)
prot2.1 <- table(prot2)
prot.tab <- prop.table(prot2.1)
round(prot.tab, digits=2)


###Restrictions###
table(sa.only$q24f)
rest2.x <- numeric()
rest2.x[sa.only$q24f==1] <- 1
rest2.x[sa.only$q24f==2] <- 1
rest2.x[sa.only$q24f==3] <- 0
rest2.x[sa.only$q24f==4] <- 0
rest2.x[sa.only$q24f==8] <- NA
rest2.x[sa.only$q24f==9] <- NA

table(rest2.x)
rest2.labs <- c("Agree", "Disagree")
rest2 <- factor(rest2.x, levels=c(1,0), labels=rest2.labs)
rest2.1 <- table(rest2)
rest.tab <- prop.table(rest2.1)
round(rest.tab, digits=2)


###Conflict###
table(sa.only$q14b)
conf2.x <- numeric()
conf2.x[sa.only$q14b==1] <- 1
conf2.x[sa.only$q14b==2] <- 1
conf2.x[sa.only$q14b==3] <- 0
conf2.x[sa.only$q14b==4] <- 0
conf2.x[sa.only$q14b==8] <- NA

table(conf2.x)
conf2.labs <- c("Problem", "Not a problem")
conf2 <- factor(conf2.x, levels=c(1,0), labels=conf2.labs)
conf2.1 <- table(conf2)
conf.tab <- prop.table(conf2.1)
round(conf.tab, digits=2)


###Traditional way of life###
table(sa.only$q20)
life2.x <- numeric()
life2.x[sa.only$q20==1] <- 0
life2.x[sa.only$q20==2] <- 1
life2.x[sa.only$q20==8] <- NA
life2.x[sa.only$q20==9] <- NA

table(life2.x)
life2.labs <- c("Getting Lost", "Remains Strong")
life2 <- factor(life2.x, levels=c(0,1), labels=life2.labs)
life2.1 <- table(life2)
life.tab <- prop.table(life2.1)
round(life.tab, digits=2)


###S.A Employment levels###
table(sa.only$q118safr)
emp2.x <- numeric()
emp2.x [sa.only$q118safr==1] <- 1
emp2.x [sa.only$q118safr==2] <- 1
emp2.x [sa.only$q118safr==3] <- 1
emp2.x [sa.only$q118safr==4] <- 1
emp2.x [sa.only$q118safr==5] <- 0
emp2.x [sa.only$q118safr==6] <- 0
emp2.x [sa.only$q118safr==7] <- 0
emp2.x [sa.only$q118safr==8] <- 0
emp2.x [sa.only$q118safr==9] <- 0
emp2.x [sa.only$q118safr==99] <- NA
table(emp2.x)

emp2.labs  <- c("Low level of Employment", "High level of Employment")
emp2.1 <- factor(emp2.x, levels=0:1, labels=emp2.labs)

emp2.1.tab <- table(emp2.1)
emp2.tab <- prop.table(emp2.1.tab)
emp2.tab
barplot(emp2.tab, ylab="proportion")

###S.A Educations levels###
table(sa.only$q118safr)
edu2.x <- numeric()
edu2.x [sa.only$q118safr==1] <- 0
edu2.x [sa.only$q118safr==2] <- 0
edu2.x [sa.only$q118safr==3] <- 0
edu2.x [sa.only$q118safr==4] <- 0
edu2.x [sa.only$q118safr==5] <- 1
edu2.x [sa.only$q118safr==6] <- 1
edu2.x [sa.only$q118safr==98] <- NA
table(edu2.x)

edu2.labs  <- c("Low level of Education", "High level of Education")
edu2.1 <- factor(edu2.x, levels=0:1, labels=edu2.labs)

edu2.1.tab <- table(edu2.1)
edu2.tab <- prop.table(edu2.1.tab)
edu2.tab
barplot(edu2.tab, ylab="proportion")

###immigration###
ptab1.1.2x2 <- table(immi2, emp2.1)
ptab1.2x2 <- prop.table(ptab1.1.2x2)
ptab1.2x2
mosaicplot(t(ptab1.2x2), las=1,
           col=TRUE,
           main="",
           xlab="Employment",
           ylab="attitude toward immigration")

d <- ptab1.2x2[1,2] - ptab1.2x2[1,1]
round(d, digits=2)

ptab2.1.2x2 <- table(immi2, edu2.1)
ptab2.2x2 <- prop.table(ptab2.1.2x2)

mosaicplot(t(ptab2.2x2), las=1,
           col=TRUE,
           main="",
           xlab="Education",
           ylab="attitude toward immigration")

d <- ptab2.2x2[1,2] - ptab2.2x2[1,1]
round(d, digits=2)

###Immigrants###
ptab3.1.2x2 <- table(migr2, emp2.1)
ptab3.2x2 <- prop.table(ptab3.1.2x2)

mosaicplot(t(ptab3.2x2), las=1,
           col=TRUE,
           main="",
           xlab="Employment",
           ylab="attitude toward immigrants")

d <- ptab3.2x2[1,2] - ptab3.2x2[1,1]
round(d, digits=2)

ptab4.1.2x2 <- table(migr2, edu2.1)
ptab4.2x2 <- prop.table(ptab4.1.2x2)

mosaicplot(t(ptab4.2x2), las=1,
           col=TRUE,
           main="",
           xlab="Education",
           ylab="attitude toward immigrants")

d <- ptab4.2x2[1,2] - ptab4.2x2[1,1]
round(d, digits=2)


###Protection###
ptab5.1.2x2 <- table(prot2, emp2.1)
ptab5.2x2 <- prop.table(ptab5.1.2x2)

mosaicplot(t(ptab5.2x2), las=1,
           col=TRUE,
           main="",
           xlab="Employment",
           ylab="attitude toward foreign protection")

d <- ptab5.2x2[1,2] - ptab5.2x2[1,1]
round(d, digits=2)

ptab6.1.2x2 <- table(prot2, edu2.1)
ptab6.2x2 <- prop.table(ptab6.1.2x2)

mosaicplot(t(ptab6.2x2), las=1,
           col=TRUE,
           main="",
           xlab="Education",
           ylab="attitude toward foreign protection")

d <- ptab6.2x2[1,2] - ptab6.2x2[1,1]
round(d, digits=2)

###Restrictions###
ptab7.1.2x2 <- table(rest2, emp2.1)
ptab7.2x2 <- prop.table(ptab7.1.2x2)

mosaicplot(t(ptab7.2x2), las=1,
           col=TRUE,
           main="",
           xlab="Employment",
           ylab="attitude toward border restrictions")

d <- ptab7.2x2[1,2] - ptab7.2x2[1,1]
round(d, digits=2)

ptab8.1.2x2 <- table(rest2, edu2.1)
ptab8.2x2 <- prop.table(ptab8.1.2x2)

mosaicplot(t(ptab8.2x2), las=1,
           col=TRUE,
           main="",
           xlab="Education",
           ylab="attitude toward border restrictions")

d <- ptab8.2x2[1,2] - ptab8.2x2[1,1]
round(d, digits=2)

###Way of life###
ptab9.1.2x2 <- table(rest2, emp2.1)
ptab9.2x2 <- prop.table(ptab9.1.2x2)

mosaicplot(t(ptab9.2x2), las=1,
           col=TRUE,
           main="South Africa, 2007",
           xlab="Employment",
           ylab="attitude toward border restrictions")

d <- ptab9.2x2[1,2] - ptab9.2x2[1,1]
round(d, digits=2)

ptab10.1.2x2 <- table(rest2, edu2.1)
ptab10.2x2 <- prop.table(ptab10.1.2x2)

mosaicplot(t(ptab8.2x2), las=1,
           col=TRUE,
           main="South Africa, 2007",
           xlab="Education",
           ylab="attitude toward border restrictions")

d <- ptab10.2x2[1,2] - ptab10.2x2[1,1]
round(d, digits=2)

###Conflict###
ptab11.1.2x2 <- table(conf2, emp2.1)
ptab11.2x2 <- prop.table(ptab11.1.2x2)

mosaicplot(t(ptab11.2x2), las=1,
           col=TRUE,
           main="",
           xlab="Employment",
           ylab="attitude toward conflict")

d <- ptab11.2x2[1,2] - ptab11.2x2[1,1]
round(d, digits=2)

ptab12.1.2x2 <- table(conf2, edu2.1)
ptab12.2x2 <- prop.table(ptab12.1.2x2)

mosaicplot(t(ptab12.2x2), las=1,
           col=TRUE,
           main="",
           xlab="Education",
           ylab="attitude toward conflict ")

d <- ptab12.2x2[1,2] - ptab12.2x2[1,1]
round(d, digits=2)



######################
###Education levels###
######################
## Dichotomous approaches
#Every value >= cutpoint becomes 0
recode.abs <- function(x, cutpoint, missing=c(98, 99)) {
    new <- x
    new[new %in% missing] <- NA
    new[new < cutpoint] <- 0
    new[new >= cutpoint] <- 1
    return(new)
}

#Every value above median becomes 1
recode.rel <- function(x, missing=c(98,99)) {
    new <- x
    new[new %in% missing] <- NA
    med <- median(new, na.rm=TRUE)
    new[new < med] <- 1
    new[new >=  med] <- 0
    return(new)
}


newgap$edu.gen.a <- NA
newgap$edu.gen.a[newgap$country == "Argentina"] <- recode.abs(newgap$q118arg[newgap$country=="Argentina"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Bangladesh"] <- recode.abs(newgap$q118ban[newgap$country=="Bangladesh"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Bolivia"] <- recode.abs(newgap$q118bol[newgap$country=="Bolivia"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Brazil"] <- recode.abs(newgap$q118bra[newgap$country=="Brazil"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Bulgaria"] <- recode.abs(newgap$q118bul[newgap$country=="Bulgaria"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Canada"] <- recode.abs(newgap$q118can[newgap$country=="Canada"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Chile"] <- recode.abs(newgap$q118chil[newgap$country=="Chile"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "China"] <- recode.abs(newgap$q118chin[newgap$country=="China"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Czech Republic"] <- recode.abs(newgap$q118cze[newgap$country=="Czech Republic"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Egypt"] <- recode.abs(newgap$q118egy[newgap$country=="Egypt"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Ethiopia"] <- recode.abs(newgap$q118eth[newgap$country=="Ethiopia"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "France"] <- recode.abs(newgap$q118fra[newgap$country=="France"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Germany"] <- recode.abs(newgap$q118ger[newgap$country=="Germany"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Ghana"] <- recode.abs(newgap$q118gha[newgap$country=="Ghana"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Great Britain"] <- recode.abs(newgap$q118gbr[newgap$country=="Great Britain"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "India"] <- recode.abs(newgap$q118inda[newgap$country=="India"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Indonesia"] <- recode.abs(newgap$q118indo[newgap$country=="Indonesia"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Israel"] <- recode.abs(newgap$q118isr[newgap$country=="Israel"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Italy"] <- recode.abs(newgap$q118ita[newgap$country=="Italy"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Ivory Coast"] <- recode.abs(newgap$q118ivo[newgap$country=="Ivory Coast"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Japan"] <- recode.abs(newgap$q118jap[newgap$country=="Japan"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Jordan"] <- recode.abs(newgap$q118jor[newgap$country=="Jordan"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Kenya"] <- recode.abs(newgap$q118ken[newgap$country=="Kenya"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Kuwait"] <- recode.abs(newgap$q118kuw[newgap$country=="Kuwait"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Lebanon"] <- recode.abs(newgap$q118leb[newgap$country=="Lebanon"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Malaysia"] <- recode.abs(newgap$q118mala[newgap$country=="Malaysia"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Mali"] <- recode.abs(newgap$q118mali[newgap$country=="Mali"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Mexico"] <- recode.abs(newgap$q118mex[newgap$country=="Mexico"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Morocco"] <- recode.abs(newgap$q118mor[newgap$country=="Morocco"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Nigeria"] <- recode.abs(newgap$q118nig[newgap$country=="Nigeria"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Pakistan"] <- recode.abs(newgap$q118pak[newgap$country=="Pakistan"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Palestinian Territories"] <- recode.abs(newgap$q118pal[newgap$country=="Palestinian Territories"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Peru"] <- recode.abs(newgap$q118per[newgap$country=="Peru"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Poland"] <- recode.abs(newgap$q118pol[newgap$country=="Poland"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Russia"] <- recode.abs(newgap$q118rus[newgap$country=="Russia"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Senegal"] <- recode.abs(newgap$q118sen[newgap$country=="Senegal"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Slovakia"] <- recode.abs(newgap$q118slo[newgap$country=="Slovakia"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "South Africa"] <- recode.abs(newgap$q118safr[newgap$country=="South Africa"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "South Korea"] <- recode.abs(newgap$q118skor[newgap$country=="South Korea"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Spain"] <- recode.abs(newgap$q118spa[newgap$country=="Spain"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Sweden"] <- recode.abs(newgap$q118swe[newgap$country=="Sweden"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Tanzania"] <- recode.abs(newgap$q118tan[newgap$country=="Tanzania"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Turkey"] <- recode.abs(newgap$q118tur[newgap$country=="Turkey"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "United States"] <- recode.abs(newgap$q118us[newgap$country=="United States"], cutpoint=2, missing=9) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Uganda"] <- recode.abs(newgap$q118uga[newgap$country=="Uganda"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Ukraine"] <- recode.abs(newgap$q118ukr[newgap$country=="Ukraine"], cutpoint=4) #absolute cutpoint
newgap$edu.gen.a[newgap$country == "Venezuela"] <- recode.abs(newgap$q118ven[newgap$country=="Venezuela"], cutpoint=4) #absolute cutpoint

table(newgap$country, newgap$edu.gen.a)
edu1 <- tapply(newgap$edu.gen.a, newgap$country, mean,na.rm=TRUE)
dotchart(edu1[order(edu1)], color=c("red","blue","darkgreen"), main="Mean of absolute levels of education ", cex=0.8)
plot(density(edu1, na.rm=TRUE))

###Relative cutpoint###
newgap$edu.gen.r <- NA
newgap$edu.gen.r[newgap$country == "Argentina"] <- recode.rel(newgap$q118arg[newgap$country=="Argentina"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Bangladesh"] <- recode.rel(newgap$q118ban[newgap$country=="Bangladesh"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Bolivia"] <- recode.rel(newgap$q118bol[newgap$country=="Bolivia"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Brazil"] <- recode.rel(newgap$q118bra[newgap$country=="Brazil"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Bulgaria"] <- recode.rel(newgap$q118bul[newgap$country=="Bulgaria"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Canada"] <- recode.rel(newgap$q118can[newgap$country=="Canada"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Chile"] <- recode.rel(newgap$q118chil[newgap$country=="Chile"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "China"] <- recode.rel(newgap$q118chin[newgap$country=="China"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Czech Republic"] <- recode.rel(newgap$q118cze[newgap$country=="Czech Republic"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Egypt"] <- recode.rel(newgap$q118egy[newgap$country=="Egypt"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Ethiopia"] <- recode.rel(newgap$q118eth[newgap$country=="Ethiopia"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "France"] <- recode.rel(newgap$q118fra[newgap$country=="France"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Germany"] <- recode.rel(newgap$q118ger[newgap$country=="Germany"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Ghana"] <- recode.rel(newgap$q118gha[newgap$country=="Ghana"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Great Britain"] <- recode.rel(newgap$q118gbr[newgap$country=="Great Britain"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "India"] <- recode.rel(newgap$q118inda[newgap$country=="India"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Indonesia"] <- recode.rel(newgap$q118indo[newgap$country=="Indonesia"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Israel"] <- recode.rel(newgap$q118isr[newgap$country=="Israel"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Italy"] <- recode.rel(newgap$q118ita[newgap$country=="Italy"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Ivory Coast"] <- recode.rel(newgap$q118ivo[newgap$country=="Ivory Coast"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Japan"] <- recode.rel(newgap$q118jap[newgap$country=="Japan"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Jordan"] <- recode.rel(newgap$q118jor[newgap$country=="Jordan"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Kenya"] <- recode.rel(newgap$q118ken[newgap$country=="Kenya"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Kuwait"] <- recode.rel(newgap$q118kuw[newgap$country=="Kuwait"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Lebanon"] <- recode.rel(newgap$q118leb[newgap$country=="Lebanon"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Malaysia"] <- recode.rel(newgap$q118mala[newgap$country=="Malaysia"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Mali"] <- recode.rel(newgap$q118mali[newgap$country=="Mali"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Mexico"] <- recode.rel(newgap$q118mex[newgap$country=="Mexico"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Morocco"] <- recode.rel(newgap$q118mor[newgap$country=="Morocco"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Nigeria"] <- recode.rel(newgap$q118nig[newgap$country=="Nigeria"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Pakistan"] <- recode.rel(newgap$q118pak[newgap$country=="Pakistan"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Palestinian Territories"] <- recode.rel(newgap$q118pal[newgap$country=="Palestinian Territories"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Peru"] <- recode.rel(newgap$q118per[newgap$country=="Peru"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Poland"] <- recode.rel(newgap$q118pol[newgap$country=="Poland"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Russia"] <- recode.rel(newgap$q118rus[newgap$country=="Russia"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Senegal"] <- recode.rel(newgap$q118sen[newgap$country=="Senegal"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Slovakia"] <- recode.rel(newgap$q118slo[newgap$country=="Slovakia"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "South Africa"] <- recode.rel(newgap$q118safr[newgap$country=="South Africa"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "South Korea"] <- recode.rel(newgap$q118skor[newgap$country=="South Korea"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Spain"] <- recode.rel(newgap$q118spa[newgap$country=="Spain"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Sweden"] <- recode.rel(newgap$q118swe[newgap$country=="Sweden"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Tanzania"] <- recode.rel(newgap$q118tan[newgap$country=="Tanzania"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Turkey"] <- recode.rel(newgap$q118tur[newgap$country=="Turkey"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "United States"] <- recode.rel(newgap$q118us[newgap$country=="United States"],missing=9)  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Uganda"] <- recode.rel(newgap$q118uga[newgap$country=="Uganda"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Ukraine"] <- recode.rel(newgap$q118ukr[newgap$country=="Ukraine"])  #relative cutpoint
newgap$edu.gen.r[newgap$country == "Venezuela"] <- recode.rel(newgap$q118ven[newgap$country=="Venezuela"])  #relative cutpoint

table(newgap$country, newgap$edu.gen.r)
edu2 <- tapply(newgap$edu.gen.r, newgap$country, mean,na.rm=TRUE)
dotchart(edu2[order(edu2)], color=c("red","blue","darkgreen"), main="Mean of relative levels of education ", cex=0.8)





#######################
###Employment levles###
#######################

table(newgap$country,newgap$q120)
#Every value >1 cutpoint becomes 1
recode.abs2 <- function(x, cutpoint, missing=c(98, 99)) {
    new <- x
    new[new %in% missing] <- NA
    new[new <= cutpoint] <- 1
    new[new > cutpoint] <- 0
    return(new)
}

table(newgap$country, newgap$q120)
newgap$emp.gen.a <- NA
newgap$emp.gen.a[newgap$country == "Argentina"] <- recode.abs2(newgap$q120[newgap$country=="Argentina"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Bangladesh"] <- recode.abs2(newgap$q120[newgap$country=="Bangladesh"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Bolivia"] <- recode.abs2(newgap$q120[newgap$country=="Bolivia"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Brazil"] <- recode.abs2(newgap$q120[newgap$country=="Brazil"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Bulgaria"] <- recode.abs2(newgap$q120[newgap$country=="Bulgaria"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Canada"] <- recode.abs2(newgap$q120[newgap$country=="Canada"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Chile"] <- recode.abs2(newgap$q120[newgap$country=="Chile"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "China"] <- recode.abs2(newgap$q120[newgap$country=="China"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Czech Republic"] <- recode.abs2(newgap$q120[newgap$country=="Czech Republic"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Egypt"] <- recode.abs2(newgap$q120[newgap$country=="Egypt"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Ethiopia"] <- recode.abs2(newgap$q120[newgap$country=="Ethiopia"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "France"] <- recode.abs2(newgap$q120[newgap$country=="France"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Germany"] <- recode.abs2(newgap$q120[newgap$country=="Germany"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Ghana"] <- recode.abs2(newgap$q120[newgap$country=="Ghana"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Great Britain"] <- recode.abs2(newgap$q120[newgap$country=="Great Britain"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "India"] <- recode.abs2(newgap$q120inda[newgap$country=="India"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Indonesia"] <- recode.abs2(newgap$q120[newgap$country=="Indonesia"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Israel"] <- recode.abs2(newgap$q120[newgap$country=="Israel"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Italy"] <- recode.abs2(newgap$q120[newgap$country=="Italy"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Ivory Coast"] <- recode.abs2(newgap$q120[newgap$country=="Ivory Coast"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Japan"] <- recode.abs2(newgap$q120[newgap$country=="Japan"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Jordan"] <- recode.abs2(newgap$q120[newgap$country=="Jordan"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Kenya"] <- recode.abs2(newgap$q120[newgap$country=="Kenya"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Kuwait"] <- recode.abs2(newgap$q120[newgap$country=="Kuwait"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Lebanon"] <- recode.abs2(newgap$q120[newgap$country=="Lebanon"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Malaysia"] <- recode.abs2(newgap$q120[newgap$country=="Malaysia"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Mali"] <- recode.abs2(newgap$q120[newgap$country=="Mali"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Mexico"] <- recode.abs2(newgap$q120[newgap$country=="Mexico"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Morocco"] <- recode.abs2(newgap$q120[newgap$country=="Morocco"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Nigeria"] <- recode.abs2(newgap$q120[newgap$country=="Nigeria"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Pakistan"] <- recode.abs2(newgap$q120[newgap$country=="Pakistan"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Palestinian Territories"] <- recode.abs2(newgap$q120[newgap$country=="Palestinian Territories"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Peru"] <- recode.abs2(newgap$q120[newgap$country=="Peru"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Poland"] <- recode.abs2(newgap$q120[newgap$country=="Poland"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Russia"] <- recode.abs2(newgap$q120[newgap$country=="Russia"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Senegal"] <- recode.abs2(newgap$q120[newgap$country=="Senegal"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Slovakia"] <- recode.abs2(newgap$q120[newgap$country=="Slovakia"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "South Africa"] <- recode.abs2(newgap$q120[newgap$country=="South Africa"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "South Korea"] <- recode.abs2(newgap$q120[newgap$country=="South Korea"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Spain"] <- recode.abs2(newgap$q120spa[newgap$country=="Spain"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Sweden"] <- recode.abs2(newgap$q120[newgap$country=="Sweden"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Tanzania"] <- recode.abs2(newgap$q120[newgap$country=="Tanzania"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Turkey"] <- recode.abs2(newgap$q120[newgap$country=="Turkey"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "United States"] <- recode.abs2(newgap$q120us[newgap$country=="United States"], cutpoint=2, missing=c(8,9)) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Uganda"] <- recode.abs2(newgap$q120[newgap$country=="Uganda"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Ukraine"] <- recode.abs2(newgap$q120[newgap$country=="Ukraine"], cutpoint=2) #absolute cutpoint
newgap$emp.gen.a[newgap$country == "Venezuela"] <- recode.abs2(newgap$q120[newgap$country=="Venezuela"], cutpoint=2) #absolute cutpoint

table(newgap$country, newgap$emp.gen.a)
emp1 <- tapply(newgap$emp.gen.a, newgap$country, mean,na.rm=TRUE)

dotchart(emp1[order(emp1)], color=c("red","blue","darkgreen"), main="Mean of absolute levels of employment ", cex=0.8)
plot(density(emp1, na.rm=TRUE))










####EDU.R

###edu vs migr###
par(mfrow=c(2,3))
newgap$edu.gen.r[newgap$edu.gen.r==0] <- "Low education level"
newgap$edu.gen.r[newgap$edu.gen.r==1] <- "High education level"
table(newgap$edu.gen.r)

migr.x2.labs <- c("Good","Bad")
migr.x2 <- numeric()
migr.x2[newgap$q23f==1] <- 1
migr.x2[newgap$q23f==2] <- 1
migr.x2[newgap$q23f==3] <- 2
migr.x2[newgap$q23f==4] <- 2
migr.x2[newgap$q23f==8] <- NA
migr.x2[newgap$q23f==9] <- NA
table(newgap$country, migr.x2)

migr.di <- factor(migr.x2, levels=1:2, labels=migr.x2.labs)
table(migr.di)
migrvsedu <- table(migr.di, newgap$edu.gen.r)
prop.table(migrvsedu, margin=2)
migrvsedu.d <- prop.table(migrvsedu, margin=2)
d <- migrvsedu.d[1,1] - migrvsedu.d[1,2]
d
mosaicplot(t(migrvsedu.d), main="Relative Education", ylab="Influence of immigrants")

table(migr.di, newgap$edu.gen.r)
prop.test(x=c(11568, 6293), n=c(28789,16379))

###edu vs immi###
newgap$edu.gen.r[newgap$edu.gen.r==0] <- "Low education level"
newgap$edu.gen.r[newgap$edu.gen.r==1] <- "High education level"
table(newgap$edu.gen.r)

immi.x2.labs <- c("Problem","Not a Problem")
immi.x2 <- numeric()
immi.x2[newgap$q14h==1] <- 1
immi.x2[newgap$q14h==2] <- 1
immi.x2[newgap$q14h==3] <- 2
immi.x2[newgap$q14h==4] <- 2
immi.x2[newgap$q14h==8] <- NA
immi.x2[newgap$q14h==9] <- NA
table(newgap$country, immi.x2)

immi.di <- factor(immi.x2, levels=1:2, labels=immi.x2.labs)
table(immi.di)
immivsedu <- table(immi.di, newgap$edu.gen.r)
prop.table(immivsedu, margin=2)
immivsedu.d <- prop.table(immivsedu, margin=2)
d <- immivsedu.d[1,1] - immivsedu.d[1,2]
d
mosaicplot(t(immivsedu.d), main="Relative Education", ylab="Immigration is a problem")

table(immi.di, newgap$edu.gen.r)
prop.test(x=c(14148, 7784), n=c(24910,13531))

###edu vs prot###

newgap$edu.gen.r[newgap$edu.gen.r==0] <- "Low education level"
newgap$edu.gen.r[newgap$edu.gen.r==1] <- "High education level"
table(newgap$edu.gen.r)

prot.x2.labs <- c("Agree","Disagree")
prot.x2 <- numeric()
prot.x2[newgap$q24e==1] <- 1
prot.x2[newgap$q24e==2] <- 1
prot.x2[newgap$q24e==3] <- 2
prot.x2[newgap$q24e==4] <- 2
prot.x2[newgap$q24e==8] <- NA
prot.x2[newgap$q24e==9] <- NA
table(newgap$country, prot.x2)

prot.di <- factor(prot.x2, levels=1:2, labels=prot.x2.labs)
table(prot.di)
protvsedu <- table(prot.di, newgap$edu.gen.r)
prop.table(protvsedu, margin=2)
protvsedu.d <- prop.table(protvsedu, margin=2)
d <- protvsedu.d[1,1] - protvsedu.d[1,2]
d
mosaicplot(t(protvsedu.d), main="Relative Education", ylab="This country needs protection from foreign influence")

table(prot.di, newgap$edu.gen.r)
prop.test(x=c(19464, 11290), n=c(25320,13983))

###edu vs life###
newgap$edu.gen.r[newgap$edu.gen.r==0] <- "Low education level"
newgap$edu.gen.r[newgap$edu.gen.r==1] <- "High education level"
table(newgap$edu.gen.r)

life.x2.labs <- c("Getting Lost","Remains Strong")
life.x2 <- numeric()
life.x2[newgap$q20==1] <- 1
life.x2[newgap$q20==2] <- 2
life.x2[newgap$q20==8] <- NA
life.x2[newgap$q20==9] <- NA
table(newgap$country, life.x2)

life.di <- factor(life.x2, levels=1:2, labels=life.x2.labs)
table(life.di)
lifevsedu <- table(life.di, newgap$edu.gen.r)
prop.table(lifevsedu, margin=2)
lifevsedu.d <- prop.table(lifevsedu, margin=2)
d <- lifevsedu.d[1,1] - lifevsedu.d[1,2]
d
mosaicplot(t(lifevsedu.d), main="Relative Education", ylab="Traditional way of life is getting lost")

table(life.di, newgap$edu.gen.r)
prop.test(x=c(19446, 10570), n=c(25219,14226))

###edu vs rest###
newgap$edu.gen.r[newgap$edu.gen.r==0] <- "Low education level"
newgap$edu.gen.r[newgap$edu.gen.r==1] <- "High education level"
table(newgap$edu.gen.r)

rest.x2.labs <- c("Agree","Disagree")
rest.x2 <- numeric()
rest.x2[newgap$q24f==1] <- 1
rest.x2[newgap$q24f==2] <- 1
rest.x2[newgap$q24f==3] <- 2
rest.x2[newgap$q24f==4] <- 2
rest.x2[newgap$q24f==8] <- NA
rest.x2[newgap$q24f==9] <- NA
table(newgap$country, rest.x2)

rest.di <- factor(rest.x2, levels=1:2, labels=rest.x2.labs)
table(rest.di)
restvsedu <- table(rest.di, newgap$edu.gen.r)
prop.table(restvsedu, margin=2)
restvsedu.d <- prop.table(restvsedu, margin=2)
d <- restvsedu.d[1,1] - restvsedu.d[1,2]
d
mosaicplot(t(restvsedu.d), main="Relative Education", ylab="Country needs more restrictions and control of entry")

table(rest.di, newgap$edu.gen.r)
prop.test(x=c(18208, 10432), n=c(25143,18208))

###edu vs conf###
newgap$edu.gen.r[newgap$edu.gen.r==0] <- "Low education level"
newgap$edu.gen.r[newgap$edu.gen.r==1] <- "High education level"
table(newgap$edu.gen.r)

conf.x2.labs <- c("Problem", "Not a Problem")
conf.x2 <- numeric()
conf.x2[newgap$q14b==1] <- 1
conf.x2[newgap$q14b==2] <- 1
conf.x2[newgap$q14b==3] <- 2
conf.x2[newgap$q14b==4] <- 2
conf.x2[newgap$q14b==8] <- NA
conf.x2[newgap$q14b==9] <- NA
table(newgap$country, conf.x2)

conf.di <- factor(conf.x2, levels=1:2, labels=conf.x2.labs)
table(conf.di)
confvsedu <- table(conf.di, newgap$edu.gen.r)
prop.table(confvsedu, margin=2)
confvsedu.d <- prop.table(confvsedu, margin=2)
d <- confvsedu.d[1,1] - confvsedu.d[1,2]
d
mosaicplot(t(confvsedu.d), main="Relative Education", ylab="Conflict is not a problem")

table(conf.di, newgap$edu.gen.r)
prop.test(x=c(16668, 9032), n=c(23885,12711))
dev.off()

###EDU.A###
par(mfrow=c(2,3))
newgap$edu.gen.a[newgap$edu.gen.a==0] <- "Low education level"
newgap$edu.gen.a[newgap$edu.gen.a==1] <- "High education level"
table(newgap$edu.gen.a)

migr.x2.labs <- c("Good","Bad")
migr.x2 <- numeric()
migr.x2[newgap$q23f==1] <- 1
migr.x2[newgap$q23f==2] <- 1
migr.x2[newgap$q23f==3] <- 2
migr.x2[newgap$q23f==4] <- 2
migr.x2[newgap$q23f==8] <- NA
migr.x2[newgap$q23f==9] <- NA
table(newgap$country, migr.x2)

migr.di <- factor(migr.x2, levels=1:2, labels=migr.x2.labs)
table(migr.di)
migrvsedu <- table(migr.di, newgap$edu.gen.a)
prop.table(migrvsedu, margin=2)
migrvsedu.d <- prop.table(migrvsedu, margin=2)
d <- migrvsedu.d[1,1] - migrvsedu.d[1,2]
d
mosaicplot(t(migrvsedu.d), main="Absolute Education", ylab="Immigrants are a bad influence")


###edu vs immi###
newgap$edu.gen.a[newgap$edu.gen.a==0] <- "Low education level"
newgap$edu.gen.a[newgap$edu.gen.a==1] <- "High education level"
table(newgap$edu.gen.a)

immi.x2.labs <- c("Problem","Not a Problem")
immi.x2 <- numeric()
immi.x2[newgap$q14h==1] <- 1
immi.x2[newgap$q14h==2] <- 1
immi.x2[newgap$q14h==3] <- 2
immi.x2[newgap$q14h==4] <- 2
immi.x2[newgap$q14h==8] <- NA
immi.x2[newgap$q14h==9] <- NA
table(newgap$country, immi.x2)

immi.di <- factor(immi.x2, levels=1:2, labels=immi.x2.labs)
table(immi.di)
immivsedu <- table(immi.di, newgap$edu.gen.a)
prop.table(immivsedu, margin=2)
immivsedu.d <- prop.table(immivsedu, margin=2)
d <- immivsedu.d[1,1] - immivsedu.d[1,2]
d
mosaicplot(t(immivsedu.d), main="Absolute Education", ylab="Immigration is a problem")


###edu vs prot###

newgap$edu.gen.a[newgap$edu.gen.a==0] <- "Low education level"
newgap$edu.gen.a[newgap$edu.gen.a==1] <- "High education level"
table(newgap$edu.gen.a)

prot.x2.labs <- c("Agree","Disagree")
prot.x2 <- numeric()
prot.x2[newgap$q24e==1] <- 1
prot.x2[newgap$q24e==2] <- 1
prot.x2[newgap$q24e==3] <- 2
prot.x2[newgap$q24e==4] <- 2
prot.x2[newgap$q24e==8] <- NA
prot.x2[newgap$q24e==9] <- NA
table(newgap$country, prot.x2)

prot.di <- factor(prot.x2, levels=1:2, labels=prot.x2.labs)
table(prot.di)
protvsedu <- table(prot.di, newgap$edu.gen.a)
prop.table(protvsedu, margin=2)
protvsedu.d <- prop.table(protvsedu, margin=2)
d <- protvsedu.d[1,1] - protvsedu.d[1,2]
d
mosaicplot(t(protvsedu.d), main="Absolute Education", ylab="This country needs protection from foreign influence")


###edu vs life###
newgap$edu.gen.a[newgap$edu.gen.a==0] <- "Low education level"
newgap$edu.gen.a[newgap$edu.gen.a==1] <- "High education level"
table(newgap$edu.gen.a)

life.x2.labs <- c("Getting Lost","Remains Strong")
life.x2 <- numeric()
life.x2[newgap$q20==1] <- 1
life.x2[newgap$q20==2] <- 2
life.x2[newgap$q20==8] <- NA
life.x2[newgap$q20==9] <- NA
table(newgap$country, life.x2)

life.di <- factor(life.x2, levels=1:2, labels=life.x2.labs)
table(life.di)
lifevsedu <- table(life.di, newgap$edu.gen.a)
prop.table(lifevsedu, margin=2)
lifevsedu.d <- prop.table(lifevsedu, margin=2)
d <- lifevsedu.d[1,1] - lifevsedu.d[1,2]
d
mosaicplot(t(lifevsedu.d), main="Absolute Education", ylab="Traditional way of life is getting lost")


###edu vs rest###
newgap$edu.gen.a[newgap$edu.gen.a==0] <- "Low education level"
newgap$edu.gen.a[newgap$edu.gen.a==1] <- "High education level"
table(newgap$edu.gen.a)

rest.x2.labs <- c("Agree","Disagree")
rest.x2 <- numeric()
rest.x2[newgap$q24f==1] <- 1
rest.x2[newgap$q24f==2] <- 1
rest.x2[newgap$q24f==3] <- 2
rest.x2[newgap$q24f==4] <- 2
rest.x2[newgap$q24f==8] <- NA
rest.x2[newgap$q24f==9] <- NA
table(newgap$country, rest.x2)

rest.di <- factor(rest.x2, levels=1:2, labels=rest.x2.labs)
table(rest.di)
restvsedu <- table(rest.di, newgap$edu.gen.a)
prop.table(restvsedu, margin=2)
restvsedu.d <- prop.table(restvsedu, margin=2)
d <- restvsedu.d[1,1] - restvsedu.d[1,2]
d
mosaicplot(t(restvsedu.d),main="Absolute Education", ylab="Country needs more restrictions and control of entry")


###edu vs conf###
newgap$edu.gen.a[newgap$edu.gen.a==0] <- "Low education level"
newgap$edu.gen.a[newgap$edu.gen.a==1] <- "High education level"
table(newgap$edu.gen.a)

conf.x2.labs <- c("Problem", "Not a Problem")
conf.x2 <- numeric()
conf.x2[newgap$q14b==1] <- 1
conf.x2[newgap$q14b==2] <- 1
conf.x2[newgap$q14b==3] <- 2
conf.x2[newgap$q14b==4] <- 2
conf.x2[newgap$q14b==8] <- NA
conf.x2[newgap$q14b==9] <- NA
table(newgap$country, conf.x2)

conf.di <- factor(conf.x2, levels=1:2, labels=conf.x2.labs)
table(conf.di)
confvsedu <- table(conf.di, newgap$edu.gen.a)
prop.table(confvsedu, margin=2)
confvsedu.d <- prop.table(confvsedu, margin=2)
d <- confvsedu.d[1,1] - confvsedu.d[1,2]
d
mosaicplot(t(confvsedu.d),main="Absolute Education", ylab="Conflict is not a problem")
dev.off()


####EMP###


###emp vs migr ###
par(mfrow=c(2,3))
newgap$emp.gen.a[newgap$emp.gen.a==0] <- "Unemployed"
newgap$emp.gen.a[newgap$emp.gen.a==1] <- "Employed"
table(newgap$emp.gen.a)

migr.x2.labs <- c("Good","Bad")
migr.x2 <- numeric()
migr.x2[newgap$q23f==1] <- 1
migr.x2[newgap$q23f==2] <- 1
migr.x2[newgap$q23f==3] <- 2
migr.x2[newgap$q23f==4] <- 2
migr.x2[newgap$q23f==8] <- NA
migr.x2[newgap$q23f==9] <- NA
table(newgap$country, migr.x2)

migr.di <- factor(migr.x2, levels=1:2, labels=migr.x2.labs)
table(migr.di)
migrvsemp <- table(migr.di, newgap$emp.gen.a)
prop.table(migrvsemp, margin=2)
migrvsemp.d <- prop.table(migrvsemp, margin=2)
d <- migrvsemp.d[1,1] - migrvsemp.d[1,2]
d
mosaicplot(t(migrvsemp.d), main="Employment", ylab="Immigrants are a bad influence in this country")

table(migr.di, newgap$emp.gen.a)
prop.test(x=c(9607, 6866), n=c(18795,13630))

###emp vs immi ###
newgap$emp.gen.a[newgap$emp.gen.a==0] <- "Unemployed"
newgap$emp.gen.a[newgap$emp.gen.a==1] <- "Employed"
table(newgap$emp.gen.a)

immi.x2.labs <- c("Problem","Not a Problem")
immi.x2 <- numeric()
immi.x2[newgap$q14h==1] <- 1
immi.x2[newgap$q14h==2] <- 1
immi.x2[newgap$q14h==3] <- 2
immi.x2[newgap$q14h==4] <- 2
immi.x2[newgap$q14h==8] <- NA
immi.x2[newgap$q14h==9] <- NA
table(newgap$country, immi.x2)

immi.di <- factor(immi.x2, levels=1:2, labels=immi.x2.labs)
table(immi.di)
immivsemp <- table(immi.di, newgap$emp.gen.a)
prop.table(immivsemp, margin=2)
immivsemp.d <- prop.table(immivsemp, margin=2)
d <- immivsemp.d[1,1] - immivsemp.d[1,2]
d
mosaicplot(t(immivsemp.d),main="Employment",ylab="Immigration is a problem in this country")

table(immi.di, newgap$emp.gen.a)
prop.test(x=c(12325, 8772), n=c(20735,14677))

###emp vs prot ###
newgap$emp.gen.a[newgap$emp.gen.a==0] <- "Unemployed"
newgap$emp.gen.a[newgap$emp.gen.a==1] <- "Employed"
table(newgap$emp.gen.a)

prot.x2.labs <- c("Agree","Disagree")
prot.x2 <- numeric()
prot.x2[newgap$q24e==1] <- 1
prot.x2[newgap$q24e==2] <- 1
prot.x2[newgap$q24e==3] <- 2
prot.x2[newgap$q24e==4] <- 2
prot.x2[newgap$q24e==8] <- NA
prot.x2[newgap$q24e==9] <- NA
table(newgap$country, prot.x2)

prot.di <- factor(prot.x2, levels=1:2, labels=prot.x2.labs)
table(prot.di)
protvsemp <- table(prot.di, newgap$emp.gen.a)
prop.table(protvsemp, margin=2)
protvsemp.d <- prop.table(protvsemp, margin=2)
d <- protvsemp.d[1,1] - protvsemp.d[1,2]
d
mosaicplot(t(protvsemp.d),main="Emplyment",ylab="Protection from foreign influence")

table(prot.di, newgap$emp.gen.a)
prop.test(x=c(17089, 11262), n=c(21266,14813))

###emp vs rest ###
newgap$emp.gen.a[newgap$emp.gen.a==0] <- "Unemployed"
newgap$emp.gen.a[newgap$emp.gen.a==1] <- "Employed"
table(newgap$emp.gen.a)

rest.x2.labs <- c("Agree","Disagree")
rest.x2 <- numeric()
rest.x2[newgap$q24f==1] <- 1
rest.x2[newgap$q24f==2] <- 1
rest.x2[newgap$q24f==3] <- 2
rest.x2[newgap$q24f==4] <- 2
rest.x2[newgap$q24f==8] <- NA
rest.x2[newgap$q24f==9] <- NA
table(newgap$country, rest.x2)

rest.di <- factor(rest.x2, levels=1:2, labels=rest.x2.labs)
table(rest.di)
restvsemp <- table(rest.di, newgap$emp.gen.a)
prop.table(restvsemp, margin=2)
restvsemp.d <- prop.table(restvsemp, margin=2)
d <- restvsemp.d[1,1] - restvsemp.d[1,2]
d
mosaicplot(t(restvsemp.d),main="Employment",ylab="Country needs more restrictions")

table(rest.di, newgap$emp.gen.a)
prop.test(x=c(16054, 10784), n=c(21039,14774))

###emp vs life ###
newgap$emp.gen.a[newgap$emp.gen.a==0] <- "Unemployed"
newgap$emp.gen.a[newgap$emp.gen.a==1] <- "Employed"
table(newgap$emp.gen.a)

life.x2.labs <- c("Getting Lost", "Remains Strong")
life.x2 <- numeric()
life.x2[newgap$q20==1] <- 1
life.x2[newgap$q20==2] <- 2
life.x2[newgap$q20==8] <- NA
life.x2[newgap$q20==9] <- NA
table(newgap$country, life.x2)

life.di <- factor(life.x2, levels=1:2, labels=life.x2.labs)
table(life.di)
lifevsemp <- table(life.di, newgap$emp.gen.a)
prop.table(lifevsemp, margin=2)
lifevsemp.d <- prop.table(lifevsemp, margin=2)
d <- lifevsemp.d[1,1] - lifevsemp.d[1,2]
d
mosaicplot(t(lifevsemp.d),main="Employment",ylab="Traditional way of life is changing")
prop.test(lifevsemp)

###emp vs conf ###
newgap$emp.gen.a[newgap$emp.gen.a==0] <- "Unemployed"
newgap$emp.gen.a[newgap$emp.gen.a==1] <- "Employed"
table(newgap$emp.gen.a)

conf.x2.labs <- c("Problem", "Not a Problem")
conf.x2 <- numeric()
conf.x2[newgap$q14b==1] <- 1
conf.x2[newgap$q14b==2] <- 1
conf.x2[newgap$q14b==3] <- 2
conf.x2[newgap$q14b==4] <- 2
conf.x2[newgap$q14b==8] <- NA
conf.x2[newgap$q14b==9] <- NA
table(newgap$country, conf.x2)

conf.di <- factor(conf.x2, levels=1:2, labels=conf.x2.labs)
table(conf.di)
confvsemp <- table(conf.di, newgap$emp.gen.a)
prop.table(confvsemp, margin=2)
confvsemp.d <- prop.table(confvsemp, margin=2)
d <- confvsemp.d[1,1] - confvsemp.d[1,2]
d
mosaicplot(t(confvsemp.d),main="Employment",ylab="conflict is not a problem")
prop.test(confvsemp)
dev.off()









##########################

## PROP TEST FUNCTION WITH CONFIDENCE INTERVAL, FOR MANY SAMPLESa
## x must be a dichotomous ("outcome") vector
## group must be a dichotomous ("grouping") vector
## by must be a vector identifying the "samples" (e.g., countries)
propDandCI <- function(x, group, by, conf.level=0.95) {
  df <- na.omit(data.frame(x=x, group=group, by=by))
  df1 <- df[as.numeric(as.factor(df$group))==1, ]
  df2 <- df[as.numeric(as.factor(df$group))==2, ]
  dtable1 <- addmargins(table(df1$by, df1$x), margin=2)
  dtable2 <- addmargins(table(df2$by, df2$x), margin=2)
  dtable <- cbind(dtable1, dtable2)
  gnames <- rownames(dtable)
  get.d <- function(x) {
    pt.out <- prop.test(x=x[c(2,5)], n=x[c(3,6)], conf.level=conf.level)
    return(diff(pt.out$estimate))
  }
  get.ciLow <- function(x) {
    pt.out <- prop.test(x=x[c(2,5)], n=x[c(3,6)], conf.level=conf.level)
    return(-pt.out$conf.int[2])
  }
  get.ciHigh <- function(x) {
    pt.out <- prop.test(x=x[c(2,5)], n=x[c(3,6)], conf.level=conf.level)
    return(-pt.out$conf.int[1])
  }
  d <- apply(dtable, 1, get.d)
  ciLow <- apply(dtable, 1, get.ciLow)
  ciHigh <- apply(dtable, 1, get.ciHigh)
  output <- data.frame(by=gnames, d=d, ciLow=ciLow, ciHigh=ciHigh)
  return(output)
}


#Difference of proportions analysis

#Put existing dichotomous variable on immigration into dataframe
newgap$immi.di <- immi.di

#"Relative" education is already in the data frame
#Use new function to run the table
dCompTab1 <- propDandCI(x=newgap$immi.di,
                       group=newgap$edu.gen.r,
                       by=newgap$country,
                       conf.level=.9)
round(dCompTab1[ , -1], 3)
dCompTab1 <- dCompTab1[order(dCompTab1$d), ]

png("newgraph.png", width=480, height=720)
dotchart(dCompTab1$d, labels=dCompTab1$by,
         xlim=c(min(dCompTab1$ciLow), max(dCompTab1$ciHigh)),
         xlab="proportion difference in anti-immigrant sentiment\np(low educ) - p(high educ)",
         pch=16)
abline(v=0)
segments(x0=dCompTab1$ciLow,
         x1=dCompTab1$ciHigh,
         y0=1:nrow(dCompTab1),
         y1=1:nrow(dCompTab1))
dev.off()

###Immigration
newgap$migr.di <- migr.di
dCompTab2 <- propDandCI(x=newgap$migr.di,
                       group=newgap$edu.gen.r,
                       by=newgap$country,
                       conf.level=.9)
round(dCompTab2[ , -1], 3)
dCompTab2 <- dCompTab2[order(dCompTab2$d), ]

png("newgraph2.png", width=480, height=720)
dotchart(dCompTab2$d, labels=dCompTab2$by,
         xlim=c(min(dCompTab2$ciLow), max(dCompTab2$ciHigh)),
         xlab="proportion difference in anti-immigration sentiment\np(low educ) - p(high educ)",
         pch=16)
abline(v=0)
segments(x0=dCompTab2$ciLow,
         x1=dCompTab2$ciHigh,
         y0=1:nrow(dCompTab2),
         y1=1:nrow(dCompTab2))
dev.off()


###Restriction
newgap$rest.di <- rest.di
dCompTab3 <- propDandCI(x=newgap$rest.di,
                       group=newgap$edu.gen.r,
                       by=newgap$country,
                       conf.level=.9)
round(dCompTab3[ , -1], 3)
dCompTab3 <- dCompTab3[order(dCompTab3$d), ]

png("newgraph3.png", width=480, height=720)
dotchart(dCompTab3$d, labels=dCompTab3$by,
         xlim=c(min(dCompTab3$ciLow), max(dCompTab3$ciHigh)),
         xlab="proportion difference in restriction of entry\np(low educ) - p(high educ)",
         pch=16)
abline(v=0)
segments(x0=dCompTab3$ciLow,
         x1=dCompTab3$ciHigh,
         y0=1:nrow(dCompTab3),
         y1=1:nrow(dCompTab3))
dev.off()


###Protection
newgap$prot.di <- prot.di
dCompTab4 <- propDandCI(x=newgap$prot.di,
                       group=newgap$edu.gen.r,
                       by=newgap$country,
                       conf.level=.9)
round(dCompTab4[ , -1], 3)
dCompTab4 <- dCompTab4[order(dCompTab4$d), ]

png("newgraph4.png", width=480, height=720)
dotchart(dCompTab4$d, labels=dCompTab4$by,
         xlim=c(min(dCompTab4$ciLow), max(dCompTab4$ciHigh)),
         xlab="proportion difference in protection from foreign influence\np(low educ) - p(high educ)",
         pch=16)
abline(v=0)
segments(x0=dCompTab4$ciLow,
         x1=dCompTab4$ciHigh,
         y0=1:nrow(dCompTab4),
         y1=1:nrow(dCompTab4))
dev.off()

###Way of Life
newgap$life.di <- life.di
dCompTab5 <- propDandCI(x=newgap$life.di,
                       group=newgap$edu.gen.r,
                       by=newgap$country,
                       conf.level=.9)
round(dCompTab5[ , -1], 3)
dCompTab5 <- dCompTab5[order(dCompTab5$d), ]

png("newgraph5.png", width=480, height=720)
dotchart(dCompTab5$d, labels=dCompTab5$by,
         xlim=c(min(dCompTab5$ciLow), max(dCompTab5$ciHigh)),
         xlab="proportion difference in attitude towards way of life remaining strong\np(low educ) - p(high educ)",
         pch=16)
abline(v=0)
segments(x0=dCompTab5$ciLow,
         x1=dCompTab5$ciHigh,
         y0=1:nrow(dCompTab5),
         y1=1:nrow(dCompTab5))
dev.off()

###Conflict
newgap$conf.di <- conf.di
dCompTab6 <- propDandCI(x=newgap$conf.di,
                       group=newgap$edu.gen.r,
                       by=newgap$country,
                       conf.level=.9)
round(dCompTab6[ , -1], 3)
dCompTab6 <- dCompTab6[order(dCompTab6$d), ]

png("newgraph6.png", width=480, height=720)
dotchart(dCompTab6$d, labels=dCompTab6$by,
         xlim=c(min(dCompTab6$ciLow), max(dCompTab6$ciHigh)),
         xlab="proportion difference in attitude towards conflict \np(low educ) - p(high educ)",
         pch=16)
abline(v=0)
segments(x0=dCompTab6$ciLow,
         x1=dCompTab6$ciHigh,
         y0=1:nrow(dCompTab6),
         y1=1:nrow(dCompTab6))
dev.off()


###Employment
nochina <- newgap[newgap$country != "China", ]
nochina$country <- factor(nochina$country)

#####

dCompTab7 <- propDandCI(x=nochina$immi.di,
                       group=nochina$emp.gen.a,
                       by=nochina$country,
                       conf.level=.9)

round(dCompTab7[ , -1], 3)
dCompTab7 <- dCompTab7[order(dCompTab7$d), ]

png("newgraph7.png", width=480, height=720)
dotchart(dCompTab7$d, labels=dCompTab7$by,
         xlim=c(min(dCompTab7$ciLow), max(dCompTab7$ciHigh)),
         xlab="proportion difference in anti-immigrant sentiment\np(low emp) - p(high emp)",
         pch=16)
abline(v=0)
segments(x0=dCompTab7$ciLow,
         x1=dCompTab7$ciHigh,
         y0=1:nrow(dCompTab7),
         y1=1:nrow(dCompTab7))
dev.off()


###Immigration#####################################
dCompTab8 <- propDandCI(x=nochina$migr.di,
                       group=nochina$emp.gen.a,
                       by=nochina$country,
                       conf.level=.9)

round(dCompTab8[ , -1], 3)
dCompTab8 <- dCompTab8[order(dCompTab8$d), ]

png("newgraph8.png", width=480, height=720)
dotchart(dCompTab8$d, labels=dCompTab8$by,
         xlim=c(min(dCompTab8$ciLow), max(dCompTab8$ciHigh)),
         xlab="proportion difference in anti-immigration sentiment\np(low emp) - p(high emp)",
         pch=16)
abline(v=0)
segments(x0=dCompTab8$ciLow,
         x1=dCompTab8$ciHigh,
         y0=1:nrow(dCompTab8),
         y1=1:nrow(dCompTab8))
dev.off()


###Restriction
dCompTab9 <- propDandCI(x=nochina$rest.di,
                       group=nochina$emp.gen.a,
                       by=nochina$country,
                       conf.level=.9)
round(dCompTab9[ , -1], 3)
dCompTab9 <- dCompTab9[order(dCompTab9$d), ]

png("newgraph9.png", width=480, height=720)
dotchart(dCompTab9$d, labels=dCompTab9$by,
         xlim=c(min(dCompTab9$ciLow), max(dCompTab9$ciHigh)),
         xlab="proportion difference in restriction of entry\np(low emp) - p(high emp)",
         pch=16)
abline(v=0)
segments(x0=dCompTab9$ciLow,
         x1=dCompTab9$ciHigh,
         y0=1:nrow(dCompTab9),
         y1=1:nrow(dCompTab9))
dev.off()


###Protection
dCompTab10 <- propDandCI(x=nochina$prot.di,
                       group=nochina$emp.gen.a,
                       by=nochina$country,
                       conf.level=.9)
round(dCompTab10[ , -1], 3)
dCompTab10 <- dCompTab10[order(dCompTab10$d), ]

png("newgraph10.png", width=500, height=800)
dotchart(dCompTab10$d, labels=dCompTab10$by,
         xlim=c(min(dCompTab10$ciLow), max(dCompTab10$ciHigh)),
         xlab="proportion difference in protection\np(low emp) - p(high emp)",
         pch=16)
abline(v=0)
segments(x0=dCompTab10$ciLow,
         x1=dCompTab10$ciHigh,
         y0=1:nrow(dCompTab10),
         y1=1:nrow(dCompTab10))
dev.off()

###Way of Life##########################################
dCompTab11 <- propDandCI(x=nochina$life.di,
                       group=nochina$emp.gen.a,
                       by=nochina$country,
                       conf.level=.9)
round(dCompTab11[ , -1], 3)
dCompTab11 <- dCompTab11[order(dCompTab11$d), ]

png("newgraph11.png", width=480, height=720)
dotchart(dCompTab11$d, labels=dCompTab11$by,
         xlim=c(min(dCompTab11$ciLow), max(dCompTab11$ciHigh)),
         xlab="proportion difference in attitude towards way of life remaining strong\np(low emp) - p(high emp)",
         pch=16)
abline(v=0)
segments(x0=dCompTab11$ciLow,
         x1=dCompTab11$ciHigh,
         y0=1:nrow(dCompTab11),
         y1=1:nrow(dCompTab11))
dev.off()

###Conflict################################3

dCompTab12 <- propDandCI(x=nochina$conf.di,
                       group=nochina$emp.gen.a,
                       by=nochina$country,
                       conf.level=.9)
round(dCompTab12[ , -1], 3)
dCompTab12 <- dCompTab12[order(dCompTab12$d), ]

png("newgraph12.png", width=480, height=720)
dotchart(dCompTab12$d, labels=dCompTab12$by,
         xlim=c(min(dCompTab12$ciLow), max(dCompTab12$ciHigh)),
         xlab="proportion difference in attitude towards conflict \np(low emp) - p(high emp)",
         pch=16)
abline(v=0)
segments(x0=dCompTab12$ciLow,
         x1=dCompTab12$ciHigh,
         y0=1:nrow(dCompTab12),
         y1=1:nrow(dCompTab12))
dev.off()


####################

###Stastistics Assignment###

###Load a foreign spss data - Country specific (South Africa) Round 4 Afrobarometer data)

library(foreign)
data <- read.spss(file="saf.sav",to.data.frame=FALSE,
                        use.value.labels=TRUE)

objects()
names(data)
summary(data)
###Chang names to lower case###
names(data) <- tolower(names(data))
names(data)

###Subseting Columns###
sa <- data[c("q94", "q89", "q1", "q84d_saf", "q76a_saf",  "q76b_saf",  "q57t_saf", "q77c_saf", "q77a_saf", "q101", "q102", "region", "urbrur")]
summary(sa)



###Recoding data###

#########################################
###Trust towards immigrants/foreigners###
#########################################

sa$q84d_saf <- sa$q84d
uncoded <- table(sa$q84d_saf)
uncoded
barplot((uncoded), main="Trust towards immigrants")


atti.x <- numeric() ### Initialize vector
atti.x[sa$q84d== "Not at all"] <- 1
atti.x[sa$q84d== "Just a little"] <- 0.75
atti.x[sa$q84d== "Somewhat"] <- 0.5
atti.x[sa$q84d== "A lot"] <- 0.25
atti.x[sa$q84d== "Don't know"] <- 0
atti.x[sa$q84d== "Refused"] <- NA
atti.x[sa$q84d== "Missing"] <-NA
table(atti.x)

###Make labels###
atti.labs <- c("Not at all","Just a little", "Somewhat", "A lot", "Don't know")
atti.labs
###Make factor###
q84 <- factor(atti.x, levels=c(1,0.75,0.5,0.25,0), labels=atti.labs)

###Check the coding###
table(q84, sa$q84d, exclude=NULL)

###Add to the data frame###
sa <- data.frame(sa, q84)

###Graphing###
coded <- table(q84)

coded <- prop.table(coded)
barplot((coded), main="Trust towards immigrants", ylab="Proportions")

###Dichotomous###
atti.x2 <- numeric() ### Initialize vector
atti.x2[sa$q84d== "Not at all"] <- 1
atti.x2[sa$q84d== "Just a little"] <- 1
atti.x2[sa$q84d== "Somewhat"] <- 0
atti.x2[sa$q84d== "A lot"] <- 0
atti.x2[sa$q84d== "Don't know"] <- NA
atti.x2[sa$q84d== "Refused"] <- NA
atti.x2[sa$q84d== "Missing"] <-NA
table(atti.x2)

###Make labels###
atti.labs2 <- c("No","Yes")
atti.labs2
###Make factor###
atti <- factor(atti.x2, levels=c(1,0), labels=atti.labs2)
table(atti)

###################################
###Tolerance towards immigration###
###################################

sa$q76a_saf <- sa$q76a
uncodedtol <- table(sa$q76a)
uncodedtol
barplot((uncodedtol), main="Tolerance towards immigration")


tol.x <- numeric() ### Initialize vector
tol.x[sa$q76a== "Prohibit people entering from other countries"]  <- 1
tol.x[sa$q76a== "Place strict restrictions on the number of foreigners who ca"] <- 0.75
tol.x[sa$q76a== "Let people in as long as there are jobs available"]  <- 0.5
tol.x[sa$q76a== "Let anyone in who wants to enter"] <- 0.25
tol.x[sa$q76a== "Don't know"] <- 0
tol.x[sa$q76a== "Refused to answer"] <- NA
tol.x[sa$q76a== "Missing"]  <- NA
tol.x[sa$q76a== "Other"] <- NA
tol.x[sa$q76a== "Refused"] <- NA
table(tol.x)


###Make labels###
tol.labs <- c("Prohibit people", "Place strict restrictions", "Let people in if there are jobs", "Let anyone in", "Don't know")

###Make factor###
q76a <- factor(tol.x, levels=c(1,0.75,0.5,0.25,0), labels=tol.labs)

###Check coding###
table(q76a, sa$q76a, exclude=NULL)

###Add to data frame###
sa <- data.frame(sa, q76a)

###Graphing###
codedtol <- table(q76a)
codedtol <- prop.table(codedtol)
barplot((codedtol), main="Tolerance towards immigration", ylab="Proportions")

###Dichotomous###
tol.x2 <- numeric() ### Initialize vector
tol.x2[sa$q76a== "Prohibit people"]  <- 1
tol.x2[sa$q76a== "Place strict restrictions"] <- 1
tol.x2[sa$q76a== "Let people in as long as there are jobs"]  <- 0
tol.x2[sa$q76a== "Let anyone in"] <- 0
tol.x2[sa$q76a== "Don't know"] <- NA
tol.x2[sa$q76a== "Refused to answer"] <- NA
tol.x2[sa$q76a== "Missing data"]  <- NA
table(tol.x2)

###Make labels###
tol.labs2 <- c("Prohibit people",  "Let people in")

###Make factor###
tol <- factor(tol.x2, levels=c(1, 0), labels=tol.labs2)
table(tol)

###################################
###Tolerance towards deportation###
###################################

sa$q76b_saf <- sa$q76b
uncodeddep <- table(sa$q76b)
uncodeddep
barplot((uncodeddep), main="Tolerance towards deportation")


dep.x <- numeric() ### Initialize vector
dep.x[sa$q76b== "All of these people"] <- 1
dep.x[sa$q76b== "Only those who are not contributing to the economy"] <- 0.75
dep.x[sa$q76b== "Only those who are here without the permission of the SAF"] <- 0.5
dep.x[sa$q76b== "Only those who have committed crimes"] <- 0.25
dep.x[sa$q76b== "The govt. should not send back people to their own countries"] <- 0
dep.x[sa$q76b== "Other"] <- 0
dep.x[sa$q76b== "Don't know"] <- NA
dep.x[sa$q76b== "Missing"]  <- NA
table(dep.x)

table(sa$q76b, dep.x, exclude=NULL)
###Make labels -- in the right order###
dep.labs <- c("Every foreigner", " Those not contributing to the economy", " Those without permits", " Criminal offenders", " None")

###Make factor###
q76b <- factor(dep.x, levels=c(1,0.75,0.5,0.25,0), labels=dep.labs)

###Check coding###
table(q76b, sa$q76b, exclude=NULL)

###Add to data frame)
sa <- data.frame(sa, q76b)

###Graphing###

codeddep <- table(q76b)
codeddep <- prop.table(codeddep)
barplot((codeddep), main="Tolerance towards deportation", ylab="Proportions")



#################################################################################
###Attitudes towards the management of immigration by the Mbeki administration###
#################################################################################

sa$q57t_saf <- sa$q57t
uncodedman <- table(sa$q57t)
uncodedman
barplot((uncodedman), main="Management of immigration by Mbeki admin")



man.x <- numeric() ### Initialize vector
man.x[sa$q57t== "Very Badly"] <- 1
man.x[sa$q57t== "Fairly Badly"] <- 0.75
man.x[sa$q57t== "Fairly Well"] <- 0.5
man.x[sa$q57t== "Very Well"] <- 0.25
man.x[sa$q57t== "Don't know/Haven't heard enough"] <- 0
man.x[sa$q57t== "Refused"] <- NA
man.x[sa$q57t== "Missing"]  <- NA
table(man.x)


###Make labels -- in the right order###
man.labs <- c("Very Badly", "Fairly Badly", "Fairly Well", "Very Well", "Don't Know/Haven't heard enough ")

###Make factor###
q57t <- factor(man.x, levels=c(1,0.75,0.5,0.25,0), labels=man.labs)

###Check coding###
table(q57t, sa$q57t, exclude=NULL)

###Add to data frame)
sa <- data.frame(sa, q57t)

###Graphing###
codedman <- table(q57t)
codedman <- prop.table(codedman)
barplot((codedman), main=" Management of immigration by Mbeki admin", ylab="Proportions")


###Dichotomous###
man.x2 <- numeric() ### Initialize vector
man.x2[sa$q57t== "Very Badly"] <- 1
man.x2[sa$q57t== "Fairly Badly"] <- 1
man.x2[sa$q57t== "Fairly Well"] <- 0
man.x2[sa$q57t== "Very Well"] <- 0
man.x2[sa$q57t== "Don't know/Haven't heard enough"] <- NA
man.x2[sa$q57t== "Refused"] <- NA
man.x2[sa$q57t== "Missing"]  <- NA
table(man.x2)

###Make labels -- in the right order###
man.labs2 <- c("Badly","Well")

###Make factor###
man <- factor(man.x2, levels=c(1,0), labels=man.labs2)
table(man)


############################################################################
###Likelihood of action against foreigners to move into your neighborhood###
############################################################################

sa$q77a_saf <- sa$q77a
uncodedact <- table(sa$q77a)
uncodedact
barplot((uncodedact), main=" Likelihood of action against foreigners to move into your neighborhood")



act.x <- numeric() ### Initialize vector
act.x[sa$q77a=="Very likely"] <- 1
act.x[sa$q77a=="Likely"] <- 0.75
act.x[sa$q77a=="Not very likely"] <- 0.5
act.x[sa$q77a=="Not likely at all"] <- 0.25
act.x[sa$q77a=="Don't know"] <- 0
act.x[sa$q77a=="Refused"] <- NA
act.x[sa$q77a=="Missing"]  <- NA
table(act.x)


###Make labels###
act.labs <- c("Very likely", "likely", "Not very likely", "Not likely at all", "Don't Know")

###Make factor###
q77a <- factor(act.x, levels=c(1,0.75,0.5,0.25,0), labels=act.labs)

###Check coding###
table(q77a, sa$q77a, exclude=NULL)

###Add to data frame###
sa <- data.frame(sa, q77a)

###Graphing###
codedact <- table(q77a)
codedact <- prop.table(codedact)
barplot((codedact), main="Likely of action against foreigners moving into your neighborhood", ylab="Proportions")



###Dichotomous###
act.x2 <- numeric() ### Initialize vector
act.x2[sa$q77a=="Very likely"] <- 1
act.x2[sa$q77a=="Likely"] <- 1
act.x2[sa$q77a=="Not very likely"] <- 0
act.x2[sa$q77a=="Not likely at all"] <- 0
act.x2[sa$q77a=="Don't know"] <- NA
act.x2[sa$q77a=="Refused"] <- NA
act.x2[sa$q77a=="Missing"]  <- NA
table(act.x2)

###Make labels###
act.labs2 <- c("Likely", "Not likely")

###Make factor###
act <- factor(act.x2, levels=c(1,0), labels=act.labs2)
table(act)



###Recoding education###
table(sa$q89)
educ.x <- numeric()
educ.x[sa$q89=="No formal schooling"] <- 0
educ.x[sa$q89=="Informal schooling only"] <- 0
educ.x[sa$q89=="Some primary schooling"] <- 0.25
educ.x[sa$q89=="Primary school completed"] <- 0.25
educ.x[sa$q89=="Some secondary school/high school"] <- 0.5
educ.x[sa$q89=="Secondary school completed/high school"] <- 0.5
educ.x[sa$q89=="Post-secondary qualifications, not university"] <- 0.75
educ.x[sa$q89=="Some university"] <- 0.75
educ.x[sa$q89=="University completed"] <- 1
educ.x[sa$q89=="Post-graduate"] <- 1
educ.x[sa$q89=="Don't know"] <- NA
educ.x[sa$q89=="Refused"] <- NA
educ.x[sa$q89=="Missing"] <- NA
table(educ.x)


###Make labels -- in the right order###
edu.labs <- c("No formal education, Incomplete primary school","Complete primary education, Incomplete secondary school","Complete secondary school, Incomplete technical school","Complete technical school, Some University-level, without a degree", "University-level education, with degree")

###Make factor###
q89 <- factor(educ.x, levels=c(1,0.75,0.5,0.25,0), labels=edu.labs)

###Check coding###
table(q89, sa$q89, exclude=NULL)

###Add to data frame###
sa <- data.frame(sa, q89)

###Graphing###
codedemp<- table(q89)
codedemp <- prop.table(codedemp)
barplot((codedemp), main="Levels of education")


###Dichotomous###

educ2.x <- numeric()
educ2.x[sa$q89=="No formal schooling"] <- 0
educ2.x[sa$q89=="Informal schooling only"] <- 0
educ2.x[sa$q89=="Some primary schooling"] <- 0
educ2.x[sa$q89=="Primary school completed"] <- 0
educ2.x[sa$q89=="Some secondary school/high school"] <- 0
educ2.x[sa$q89=="Secondary school completed/high school"] <- 1
educ2.x[sa$q89=="Post-secondary qualifications, not university"] <- 1
educ2.x[sa$q89=="Some university"] <- 1
educ2.x[sa$q89=="University completed"] <- 1
educ2.x[sa$q89=="Post-graduate"] <- 1
educ2.x[sa$q89=="Don't know"] <- NA
educ2.x[sa$q89=="Refused"] <- NA
educ2.x[sa$q89=="Missing"] <- NA
table(educ2.x)

###Make labels -- in the right order###
educ2.labs <- c("Low level of education", "High level of education")

###Make factor###
educ2 <- factor(educ2.x, levels=c(0:1), labels=educ2.labs)

###Check coding###
table(educ2, sa$q89, exclude=NULL)

###Add to data frame###
sa <- data.frame(sa, educ2)

###Graphing###
codededuc2<- table(educ2)
codededuc2 <- prop.table(codededuc2)
barplot((codededuc2), main="Levels of education")
dotchart(rev(codededuc2), main=" Levels of education", xlab="Relative Propotions",cex=0.8, color=c("red", "blue"))


###Recoding employment###
table(sa$q94)
empl.x <- numeric()
empl.x[sa$q94=="Missing"] <- NA
empl.x[sa$q94=="No (not looking)"] <- 1
empl.x[sa$q94=="No (looking) Yes"] <- 2
empl.x[sa$q94=="part time (not looking)"] <- 3
empl.x[sa$q94=="Yes, part time (looking)"] <- 4
empl.x[sa$q94=="Yes, full time (not looking)"] <- 5
empl.x[sa$q94=="Yes, full time (looking)"] <- 6
empl.x[sa$q94=="Don't know"] <- NA
empl.x[sa$q94=="Refused"] <- NA

table(empl.x)

###Making labels -- in the right order###
empl.labs <- c("Not employed (not looking)","Not employed (looking)", "Yes,part time (not looking)","Yes, part time (looking)","Yes, full time (not looking)","Yes, full time (looking)")

###Make factor###
q94 <- factor(empl.x, levels=1:6, labels=empl.labs)
table(q94, sa$q94, exclude=NULL)

###Add to data frame###
sa <- data.frame(sa, q94)

codedempl <- table(q94)
codedempl
barplot((codedempl), main="Levels of employment")
dotchart(rev(codedempl), main="Levels of employment", xlab="Relative Propotions",cex=0.8, color=c("red", "blue"))


table(sa$q94)
empl2.x <- numeric()
empl2.x[sa$q94=="Missing"] <- NA
empl2.x[sa$q94=="No (not looking)"] <- 0
empl2.x[sa$q94=="No (looking) Yes"] <- 0
empl2.x[sa$q94=="part time (not looking)"] <- 0
empl2.x[sa$q94=="Yes, part time (looking)"] <- 0
empl2.x[sa$q94=="Yes, full time (not looking)"] <- 1
empl2.x[sa$q94=="Yes, full time (looking)"] <- 1
empl2.x[sa$q94=="Don't know"] <- NA
empl2.x[sa$q94=="Refused"] <- NA

table(empl2.x)

###Making labels -- in the right order###
empl2.labs <- c("Not employed","Employed(full time)")

###Make factor###
empl2 <- factor(empl2.x, levels=0:1, labels=empl2.labs)
table(empl2, sa$q94, exclude=NULL)

###Add to data frame###
sa <- data.frame(sa, empl2)

###Graphing###
codedempl2 <- table(empl2)
barplot((codedempl2), main="Levels of employment")
dotchart(rev(codedempl2), main=" Levels of employment", xlab="Relative Propotions",cex=0.8, color=c("red", "blue"))



###Prop diff###
par(mfrow=c(1,2))
attivsedu <- table(atti, educ2)
attivsedu.d <- prop.table(attivsedu, margin=2)
attivsedu.d
d <- attivsedu.d[2,1]-attivsedu.d[1,2]
d
mosaicplot(t(attivsedu.d), main="Education", ylab="Trust towards immigrants", xlab="")
prop.test(attivsedu)


###
attivsemp <- table(atti,empl2)
attivsemp.d <- prop.table(attivsemp, margin=2)
attivsemp.d
d <- attivsemp.d[1,1]-attivsemp.d[1,2]
d
mosaicplot(t(attivsemp), main="Employment", ylab="Trust towards immigrants", xlab="")
prop.test(attivsemp)
dev.off()



####
par(mfrow=c(1,2))
tolvsedu <- table(tol, educ2)
tolvsedu.d <- prop.table(tolvsedu, margin=2)
tolvsedu.d
d <- tolvsedu.d[1,1]-tolvsedu.d[1,2]
d
mosaicplot(t(tolvsedu.d), main="Education", ylab="Tolerance towards immigration", xlab="")
prop.test(tolvsedu)

###
tolvsemp <- table(tol,empl2)
tolvsemp.d <- prop.table(tolvsemp, margin=2)
tolvsemp.d
d <- tolvsemp.d[1,1]-tolvsemp.d[1,2]
d
mosaicplot(t(tolvsemp), main="Employment", ylab="Tolerance towards immigration", xlab="")
prop.test(tolvsemp)
dev.off()



###

par(mfrow=c(1,2))
manvsedu <- table(man, educ2)
manvsedu.d <- prop.table(manvsedu, margin=2)
manvsedu.d
d <- manvsedu.d[1,1]-manvsedu.d[1,2]
d
mosaicplot(t(manvsedu.d), main="Education", ylab="Immigration management by Mbeki Admin", xlab="")

###
manvsemp <- table(man,empl2)
manvsemp.d <- prop.table(manvsemp, margin=2)
manvsemp.d
d <- manvsemp.d[1,1]-manvsemp.d[1,2]
d
mosaicplot(t(manvsemp), main="Employment", ylab="Immigration management by Mbeki Admin", xlab="")
dev.off()



###
par(mfrow=c(1,2))
actvsedu <- table(act, educ2)
actvsedu.d <- prop.table(actvsedu, margin=2)
actvsedu.d
d <- actvsedu.d[1,1]-actvsedu.d[1,2]
d
mosaicplot(t(actvsedu.d), main="Education", ylab="Likelihood of action against foreign neighbours", xlab="")

###
actvsemp <- table(act,empl2)
actvsemp.d <- prop.table(actvsemp, margin=2)
actvsemp.d
d <- actvsemp.d[1,1]-actvsemp.d[1,2]
d
mosaicplot(t(actvsemp), main="Employment", ylab="Likelihood of action against foreign neighbours", xlab="")

dev.off()