Data preparation Eurobarometer

In this script, I select, recode and create variables of the Eurobarometer

rm(list = ls())
library(plyr)
library(dplyr)
library(foreign)
library(tidyverse)
library(labelled)
library(ggplot2)
library(questionr)
library(psych)

Years

1986

# Eurobarometer has the most waves
#1986.
eb1986 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb1986.sav", use.value.labels = T,  to.data.frame = T)

eb1986sel <- eb1986 %>% filter(v5=="NETHERLANDS") %>%
            select(v1, v4, v8, v9, v218, v161, v270, v280, v281, v283, v284, v290, v291, v296, v302, v303)


# Dependent variables
table(eb1986sel$v161, useNA = "always") #19 missings. Have to recode bc now a low score means that it's not really a problem. 
eb1986sel$env_prsimp <- as.numeric(eb1986sel$v161)
eb1986sel <- eb1986sel %>% 
     mutate_at(c("env_prsimp"), funs(recode(., `1` = 3, `2` = 2, `3` = 1)))
eb1986sel$env_prsimp <- (eb1986sel$env_prsimp -1) * (4/2) + 1
table(eb1986sel$v218, useNA = "always") #60 missings. 1 means development of economy is more important than environment, 3 means protecting environment are necessary for economic development --> higher score thus more positive. 
eb1986sel$env_ec_stat <- as.numeric(eb1986sel$v218)
eb1986sel$env_ec_stat <- (eb1986sel$env_ec_stat -1) * (4/2) + 1

# Independent variables 
table(eb1986sel$v270, useNA = "always") #55 missings on rl placement. Numeric: highest score is most right. 
eb1986sel$lrplace <- as.numeric(eb1986sel$v270)
table(eb1986sel$v281, useNA = "always") #15 missings on religion. 
eb1986sel$relig <- as.numeric(eb1986sel$v281)
table(eb1986sel$v284, useNA = "always") #No missings education. 
eb1986sel$eduyrs_cat <- eb1986sel$v284 
eb1986sel$eduyrs <- revalue(eb1986sel$eduyrs_cat, c("UP TO 14 YEARS"= "14", "15 YEARS"="15", "16 YEARS" = "16", "17 YEARS" = "17", "18 YEARS" = "18", "19 YEARS" = "19", "20 YEARS" = "20", "21 YEARS" = "21", "22 YRS OR OLDER" = "22", "STILL STUDYING" = NA))

table(eb1986sel$eduyrs, useNA = "always")
eb1986sel$eduyrs <- as.numeric(as.character(eb1986sel$eduyrs))
eb1986sel$eduyrs <- eb1986sel$eduyrs - 6

# I want the people that are still studying to have the mean educational years. Do this before categorizing them so that they can be in one of the isced categories
mean(eb1986sel$eduyrs, na.rm=T)
eb1986sel$eduyrs[is.na(eb1986sel$eduyrs)] <- 11.519
eb1986sel$isced[eb1986sel$eduyrs <=4] <- 0
eb1986sel$isced[eb1986sel$eduyrs > 4 & eb1986sel$eduyrs <= 6] <- 1
eb1986sel$isced[eb1986sel$eduyrs > 6 & eb1986sel$eduyrs <= 10] <- 2
eb1986sel$isced[eb1986sel$eduyrs > 10 & eb1986sel$eduyrs <= 13] <- 3
eb1986sel$isced[eb1986sel$eduyrs > 13 & eb1986sel$eduyrs <= 15] <- 4
eb1986sel$isced[eb1986sel$eduyrs > 15 & eb1986sel$eduyrs <= 18] <- 5
eb1986sel$isced[eb1986sel$eduyrs > 18] <- 6
table(eb1986sel$isced, useNA = "always")
                            
table(eb1986sel$v290, useNA = "always") 
eb1986sel$sex <- revalue(eb1986sel$v290, c("MAN"="1", "WOMAN"="2"))
table(eb1986sel$v291, useNA = "always") #No missings on age. 
eb1986sel$age <- as.numeric(as.character(eb1986sel$v291))
table(eb1986sel$v296, useNA = "always") #159 missings income. 
eb1986sel$income <- as.numeric(eb1986sel$v296) #Household income in 12 cats
mean(eb1986sel$income, na.rm=T)
eb1986sel$income[is.na(eb1986sel$income)] <- 7.049
eb1986sel$income_quart <- with(eb1986sel, cut(income, 
                                breaks=quantile(income, probs=seq(0,1, by=0.25), na.rm=TRUE), 
                                include.lowest=TRUE))
table(eb1986sel$income_quart, useNA = "always")
eb1986sel$income_quart <- as.numeric(eb1986sel$income_quart)
table(eb1986sel$urban, useNA = "always") #no missings on community
eb1986sel$urban <- revalue(eb1986sel$v302, c("RURAL AREA - VIL"= "Low urbanity", "SM,MDL SZE TOWN"="Medium urbanity", "BIG TOWN" = "High urbanity"))
table(eb1986sel$v283, useNA = "always")

eb1986sel$v283 <- as.numeric(eb1986sel$v283)
eb1986sel$marstat[eb1986sel$v283 == 2 | eb1986sel$v283 == 3] <- 1 #Living together
eb1986sel$marstat[eb1986sel$v283 != 2 & eb1986sel$v283 != 3] <- 2 #Not living together
eb1986sel$marstat <- as.factor(eb1986sel$marstat)
table(eb1986sel$marstat, useNA = "always" )

# Missings
lapply(eb1986sel, table, useNA = "always")

mis_vars <- c("lrplace") 

for (var in mis_vars) {
  eb1986sel[is.na(eb1986sel[,var]), var] <- mean(eb1986sel[,var], na.rm = TRUE)
}

eb1986sel <- eb1986sel %>% select(v1, v8, env_ec_stat, env_prsimp, sex, eduyrs_cat, isced, lrplace, age, income, income_quart, urban, relig, marstat)

save(eb1986sel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb1986sel.Rdata")

1992

#Continue with 1992
eb1992 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb1992.sav", use.value.labels = T,  to.data.frame = T)

eb1992sel <- eb1992 %>% filter(v5=="NETHERLANDS") %>%
  select(v1, v8, v9, v456, v449, v645, v647, v649, v601, v602, v610, v611, v615, v616)

#Dependent variables 
table(eb1992sel$v456)
eb1992sel$env_ec_stat <- as.numeric(eb1992sel$v456)
eb1992sel$env_ec_stat <- (eb1992sel$env_ec_stat - 1) * (4/2) + 1
attributes(eb1992sel$v449) #After recode --> higher score means more of a problem. 
eb1992sel$env_prsimp <- as.numeric(eb1992sel$v449)
eb1992sel <- eb1992sel %>% 
     mutate_at(c("env_prsimp"), funs(recode(., `1` = 3, `2` = 2, `3` = 1)))
eb1992sel$env_prsimp <- (eb1992sel$env_prsimp -1) * (4/2) + 1

#Independent vars
table(eb1992sel$v645, useNA = "always") #9 missings on urbanity.
eb1992sel$urban <- revalue(eb1992sel$v645, c("RURAL AREA/VILLAGE"= "Low urbanity", "SMALL/MIDDLE TOWN"="Medium urbanity", "LARGE TOWN" = "High urbanity"))
eb1992sel$urban <- factor(eb1992sel$urban, levels=c("Low urbanity", "Medium urbanity", "High urbanity"),  ed=TRUE)
table(eb1992sel$v647, useNA = "always") 
eb1992sel$ch_attend <- as.numeric(eb1992sel$v647)
eb1992sel <- eb1992sel %>% 
     mutate_at(c("ch_attend"), funs(recode(., `1` = 5, `2` = 4, `3` = 3, `4` = 2, `5`= 1)))
table(eb1992sel$v649, useNA = "always") #Income has 143 missings. In the same 12 categories as first wave
eb1992sel$income <- as.numeric(eb1992sel$v649)
mean(eb1992sel$income, na.rm=T)
eb1992sel$income[is.na(eb1992sel$income)] <- 7.427
eb1992sel$income_quart <- with(eb1992sel, cut(income, 
                                breaks=quantile(income, probs=seq(0,1, by=0.25), na.rm=TRUE), 
                                include.lowest=TRUE))
table(eb1992sel$income_quart, useNA = "always")
eb1992sel$income_quart <- as.numeric(eb1992sel$income_quart)

table(eb1992sel$v601, useNA = "always") #78 missings on left right placement. 
eb1992sel$lrplace <- as.numeric(eb1992sel$v601)
table(eb1992sel$v611, useNA = "always") #No missings on education
eb1992sel$eduyrs <- as.numeric(as.character(eb1992sel$v611))
eb1992sel$eduyrs <- eb1992sel$eduyrs - 6

# I want the people that are still studying to have the mean educational years. Do this before categorizing them so that they can be in one of the isced categories
mean(eb1992sel$eduyrs, na.rm=T)
eb1992sel$eduyrs[is.na(eb1992sel$eduyrs)] <- 12.138
eb1992sel$isced[eb1992sel$eduyrs <=4] <- 0
eb1992sel$isced[eb1992sel$eduyrs > 4 & eb1992sel$eduyrs <= 6] <- 1
eb1992sel$isced[eb1992sel$eduyrs > 6 & eb1992sel$eduyrs <= 10] <- 2
eb1992sel$isced[eb1992sel$eduyrs > 10 & eb1992sel$eduyrs <= 13] <- 3
eb1992sel$isced[eb1992sel$eduyrs > 13 & eb1992sel$eduyrs <= 15] <- 4
eb1992sel$isced[eb1992sel$eduyrs > 15 & eb1992sel$eduyrs <= 18] <- 5
eb1992sel$isced[eb1992sel$eduyrs > 18] <- 6
table(eb1992sel$isced, useNA = "always")

table(eb1992sel$v615, useNA = "always")#No missings gender
eb1992sel$sex <- revalue(eb1992sel$v615, c("MALE"="1", "FEMALE"="2"))
eb1992sel$age <- as.numeric(as.character(eb1992sel$v616))

eb1992sel$v610 <- as.numeric(eb1992sel$v610)
eb1992sel$marstat[eb1992sel$v610 == 2 | eb1992sel$v610 == 3] <- 1 #Living together
eb1992sel$marstat[eb1992sel$v610 != 2 & eb1992sel$v610 != 3] <- 2 #Not living together
eb1992sel$marstat <- as.factor(eb1992sel$marstat)
table(eb1992sel$marstat, useNA = "always" )

# Missings
lapply(eb1992sel, table, useNA = "always")

mis_vars <- c("lrplace") 

for (var in mis_vars) {
  eb1992sel[is.na(eb1992sel[,var]), var] <- mean(eb1992sel[,var], na.rm = TRUE)
}

eb1992sel <- eb1992sel %>% select(v1, v8, env_ec_stat, env_prsimp, lrplace, eduyrs, isced, sex, age, urban, income, income_quart, ch_attend, marstat)

save(eb1992sel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb1992sel.Rdata")

1995

#1995
eb1995 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb1995.sav", use.value.labels = T,  to.data.frame = T)

eb1995sel <- eb1995 %>% filter(v5=="Netherlands") %>%
  select(v1, v8, v9, v182, v134, v339, v342, v343, v344, v345, v354, v355)

#Dependent variables
attributes(eb1995sel$v182) #Higher score means environment higher priority. 
eb1995sel$env_ec_stat <- as.numeric(eb1995sel$v182)
eb1995sel$env_ec_stat <- (eb1995sel$env_ec_stat-1) * (4/2) + 1
attributes(eb1995sel$v134)
eb1995sel$env_prsimp <- as.numeric(eb1995sel$v134)
eb1995sel <- eb1995sel %>% 
     mutate_at(c("env_prsimp"), funs(recode(., `1` = 3, `2` = 2, `3` = 1)))
eb1995sel$env_prsimp <- (eb1995sel$env_prsimp -1) * (4/2) + 1

#Independent variables
table(eb1995sel$v339, useNA = "always") #90 missings on lr placement
eb1995sel$lrplace <- as.numeric(eb1995sel$v339)
table(eb1995sel$v343, useNA = "always") #No missings education
eb1995sel$eduyrs_cat <- eb1995sel$v343
# This time its only asked in large categories
eb1995sel$eduyrs_cat <- as.numeric(eb1995sel$eduyrs_cat)
table(eb1995sel$eduyrs_cat, useNA = "always")
eb1995sel$eduyrs_cat[eb1995sel$eduyrs_cat ==4] <- NA
mean(eb1995sel$eduyrs_cat, na.rm=T)
eb1995sel$eduyrs_cat[is.na(eb1995sel$eduyrs_cat)] <- 2.149
eb1995sel$isced[eb1995sel$eduyrs_cat == 1] <- 2
eb1995sel$isced[eb1995sel$eduyrs_cat >= 2 & eb1995sel$eduyrs_cat < 3] <- 3
eb1995sel$isced[eb1995sel$eduyrs_cat == 3] <- 5
table(eb1995sel$isced, useNA = "always")

table(eb1995sel$v344, useNA = "always") 
eb1995sel$sex <- revalue(eb1995sel$v344, c("Male"="1", "Female"="2"))
table(eb1995sel$v345, useNA = "always") #No missings age
eb1995sel$v345 <- revalue(eb1995sel$v345, c("15 years"="15"))
eb1995sel$age <- as.numeric(as.character(eb1995sel$v345))
table(eb1995sel$v354, useNA = "always") #No missings urbanity
eb1995sel$urban <- revalue(eb1995sel$v354, c("Rural area or village"= "Low urbanity", "Small or middle size town"="Medium urbanity", "Large town" = "High urbanity"))
table(eb1995sel$v355, useNA = "always") #199 missings  income, still in 12 cats. 
eb1995sel$income <- as.numeric(eb1995sel$v355)


mean(eb1995sel$income, na.rm=T)
eb1995sel$income[is.na(eb1995sel$income)] <- 6.831
eb1995sel$income_quart <- with(eb1995sel, cut(income, 
                                breaks=quantile(income, probs=seq(0,1, by=0.25), na.rm=TRUE), 
                                include.lowest=TRUE))
table(eb1995sel$income_quart, useNA = "always")
eb1995sel$income_quart <- as.numeric(eb1995sel$income_quart)

eb1995sel$v342 <- as.numeric(eb1995sel$v342)
eb1995sel$marstat[eb1995sel$v342 == 2 | eb1995sel$v342 == 3] <- 1 #Living together
eb1995sel$marstat[eb1995sel$v342 != 2 & eb1995sel$v342 != 3] <- 2 #Not living together
eb1995sel$marstat <- as.factor(eb1995sel$marstat)
table(eb1995sel$marstat, useNA = "always" )

# Missings
lapply(eb1995sel, table, useNA = "always")

mis_vars <- c("lrplace") 

for (var in mis_vars) {
  eb1995sel[is.na(eb1995sel[,var]), var] <- mean(eb1995sel[,var], na.rm = TRUE)
}

eb1995sel <- eb1995sel %>% select(v1, v8, env_prsimp, env_ec_stat, lrplace, eduyrs_cat, isced, sex, age, urban, income, income_quart, marstat)

save(eb1995sel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb1995sel.Rdata")

2004

#2004
eb2004 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb2004.sav", use.value.labels = T,  to.data.frame = T)

eb2004sel <- eb2004 %>% filter(v6=="The Netherlands") %>%
  select(v1, v8, v307, v312, v579, v582, v583, v585, v586, v58)

#Dependent vars 

eb2004sel$env_quallife <- as.numeric(eb2004sel$v307)
eb2004sel$pers_effort <- as.numeric(eb2004sel$v312)

eb2004sel <- eb2004sel %>% 
     mutate_at(c("env_quallife", "pers_effort"), funs(recode(., `1` = 4, `2` = 3, `3` = 2, `4` = 1)))

eb2004sel$env_quallife <- (eb2004sel$env_quallife - 1) * (4/3) + 1
eb2004sel$pers_effort <- (eb2004sel$pers_effort -1) * (4/3) + 1

#Independent variables 
table(eb2004sel$v579, useNA = "always") #54 missings on left-right placement. 
eb2004sel$lrplace <- as.numeric(eb2004sel$v579)

table(eb2004sel$v583, useNA = "always") #8 missings on education. 
eb2004sel$eduyrs <- as.numeric(as.character(eb2004sel$v583))
table(eb2004sel$eduyrs, useNA = "always") #Again there are missings for the individuals who are still studying, so have to replace those later
eb2004sel$eduyrs <- eb2004sel$eduyrs - 6

# I want the people that are still studying to have the mean educational years. Do this before categorizing them so that they can be in one of the isced categories
mean(eb2004sel$eduyrs, na.rm=T)
eb2004sel$eduyrs[is.na(eb2004sel$eduyrs)] <- 14.020
eb2004sel$isced[eb2004sel$eduyrs <=4] <- 0
eb2004sel$isced[eb2004sel$eduyrs > 4 & eb2004sel$eduyrs <= 6] <- 1
eb2004sel$isced[eb2004sel$eduyrs > 6 & eb2004sel$eduyrs <= 10] <- 2
eb2004sel$isced[eb2004sel$eduyrs > 10 & eb2004sel$eduyrs <= 13] <- 3
eb2004sel$isced[eb2004sel$eduyrs > 13 & eb2004sel$eduyrs <= 15] <- 4
eb2004sel$isced[eb2004sel$eduyrs > 15 & eb2004sel$eduyrs <= 18] <- 5
eb2004sel$isced[eb2004sel$eduyrs > 18] <- 6
table(eb2004sel$isced, useNA = "always")

table(eb2004sel$v585, useNA = "always") 
eb2004sel$sex <- revalue(eb2004sel$v585, c("Male"="1", "Female"="2"))
table(eb2004sel$v586, useNA = "always") 
eb2004sel$urban <- revalue(eb2004sel$v586, c("Rural area or village"= "Low urbanity", "Small or middle sized town"="Medium urbanity", "Large town" = "High urbanity"))
table(eb2004sel$v58, useNA = "always") #No missings on age
eb2004sel$v58 <- revalue(eb2004sel$v58, c("15 years"="15"))
eb2004sel$age <- as.numeric(as.character(eb2004sel$v58))
table(eb2004sel$v582, useNA = "always" )
eb2004sel$v582 <- as.numeric(eb2004sel$v582)
eb2004sel$marstat[eb2004sel$v582 <=3] <- 1 #Living together
eb2004sel$marstat[eb2004sel$v582 > 3] <- 2 #Not living together
eb2004sel$marstat <- as.factor(eb2004sel$marstat)
table(eb2004sel$marstat, useNA = "always" )

#Missings
lapply(eb2004sel, table, useNA = "always")

mis_vars <- c("lrplace") 

for (var in mis_vars) {
  eb2004sel[is.na(eb2004sel[,var]), var] <- mean(eb2004sel[,var], na.rm = TRUE)
}


eb2004sel <- eb2004sel %>% select(v1, v8, env_quallife, pers_effort, lrplace, age, urban, sex, eduyrs, isced, marstat)

save(eb2004sel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2004sel.Rdata")

2007

eb2007 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb2007.sav", use.value.labels = T,  to.data.frame = T)

eb2007sel <- eb2007 %>% filter(v6=="The Netherlands") %>%
  select(v1, v8, v548, v543, v544, v592, v593, v594, v549, v474, v612, v615, v616, v618, v619, v624)

attributes(eb2007sel$v548) #Recode. Now a higher score means that environment has priority over economic competitiveness.
eb2007sel$env_vs_ec <- as.numeric(eb2007sel$v548)
eb2007sel <- eb2007sel %>% 
     mutate_at(c("env_vs_ec"), funs(recode(., `1` = 2, `2` = 1)))
eb2007sel$env_vs_ec <- (eb2007sel$env_vs_ec -1) * (4/1) + 1
table(eb2007sel$env_vs_ec, useNA = "always")

eb2007sel$env_quallife <- as.numeric(eb2007sel$v544)
eb2007sel$role_ind <- as.numeric(eb2007sel$v592)
eb2007sel$big_pol <- as.numeric(eb2007sel$v593)
eb2007sel$eff_daily <- as.numeric(eb2007sel$v594)
eb2007sel$buyprod <- as.numeric(eb2007sel$v549)
eb2007sel$pers_imp <- as.numeric(eb2007sel$v474)

eb2007sel <- eb2007sel %>% 
     mutate_at(c("env_quallife", "role_ind", "big_pol", "eff_daily", "buyprod", "pers_imp"), funs(recode(., `1` = 4, `2` = 3, `3` = 2, `4` = 1)))

eb2007sel$env_quallife <- (eb2007sel$env_quallife-1) * (4/3) + 1
eb2007sel$role_ind <- (eb2007sel$role_ind-1) * (4/3) + 1
eb2007sel$big_pol <- (eb2007sel$big_pol-1) * (4/3) + 1
eb2007sel$eff_daily <- (eb2007sel$eff_daily-1) * (4/3) + 1
eb2007sel$buyprod <- (eb2007sel$buyprod-1) * (4/3) + 1
eb2007sel$pers_imp <- (eb2007sel$pers_imp-1) * (4/3) + 1

# Independent variables 
table(eb2007sel$v612, useNA = "always") #45 missings on l-r placement. Replace with mean. 
eb2007sel$lrplace <- as.numeric(eb2007sel$v612)
table(eb2007sel$v616, useNA = "always") #8 missings on educyrs. 
eb2007sel$eduyrs <- as.numeric(as.character(eb2007sel$v616)) #Same as above about the people still studying

eb2007sel$eduyrs <- eb2007sel$eduyrs - 6

# I want the people that are still studying to have the mean educational years. Do this before categorizing them so that they can be in one of the isced categories
mean(eb2007sel$eduyrs, na.rm=T)
eb2007sel$eduyrs[is.na(eb2007sel$eduyrs)] <- 13.772
eb2007sel$isced[eb2007sel$eduyrs <=4] <- 0
eb2007sel$isced[eb2007sel$eduyrs > 4 & eb2007sel$eduyrs <= 6] <- 1
eb2007sel$isced[eb2007sel$eduyrs > 6 & eb2007sel$eduyrs <= 10] <- 2
eb2007sel$isced[eb2007sel$eduyrs > 10 & eb2007sel$eduyrs <= 13] <- 3
eb2007sel$isced[eb2007sel$eduyrs > 13 & eb2007sel$eduyrs <= 15] <- 4
eb2007sel$isced[eb2007sel$eduyrs > 15 & eb2007sel$eduyrs <= 18] <- 5
eb2007sel$isced[eb2007sel$eduyrs > 18] <- 6
table(eb2007sel$isced, useNA = "always")

table(eb2007sel$v618, useNA = "always") #No  missings on gender. 
eb2007sel$sex <- revalue(eb2007sel$v618, c("Male"="1", "Female"="2"))
table(eb2007sel$v619, useNA = "always") #No missings on age. 
eb2007sel$v619 <- revalue(eb2007sel$v619, c("15 years"="15"))
eb2007sel$age <- as.numeric(as.character(eb2007sel$v619))
table(eb2007sel$v624, useNA = "always") #No missings on urbanity. 
eb2007sel$urban <- revalue(eb2007sel$v624, c("Rural area or village"= "Low urbanity", "Small/middle town"="Medium urbanity", "Large town" = "High urbanity"))
table(eb2007sel$urban)

table(eb2007sel$v615, useNA = "always" )
eb2007sel$v615 <- as.numeric(eb2007sel$v615)
eb2007sel$marstat[eb2007sel$v615 <=3] <- 1 #Living together
eb2007sel$marstat[eb2007sel$v615 > 3] <- 2 #Not living together
eb2007sel$marstat <- as.factor(eb2007sel$marstat)
table(eb2007sel$marstat, useNA = "always" )

#Missings
lapply(eb2007sel, table, useNA = "always")

mis_vars <- c("lrplace") 

for (var in mis_vars) {
  eb2007sel[is.na(eb2007sel[,var]), var] <- mean(eb2007sel[,var], na.rm = TRUE)
}

eb2007sel <- eb2007sel %>% select(v1, v8, env_vs_ec, env_quallife, role_ind, big_pol, buyprod, eff_daily, lrplace, sex, age, eduyrs, isced, urban, marstat)

save(eb2007sel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2007sel.Rdata")

2008

#2008
eb2008 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb2008.sav", use.value.labels = T,  to.data.frame = T)

eb2008sel <- eb2008 %>% filter(v6=="The Netherlands") %>%
  select(v1, v8, v647, v725, v705, v720, v721, v722, v723, v761, v764, v765, v767, v768, v773)

#Dependent variables 
attributes(eb2008sel$v720) #Climate change is an unstoppable process. Difficult, bc even if you believe in it/are concerned you may think it's unstoppable. For now ill code it as more agreeness --> more positive
attributes(eb2008sel$v723) #Fighting climate change has positive impact on EU. Again, can't simply divide that into more positive. You may still think climate change is important and that it does not matter that it has a negative impact on the economy. For now I recode it, because then it means that you at least don't think the economy is an obstacle in fighting climate change. 
attributes(eb2008sel$v721) #Seriousness is exaggerated. If you disagree, you have more positive attitudes. 

eb2008sel$cc_unstop <- as.numeric(eb2008sel$v720)
eb2008sel$cc_poseu <- as.numeric(eb2008sel$v723)
eb2008sel$cc_prsact <- as.numeric(eb2008sel$v725)


eb2008sel <- eb2008sel %>% 
     mutate_at(c("cc_unstop", "cc_poseu", "cc_prsact"), funs(recode(., `1` = 4, `2` = 3, `3` = 2, `4` = 1)))

eb2008sel$cc_exag <- as.numeric(eb2008sel$v721)
eb2008sel$cc_exag <- (eb2008sel$cc_exag - 1) * (4/3) + 1
eb2008sel$cc_unstop <- (eb2008sel$cc_unstop - 1) * (4/3) + 1
eb2008sel$cc_poseu <- (eb2008sel$cc_poseu - 1) * (4/3) + 1
eb2008sel$cc_prsact <- (eb2008sel$cc_prsact - 1) * (4/3) + 1

#Independent variables 
table(eb2008sel$v761, useNA = "always") #41 missings on l-r placement. 
eb2008sel$lrplace <- as.numeric(eb2008sel$v761)
table(eb2008sel$v765, useNA = "always") 
eb2008sel$eduyrs <- as.numeric(as.character(eb2008sel$v765)) #Same as earlier about the missings on education and people still studying
eb2008sel$eduyrs <- eb2008sel$eduyrs - 6

# I want the people that are still studying to have the mean educational years. Do this before categorizing them so that they can be in one of the isced categories
mean(eb2008sel$eduyrs, na.rm=T)
eb2008sel$eduyrs[is.na(eb2008sel$eduyrs)] <- 14.175
eb2008sel$isced[eb2008sel$eduyrs <=4] <- 0
eb2008sel$isced[eb2008sel$eduyrs > 4 & eb2008sel$eduyrs <= 6] <- 1
eb2008sel$isced[eb2008sel$eduyrs > 6 & eb2008sel$eduyrs <= 10] <- 2
eb2008sel$isced[eb2008sel$eduyrs > 10 & eb2008sel$eduyrs <= 13] <- 3
eb2008sel$isced[eb2008sel$eduyrs > 13 & eb2008sel$eduyrs <= 15] <- 4
eb2008sel$isced[eb2008sel$eduyrs > 15 & eb2008sel$eduyrs <= 18] <- 5
eb2008sel$isced[eb2008sel$eduyrs > 18] <- 6
table(eb2008sel$isced, useNA = "always")

table(eb2008sel$v767, useNA = "always") #Gender no missings
eb2008sel$sex <- revalue(eb2008sel$v767, c("Male"="1", "Female"="2"))
table(eb2008sel$v768, useNA = "always") #Age neither
eb2008sel$v768 <- revalue(eb2008sel$v768, c("15 years"="15"))
eb2008sel$age <- as.numeric(as.character(eb2008sel$v768))
table(eb2008sel$v773, useNA = "always") #Urbanity neither
eb2008sel$urban <- revalue(eb2008sel$v773, c("Rural area or village"= "Low urbanity", "Small/middle town"="Medium urbanity", "Large town" = "High urbanity"))

table(eb2008sel$v764, useNA = "always" )
eb2008sel$v764 <- as.numeric(eb2008sel$v764)
eb2008sel$marstat[eb2008sel$v764 <=3] <- 1 #Living together
eb2008sel$marstat[eb2008sel$v764 > 3] <- 2 #Not living together
eb2008sel$marstat <- as.factor(eb2008sel$marstat)
table(eb2008sel$marstat, useNA = "always" )

#Missings
lapply(eb2008sel, table, useNA = "always")

mis_vars <- c("lrplace") 

for (var in mis_vars) {
  eb2008sel[is.na(eb2008sel[,var]), var] <- mean(eb2008sel[,var], na.rm = TRUE)
}


eb2008sel <- eb2008sel %>% select(v1, v8, cc_unstop, cc_exag, cc_poseu, cc_prsact, eduyrs, isced, lrplace, sex, age, urban, marstat)

save(eb2008sel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2008sel.Rdata")

2009

#2009 (2 datasets)
eb2009_1 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb2009_1.sav", use.value.labels = T,  to.data.frame = T)


eb2009asel <- eb2009_1 %>% filter(v6=="The Netherlands") %>%
  select(v1, v8, v129, v522, v526, v527, v528, v529, v530, v531, v145, v638, v641, v642, v644, v645, v650)

attributes(eb2009asel$v129) #Environmental protection is important country issue. 2 = mentioned
eb2009asel$envprotect_imp  <- as.numeric(eb2009asel$v129)
eb2009asel$envprotect_imp <- (eb2009asel$envprotect_imp -1) * (4/1) + 1

eb2009asel$pers_imp <- as.numeric(eb2009asel$v145)
eb2009asel$pers_imp <- (eb2009asel$pers_imp -1) * (4/1) + 1

eb2009asel$cc_unstop <- as.numeric(eb2009asel$v526)
eb2009asel$cc_poseu <- as.numeric(eb2009asel$v529)
eb2009asel$cc_prsact <- as.numeric(eb2009asel$v531)

eb2009asel <- eb2009asel %>% 
     mutate_at(c("cc_unstop", "cc_poseu", "cc_prsact"), funs(recode(., `1` = 4, `2` = 3, `3` = 2, `4` = 1)))

eb2009asel$cc_unstop <- (eb2009asel$cc_unstop -1) * (4/3) + 1
eb2009asel$cc_poseu <- (eb2009asel$cc_poseu -1) * (4/3) + 1
eb2009asel$cc_prsact <- (eb2009asel$cc_prsact -1) * (4/3) + 1

eb2009asel$ccpercept <- as.numeric(eb2009asel$v522)
eb2009asel$ccpercept <- (eb2009asel$ccpercept -1) * (4/9) + 1

eb2009asel$cc_exag <- as.numeric(eb2009asel$v527)
eb2009asel$cc_exag <- (eb2009asel$cc_exag -1) + (4/3) + 1

#Independent variables 
table(eb2009asel$v638, useNA = "always") #46 missings on r-l placement. 
eb2009asel$lrplace <- as.numeric(eb2009asel$v638)
table(eb2009asel$v642, useNA = "always") #Again the same as above 
eb2009asel$eduyrs <- as.numeric(as.character(eb2009asel$v642))

eb2009asel$eduyrs <- eb2009asel$eduyrs - 6

# I want the people that are still studying to have the mean educational years. Do this before categorizing them so that they can be in one of the isced categories
mean(eb2009asel$eduyrs, na.rm=T)
eb2009asel$eduyrs[is.na(eb2009asel$eduyrs)] <- 14.051
eb2009asel$isced[eb2009asel$eduyrs <=4] <- 0
eb2009asel$isced[eb2009asel$eduyrs > 4 & eb2009asel$eduyrs <= 6] <- 1
eb2009asel$isced[eb2009asel$eduyrs > 6 & eb2009asel$eduyrs <= 10] <- 2
eb2009asel$isced[eb2009asel$eduyrs > 10 & eb2009asel$eduyrs <= 13] <- 3
eb2009asel$isced[eb2009asel$eduyrs > 13 & eb2009asel$eduyrs <= 15] <- 4
eb2009asel$isced[eb2009asel$eduyrs > 15 & eb2009asel$eduyrs <= 18] <- 5
eb2009asel$isced[eb2009asel$eduyrs > 18] <- 6
table(eb2009asel$isced, useNA = "always")

table(eb2009asel$v644, useNA = "always") #no missings gender
eb2009asel$sex <- revalue(eb2009asel$v644, c("Male"="1", "Female"="2"))
table(eb2009asel$v645, useNA = "always") #No missings on age
eb2009asel$v645 <- revalue(eb2009asel$v645, c("15 years"="15"))
eb2009asel$age <- as.numeric(as.character(eb2009asel$v645))
table(eb2009asel$v650, useNA = "always") #No missings on urbanity
eb2009asel$urban <- revalue(eb2009asel$v650, c("Rural area or village"= "Low urbanity", "Small/middle town"="Medium urbanity", "Large town" = "High urbanity"))

table(eb2009asel$v641, useNA = "always" )
eb2009asel$v641 <- as.numeric(eb2009asel$v641)
eb2009asel$marstat[eb2009asel$v641 <=3] <- 1 #Living together
eb2009asel$marstat[eb2009asel$v641 > 3] <- 2 #Not living together
eb2009asel$marstat <- as.factor(eb2009asel$marstat)
table(eb2009asel$marstat, useNA = "always" )

#Missings
lapply(eb2009asel, table, useNA = "always")

mis_vars <- c("lrplace") 

for (var in mis_vars) {
  eb2009asel[is.na(eb2009asel[,var]), var] <- mean(eb2009asel[,var], na.rm = TRUE)
}

eb2009asel <- eb2009asel %>% select(v1, v8, ccpercept, cc_unstop, cc_exag, cc_poseu, cc_prsact, pers_imp, lrplace, eduyrs, isced, sex, age, urban, marstat)

save(eb2009asel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2009asel.Rdata")

eb2009_2 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb2009_2.sav", use.value.labels = T,  to.data.frame = T)


eb2009bsel <- eb2009_2 %>% filter(v6=="The Netherlands") %>%
  select(v1, v8, v385, v364, v363, v374, v392, v393, v394, v395, v396, v397, v398, v399, v387, v388, v389, v390, v391, v438, v439, v440, v441, v442, v445, v279)

# Dependent variables
eb2009bsel$ccpercept <- as.numeric(eb2009bsel$v385)
eb2009bsel$ccpercept <- (eb2009bsel$ccpercept -1) * (4/9) + 1

eb2009bsel$cchange <- NA
eb2009bsel$v364 <- as.numeric(eb2009bsel$v364)
eb2009bsel$cchange[eb2009bsel$v364==1] <- 1
eb2009bsel$cchange[eb2009bsel$v364!=1] <- 0
table(eb2009bsel$cchange, useNA = "always") 

eb2009bsel$cchange <- (eb2009bsel$cchange) * (4/1) + 1

eb2009bsel$cc_unstop <- as.numeric(eb2009bsel$v392)
eb2009bsel$cc_poseu <- as.numeric(eb2009bsel$v396)
eb2009bsel$cc_prsact <- as.numeric(eb2009bsel$v399)

eb2009bsel <- eb2009bsel %>% 
     mutate_at(c("cc_unstop", "cc_poseu", "cc_prsact"), funs(recode(., `1` = 4, `2` = 3, `3` = 2, `4` = 1)))

eb2009bsel$cc_unstop <- (eb2009bsel$cc_unstop -1) * (4/3) + 1
eb2009bsel$cc_poseu <- (eb2009bsel$cc_poseu -1) * (4/3) + 1
eb2009bsel$cc_prsact <- (eb2009bsel$cc_prsact -1) * (4/3) + 1

eb2009bsel$cc_exag <- as.numeric(eb2009bsel$v393)
eb2009bsel$cc_exag <- (eb2009bsel$cc_exag -1) * (4/3) + 1



attributes(eb2009bsel$v387) #Nat gov doing enough. Higher score means that they are not doing enough (1 is doing too much)
eb2009bsel$doprot_natgov <- as.numeric(eb2009bsel$v387)
eb2009bsel$doprot_natgov <- (eb2009bsel$doprot_natgov -1) * (4/2) + 1
attributes(eb2009bsel$v388) #European union doing enough
eb2009bsel$doprot_eu <- as.numeric(eb2009bsel$v387)
eb2009bsel$doprot_eu <- (eb2009bsel$doprot_eu -1) * (4/2) + 1
attributes(eb2009bsel$v389) #Reg/local gov doing enough
eb2009bsel$doprot_region <- as.numeric(eb2009bsel$v389)
eb2009bsel$doprot_region <- (eb2009bsel$doprot_natgov -1) * (4/2) + 1
attributes(eb2009bsel$v390) #Corporate/industry doing enough
eb2009bsel$doprot_comp <- as.numeric(eb2009bsel$v390)
eb2009bsel$doprot_comp <- (eb2009bsel$doprot_comp -1) * (4/2) + 1
attributes(eb2009bsel$v391) #Citizens doing enough
eb2009bsel$doprot_citiz <- as.numeric(eb2009bsel$v391)
eb2009bsel$doprot_citiz <- (eb2009bsel$doprot_citiz -1) * (4/2) + 1


#Independent variables 
table(eb2009bsel$v439, useNA = "always") #2 missings.
eb2009bsel$eduyrs <- as.numeric(as.character(eb2009bsel$v439)) #People who are still studying now have a missing
eb2009bsel$eduyrs <- eb2009bsel$eduyrs - 6

# I want the people that are still studying to have the mean educational years. Do this before categorizing them so that they can be in one of the isced categories
mean(eb2009bsel$eduyrs, na.rm=T)
eb2009bsel$eduyrs[is.na(eb2009bsel$eduyrs)] <- 14.393
eb2009bsel$isced[eb2009bsel$eduyrs <=4] <- 0
eb2009bsel$isced[eb2009bsel$eduyrs > 4 & eb2009bsel$eduyrs <= 6] <- 1
eb2009bsel$isced[eb2009bsel$eduyrs > 6 & eb2009bsel$eduyrs <= 10] <- 2
eb2009bsel$isced[eb2009bsel$eduyrs > 10 & eb2009bsel$eduyrs <= 13] <- 3
eb2009bsel$isced[eb2009bsel$eduyrs > 13 & eb2009bsel$eduyrs <= 15] <- 4
eb2009bsel$isced[eb2009bsel$eduyrs > 15 & eb2009bsel$eduyrs <= 18] <- 5
eb2009bsel$isced[eb2009bsel$eduyrs > 18] <- 6
table(eb2009bsel$isced, useNA = "always")

table(eb2009bsel$v441, useNA = "always") #Gender no missings
eb2009bsel$sex <- revalue(eb2009bsel$v441, c("Male"="1", "Female"="2"))
table(eb2009bsel$v442, useNA = "always")  #Age no missings
eb2009bsel$v442 <- revalue(eb2009bsel$v442, c("15 years"="15"))
eb2009bsel$age <- as.numeric(as.character(eb2009bsel$v442))
table(eb2009bsel$v445, useNA = "always") 
eb2009bsel$urban <- revalue(eb2009bsel$v445, c("Rural area or village"= "Low urbanity", "Small/middle town"="Medium urbanity", "Large town" = "High urbanity"))


table(eb2009bsel$v438, useNA = "always" )
eb2009bsel$v438 <- as.numeric(eb2009bsel$v438)
eb2009bsel$marstat[eb2009bsel$v438 <=2] <- 1 #Living together
eb2009bsel$marstat[eb2009bsel$v438 > 2] <- 2 #Not living together
eb2009bsel$marstat <- as.factor(eb2009bsel$marstat)
table(eb2009bsel$marstat, useNA = "always" )

#Missings
lapply(eb2009bsel, table, useNA = "always")

# no lrplace, no missings on other numeric independent variables

eb2009bsel <- eb2009bsel %>% select(v1, v8, ccpercept, cchange, cc_unstop, cc_exag, cc_poseu, cc_prsact, doprot_natgov:doprot_citiz, eduyrs, isced, sex, age, urban, marstat)

save(eb2009bsel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2009bsel.Rdata")

2011

#2011 consists of 2 waves as well
library(plyr)
eb2011_1 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb2011_1.sav", use.value.labels = T,  to.data.frame = T)

eb2011asel <- eb2011_1 %>% filter(v6=="The Netherlands") %>%
  select(v1, v8, v303, v304, v307, v308, v340:v343, v334:v336, v183, v309, v593, v597, v599, v601, v602, v609)

eb2011asel$env_quallife <- as.numeric(eb2011asel$v304)
eb2011asel$envp_eg <- as.numeric(eb2011asel$v307)
eb2011asel$effr_eg <- as.numeric(eb2011asel$v308)
eb2011asel$role_ind <- as.numeric(eb2011asel$v334)
eb2011asel$big_pol <- as.numeric(eb2011asel$v335)
eb2011asel$eff_daily <- as.numeric(eb2011asel$v336)
eb2011asel$pers_imp <- as.numeric(eb2011asel$v183)
eb2011asel$buyprod <- as.numeric(eb2011asel$v309)

eb2011asel <- eb2011asel %>% 
     mutate_at(c("env_quallife", "envp_eg", "effr_eg", "role_ind", "big_pol", "eff_daily", "pers_imp", "buyprod"), funs(recode(., `1` = 4, `2` = 3, `3` = 2, `4` = 1)))

eb2011asel <- eb2011asel %>% mutate_at(vars("env_quallife", "envp_eg", "effr_eg", "role_ind", "big_pol", "eff_daily", "pers_imp", "buyprod"), funs((. - 1)*(4/3) + 1))


# Independent variables 

table(eb2011asel$v593, useNA = "always") #57 missings l-r placement. 
eb2011asel$lrplace <- as.numeric(eb2011asel$v593)
table(eb2011asel$eduyrs, useNA = "always") 
eb2011asel$eduyrs <- as.numeric(as.character(eb2011asel$v599)) #Same about the individuals that are still studying

eb2011asel$eduyrs <- eb2011asel$eduyrs - 6

# I want the people that are still studying to have the mean educational years. Do this before categorizing them so that they can be in one of the isced categories
mean(eb2011asel$eduyrs, na.rm=T)
eb2011asel$eduyrs[is.na(eb2011asel$eduyrs)] <- 15.009
eb2011asel$isced[eb2011asel$eduyrs <=4] <- 0
eb2011asel$isced[eb2011asel$eduyrs > 4 & eb2011asel$eduyrs <= 6] <- 1
eb2011asel$isced[eb2011asel$eduyrs > 6 & eb2011asel$eduyrs <= 10] <- 2
eb2011asel$isced[eb2011asel$eduyrs > 10 & eb2011asel$eduyrs <= 13] <- 3
eb2011asel$isced[eb2011asel$eduyrs > 13 & eb2011asel$eduyrs <= 15] <- 4
eb2011asel$isced[eb2011asel$eduyrs > 15 & eb2011asel$eduyrs <= 18] <- 5
eb2011asel$isced[eb2011asel$eduyrs > 18] <- 6
table(eb2011asel$isced, useNA = "always")

table(eb2011asel$v601, useNA = "always") #No missings on gender
eb2011asel$sex <- revalue(eb2011asel$v601, c("Male"="1", "Female"="2"))
table(eb2011asel$v602, useNA = "always") #No missings on age
eb2011asel$v602 <- revalue(eb2011asel$v602, c("15 years"="15", "96 years" = "96"))
eb2011asel$age <- as.numeric(as.character(eb2011asel$v602))
table(eb2011asel$v609, useNA = "always") #No missings on urbanity
eb2011asel$urban <- revalue(eb2011asel$v609, c("Rural area or village"= "Low urbanity", "Small/middle town"="Medium urbanity", "Large town" = "High urbanity"))


table(eb2011asel$v597, useNA = "always" )
eb2011asel$v597 <- as.numeric(eb2011asel$v597)
eb2011asel$marstat[eb2011asel$v597 <=3] <- 1 #Living together
eb2011asel$marstat[eb2011asel$v597 > 3] <- 2 #Not living together
eb2011asel$marstat <- as.factor(eb2011asel$marstat)
table(eb2011asel$marstat, useNA = "always" )

#Missings
lapply(eb2011asel, table, useNA = "always")

mis_vars <- c("lrplace") 

for (var in mis_vars) {
  eb2011asel[is.na(eb2011asel[,var]), var] <- mean(eb2011asel[,var], na.rm = TRUE)
}


eb2011asel <- eb2011asel %>% select(v1, v8, env_quallife, envp_eg, effr_eg, role_ind, big_pol, eff_daily, pers_imp, buyprod, lrplace, eduyrs, isced, sex, age, urban, marstat)

save(eb2011asel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2011asel.Rdata")

eb2011_2 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb2011_2.sav", use.value.labels = T,  to.data.frame = T)

eb2011bsel <- eb2011_2 %>% filter(v6=="The Netherlands") %>% 
  select(v1, v8, v534, v535, v546, v557, v570, v589, v592, v594, v595, v602)

# Dependent variables 
eb2011bsel$v534 <- as.numeric(eb2011bsel$v534)
eb2011bsel$cchange[eb2011bsel$v534==1] <- 1
eb2011bsel$cchange[eb2011bsel$v534!=1] <- 0
table(eb2011bsel$cchange, useNA = "always")

eb2011bsel$v535 <- as.numeric(eb2011bsel$v535)
eb2011bsel$cchange2[eb2011bsel$v535==1] <- 1
eb2011bsel$cchange2[eb2011bsel$v535!=1] <- 0
table(eb2011bsel$cchange2, useNA = "always")

eb2011bsel$v546 <- as.numeric(eb2011bsel$v546)
eb2011bsel$cchangetot[eb2011bsel$v546==1] <- 1
eb2011bsel$cchangetot[eb2011bsel$v546!=1] <- 0
table(eb2011bsel$cchangetot, useNA = "always")

eb2011bsel$v570 <- as.numeric(eb2011bsel$v570)
eb2011bsel$prsaction[eb2011bsel$v570==1] <- 1
eb2011bsel$prsaction[eb2011bsel$v570==2] <- 0

eb2011bsel <- eb2011bsel %>% mutate_at(vars("cchange", "cchange2", "cchangetot", "prsaction"), funs((.)*(4/1) + 1))

eb2011bsel$ccpercept <- as.numeric(eb2011bsel$v557)
eb2011bsel$ccpercept <- (eb2011bsel$ccpercept -1) * (4/9) + 1

# Independent variables 
table(eb2011bsel$v592, useNA = "always") #2 missings on education. 
eb2011bsel$eduyrs <- as.numeric(as.character(eb2011bsel$v592)) #Same about individuals still studying
eb2011bsel$eduyrs <- eb2011bsel$eduyrs - 6

# I want the people that are still studying to have the mean educational years. Do this before categorizing them so that they can be in one of the isced categories
mean(eb2011bsel$eduyrs, na.rm=T)
eb2011bsel$eduyrs[is.na(eb2011bsel$eduyrs)] <- 14.833
eb2011bsel$isced[eb2011bsel$eduyrs <=4] <- 0
eb2011bsel$isced[eb2011bsel$eduyrs > 4 & eb2011bsel$eduyrs <= 6] <- 1
eb2011bsel$isced[eb2011bsel$eduyrs > 6 & eb2011bsel$eduyrs <= 10] <- 2
eb2011bsel$isced[eb2011bsel$eduyrs > 10 & eb2011bsel$eduyrs <= 13] <- 3
eb2011bsel$isced[eb2011bsel$eduyrs > 13 & eb2011bsel$eduyrs <= 15] <- 4
eb2011bsel$isced[eb2011bsel$eduyrs > 15 & eb2011bsel$eduyrs <= 18] <- 5
eb2011bsel$isced[eb2011bsel$eduyrs > 18] <- 6
table(eb2011bsel$isced, useNA = "always")

table(eb2011bsel$v594, useNA = "always") #no missings gender
eb2011bsel$sex <- revalue(eb2011bsel$v594, c("Male"="1", "Female"="2"))
table(eb2011bsel$v595, useNA = "always") #no missings age
eb2011bsel$v595 <- revalue(eb2011bsel$v595, c("15 years"="15", "98 years" = "98"))
eb2011bsel$age <- as.numeric(as.character(eb2011bsel$v595))
table(eb2011bsel$v602, useNA = "always") #no missing on urbanity
eb2011bsel$urban <- revalue(eb2011bsel$v602, c("Rural area or village"= "Low urbanity", "Small/middle town"="Medium urbanity", "Large town" = "High urbanity"))

table(eb2011bsel$v589, useNA = "always" )
eb2011bsel$v589 <- as.numeric(eb2011bsel$v589)
eb2011bsel$marstat[eb2011bsel$v589 <=3] <- 1 #Living together
eb2011bsel$marstat[eb2011bsel$v589 > 3] <- 2 #Not living together
eb2011bsel$marstat <- as.factor(eb2011bsel$marstat)
table(eb2011bsel$marstat, useNA = "always" )

#Missings
lapply(eb2011bsel, table, useNA = "always")
# Again no numeric variables left where i need to replace the missings

eb2011bsel <- eb2011bsel %>% select(v1, v8, cchange, cchange2, cchangetot, ccpercept, prsaction, eduyrs, isced, sex, age, urban, marstat)

save(eb2011bsel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2011bsel.Rdata")

2013

#2013 
eb2013 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb2013.sav", use.value.labels = T,  to.data.frame = T)


eb2013sel <- eb2013 %>% filter(country=="NL - The Netherlands") %>%
  select(studyno1, w1, qa1a, qa5, qa2, qa1b_1, qa1t_1, d10, d7r1, d8, d11, d25)

# Dependent variables 
eb2013sel$qa1a <- as.numeric(eb2013sel$qa1a)
eb2013sel$cchange[eb2013sel$qa1a==1] <- 1
eb2013sel$cchange [eb2013sel$qa1a!=1] <- 0

eb2013sel$qa1b_1 <- as.numeric(eb2013sel$qa1b_1)
eb2013sel$cchange2[eb2013sel$qa1b_1==1] <- 1
eb2013sel$cchange2[eb2013sel$qa1b_1!=1] <- 0

eb2013sel$qa1t_1 <- as.numeric(eb2013sel$qa1t_1)
eb2013sel$cchangetot[eb2013sel$qa1t_1==1] <- 1
eb2013sel$cchangetot[eb2013sel$qa1t_1!=1] <- 0

eb2013sel$qa5 <- as.numeric(eb2013sel$qa5)
eb2013sel$prsaction[eb2013sel$qa5==2] <- 0
eb2013sel$prsaction[eb2013sel$qa5==1] <- 1

eb2013sel <- eb2013sel %>% mutate_at(vars("cchange", "cchange2", "cchangetot", "prsaction"), funs((.)*(4/1) + 1))


eb2013sel$ccpercept <- as.numeric(eb2013sel$qa2)
eb2013sel$ccpercept <- (eb2013sel$ccpercept -1) * (4/9) + 1

# Independent variables 
table(eb2013sel$d10, useNA = "always") #no missings on gender
eb2013sel$sex <- revalue(eb2013sel$d10, c("Male"="1", "Female"="2"))
table(eb2013sel$d8, useNA = "always") #8 missings on educyrs
eb2013sel$eduyrs <- as.numeric(as.character(eb2013sel$d8)) #Same with resp still studying
eb2013sel$eduyrs <- eb2013sel$eduyrs - 6
mean(eb2013sel$eduyrs, na.rm=T)
eb2013sel$eduyrs[is.na(eb2013sel$eduyrs)] <- 14.753
eb2013sel$isced[eb2013sel$eduyrs <=4] <- 0
eb2013sel$isced[eb2013sel$eduyrs > 4 & eb2013sel$eduyrs <= 6] <- 1
eb2013sel$isced[eb2013sel$eduyrs > 6 & eb2013sel$eduyrs <= 10] <- 2
eb2013sel$isced[eb2013sel$eduyrs > 10 & eb2013sel$eduyrs <= 13] <- 3
eb2013sel$isced[eb2013sel$eduyrs > 13 & eb2013sel$eduyrs <= 15] <- 4
eb2013sel$isced[eb2013sel$eduyrs > 15 & eb2013sel$eduyrs <= 18] <- 5
eb2013sel$isced[eb2013sel$eduyrs > 18] <- 6
table(eb2013sel$isced, useNA = "always")

table(eb2013sel$d11, useNA = "always") #No missings on age
eb2013sel$d11 <- revalue(eb2013sel$d11, c("15 years"="15"))
eb2013sel$age <- as.numeric(as.character(eb2013sel$d11))
table(eb2013sel$d25, useNA = "always") #no missings on urbanity
eb2013sel$urban <- revalue(eb2013sel$d25, c("Rural area or village"= "Low urbanity", "Small/middle town"="Medium urbanity", "Large town" = "High urbanity"))

table(eb2013sel$d7r1, useNA = "always")
eb2013sel$d7r1 <- as.numeric(eb2013sel$d7r1)
eb2013sel$marstat[eb2013sel$d7r1 <=3] <- 1 #Living together
eb2013sel$marstat[eb2013sel$d7r1 > 3] <- 2 #Not living together
eb2013sel$marstat <- as.factor(eb2013sel$marstat)
table(eb2013sel$marstat, useNA = "always" )

#Missings
lapply(eb2013sel, table, useNA = "always")
# Again no numeric variables left where i need to replace the missings


eb2013sel <- eb2013sel %>% select(studyno1, w1, cchange, cchange2, cchangetot, ccpercept, prsaction, eduyrs, isced, sex, age, urban, marstat) 
#%>% 
  rename (v1 = studyno1, v8 = w1)

save(eb2013sel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2013sel.Rdata")

2014

eb2014 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb2014.sav", use.value.labels = T,  to.data.frame = T)

eb2014sel <- eb2014 %>% filter(country=="NL - The Netherlands") %>%
  select(studyno1, w1,qa9_1, qa9_2, qa7, qa8_1, qa16_1:qa16_6, qa13_1:qa13_3, qa1, qa10, d7r1, d8, d10, d11, d25)

eb2014sel$env_quallife <- as.numeric(eb2014sel$qa8_1)
eb2014sel$envp_eg <- as.numeric(eb2014sel$qa9_1)
eb2014sel$effr_eg <- as.numeric(eb2014sel$qa9_2)
eb2014sel$role_ind <- as.numeric(eb2014sel$qa13_1)
eb2014sel$big_pol <- as.numeric(eb2014sel$qa13_2)
eb2014sel$eff_daily <- as.numeric(eb2014sel$qa13_3)
eb2014sel$pers_imp <- as.numeric(eb2014sel$qa1)
eb2014sel$buyprod <- as.numeric(eb2014sel$qa10)

eb2014sel <- eb2014sel %>% 
     mutate_at(c("env_quallife", "envp_eg", "effr_eg", "role_ind", "big_pol", "eff_daily", "pers_imp", "buyprod"), funs(recode(., `1` = 4, `2` = 3, `3` = 2, `4` = 1)))

eb2014sel <- eb2014sel %>% mutate_at(vars("env_quallife", "envp_eg", "effr_eg", "role_ind", "big_pol", "eff_daily", "pers_imp", "buyprod"), funs((. - 1)*(4/3) + 1))

eb2014sel$doprot_comp <- as.numeric(eb2014sel$qa16_1)
eb2014sel$doprot_citiz <- as.numeric(eb2014sel$qa16_2)
eb2014sel$doprot_city <- as.numeric(eb2014sel$qa16_3)
eb2014sel$doprot_region <- as.numeric(eb2014sel$qa16_4)
eb2014sel$doprot_natgov <- as.numeric(eb2014sel$qa16_5)
eb2014sel$doprot_eu <- as.numeric(eb2014sel$qa16_6)

eb2014sel <- eb2014sel %>% mutate_at(vars("doprot_comp":"doprot_eu"), funs((. - 1)*(4/2) + 1))

# Independent variables 
table(eb2014sel$d8, useNA = "always") #1 missing on educ
eb2014sel$eduyrs <- as.numeric(as.character(eb2014sel$d8)) #same as above
eb2014sel$eduyrs <- eb2014sel$eduyrs - 6
mean(eb2014sel$eduyrs, na.rm=T)
eb2014sel$eduyrs[is.na(eb2014sel$eduyrs)] <- 14.753
eb2014sel$isced[eb2014sel$eduyrs <=4] <- 0
eb2014sel$isced[eb2014sel$eduyrs > 4 & eb2014sel$eduyrs <= 6] <- 1
eb2014sel$isced[eb2014sel$eduyrs > 6 & eb2014sel$eduyrs <= 10] <- 2
eb2014sel$isced[eb2014sel$eduyrs > 10 & eb2014sel$eduyrs <= 13] <- 3
eb2014sel$isced[eb2014sel$eduyrs > 13 & eb2014sel$eduyrs <= 15] <- 4
eb2014sel$isced[eb2014sel$eduyrs > 15 & eb2014sel$eduyrs <= 18] <- 5
eb2014sel$isced[eb2014sel$eduyrs > 18] <- 6
table(eb2014sel$isced, useNA = "always")
table(eb2014sel$d10, useNA = "always") #no missings gender
eb2014sel$sex <- revalue(eb2014sel$d10, c("Male"="1", "Female"="2"))
table(eb2014sel$d11, useNA = "always") 
eb2014sel$d11 <- revalue(eb2014sel$d11, c("15 years"="15"))
eb2014sel$age <- as.numeric(as.character(eb2014sel$d11))
table(eb2014sel$d25, useNA = "always") #no missings on urbanity
eb2014sel$urban <- revalue(eb2014sel$d25, c("Rural area or village"= "Low urbanity", "Small or medium-sized town"="Medium urbanity", "Large town/city" = "High urbanity"))

table(eb2014sel$d7r1, useNA = "always")
eb2014sel$d7r1 <- as.numeric(eb2014sel$d7r1)
eb2014sel$marstat[eb2014sel$d7r1 <=3] <- 1 #Living together
eb2014sel$marstat[eb2014sel$d7r1 > 3] <- 2 #Not living together
eb2014sel$marstat <- as.factor(eb2014sel$marstat)
table(eb2014sel$marstat, useNA = "always" )

# No missings that need substitution

eb2014sel <-  eb2014sel %>% select(studyno1, w1, envp_eg, effr_eg, role_ind, big_pol, eff_daily, env_quallife, doprot_comp:doprot_eu, pers_imp, buyprod, eduyrs, isced, sex, age, urban, marstat) 

save(eb2014sel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2014sel.Rdata")

2015

#2015
eb2015 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb2015.sav", use.value.labels = T,  to.data.frame = T)


eb2015sel <- eb2015 %>% filter(country=="NL - The Netherlands") %>%
  select(studyno1, w1, w3a4a, qa1a, qa1b_1, qa1t_1, qa5, qa2, d1, d7r1, d8, d10, d11, d25)

# Dependent variables 
eb2015sel$qa1a <- as.numeric(eb2015sel$qa1a)
eb2015sel$cchange[eb2015sel$qa1a==1] <- 1
eb2015sel$cchange [eb2015sel$qa1a!=1] <- 0

eb2015sel$qa1b_1 <- as.numeric(eb2015sel$qa1b_1)
eb2015sel$cchange2[eb2015sel$qa1b_1==1] <- 1
eb2015sel$cchange2[eb2015sel$qa1b_1!=1] <- 0

eb2015sel$qa1t_1 <- as.numeric(eb2015sel$qa1t_1)
eb2015sel$cchangetot[eb2015sel$qa1t_1==1] <- 1
eb2015sel$cchangetot[eb2015sel$qa1t_1!=1] <- 0

eb2015sel$qa5 <- as.numeric(eb2015sel$qa5)
eb2015sel$prsaction[eb2015sel$qa5==1] <- 1
eb2015sel$prsaction[eb2015sel$qa5==2] <- 0

eb2015sel <- eb2015sel %>% mutate_at(vars("cchange", "cchange2", "cchangetot", "prsaction"), funs((.)*(4/1) + 1))

eb2015sel$ccpercept <- as.numeric(eb2015sel$qa2)
eb2015sel$ccpercept <- (eb2015sel$ccpercept -1) * (4/9) + 1

# Independent variables
table(eb2015sel$d1, useNA = "always")
eb2015sel$lrplace <- as.numeric(eb2015sel$d1)
table(eb2015sel$d8, useNA = "always")
eb2015sel$eduyrs <- as.numeric(as.character(eb2015sel$d8)) #same as above
eb2015sel$eduyrs <- eb2015sel$eduyrs - 6
mean(eb2015sel$eduyrs, na.rm=T)
eb2015sel$eduyrs[is.na(eb2015sel$eduyrs)] <- 14.753
eb2015sel$isced[eb2015sel$eduyrs <=4] <- 0
eb2015sel$isced[eb2015sel$eduyrs > 4 & eb2015sel$eduyrs <= 6] <- 1
eb2015sel$isced[eb2015sel$eduyrs > 6 & eb2015sel$eduyrs <= 10] <- 2
eb2015sel$isced[eb2015sel$eduyrs > 10 & eb2015sel$eduyrs <= 13] <- 3
eb2015sel$isced[eb2015sel$eduyrs > 13 & eb2015sel$eduyrs <= 15] <- 4
eb2015sel$isced[eb2015sel$eduyrs > 15 & eb2015sel$eduyrs <= 18] <- 5
eb2015sel$isced[eb2015sel$eduyrs > 18] <- 6
table(eb2015sel$isced, useNA = "always")

table(eb2015sel$d10, useNA = "always") 
eb2015sel$sex <- revalue(eb2015sel$d10, c("Man"="1", "Woman"="2"))
table(eb2015sel$d11, useNA = "always") 
eb2015sel$d11 <- revalue(eb2015sel$d11, c("15 years"="15"))
eb2015sel$age <- as.numeric(as.character(eb2015sel$d11))
table(eb2015sel$d25, useNA = "always") 
eb2015sel$urban <- revalue(eb2015sel$d25, c("Rural area or village"= "Low urbanity", "Small or middle sized town"="Medium urbanity", "Large town" = "High urbanity"))

table(eb2015sel$d7r1, useNA = "always")
eb2015sel$d7r1 <- as.numeric(eb2015sel$d7r1)
eb2015sel$marstat[eb2015sel$d7r1 <=3] <- 1 #Living together
eb2015sel$marstat[eb2015sel$d7r1 > 3] <- 2 #Not living together
eb2015sel$marstat <- as.factor(eb2015sel$marstat)
table(eb2015sel$marstat, useNA = "always" )

# Missings
lapply(eb2015sel, table, useNA = "always")

mis_vars <- c("lrplace") 

for (var in mis_vars) {
  eb2015sel[is.na(eb2015sel[,var]), var] <- mean(eb2015sel[,var], na.rm = TRUE)
}

eb2015sel <- eb2015sel %>% select(studyno1, w1, cchange, cchange2, cchangetot, prsaction, ccpercept, eduyrs, isced, sex, age, lrplace, urban, marstat) 


save(eb2015sel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2015sel.Rdata")

2017

#2017 consists of 2 waves
eb2017_1 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb2017_1.sav", use.value.labels = T,  to.data.frame = T)

eb2017asel <- eb2017_1 %>% filter(country=="NL - The Netherlands") %>%
  select(studyno1, w1, qc1a, qc1b_1, qc1t_1, qc4_1, qc5,qc2, d1, d7r1, d8, d10, d11, d25)

# Dependent variables 
eb2017asel$qc1a <- as.numeric(eb2017asel$qc1a)
eb2017asel$cchange[eb2017asel$qc1a==1] <- 1
eb2017asel$cchange [eb2017asel$qc1a!=1] <- 0

eb2017asel$qc1b_1 <- as.numeric(eb2017asel$qc1b_1)
eb2017asel$cchange2[eb2017asel$qc1b_1==1] <- 1
eb2017asel$cchange2[eb2017asel$qc1b_1!=1] <- 0

eb2017asel$qc1t_1 <- as.numeric(eb2017asel$qc1t_1)
eb2017asel$cchangetot[eb2017asel$qc1t_1==1] <- 1
eb2017asel$cchangetot[eb2017asel$qc1t_1!=1] <- 0

eb2017asel$qc5 <- as.numeric(eb2017asel$qc5)
eb2017asel$prsaction[eb2017asel$qc5==1] <- 1
eb2017asel$prsaction[eb2017asel$qc5==2] <- 0

eb2017asel <- eb2017asel %>% mutate_at(vars("cchange", "cchange2", "cchangetot", "prsaction"), funs((.)*(4/1) + 1))

eb2017asel$ccpercept <- as.numeric(eb2017asel$qc2)
eb2017asel$ccpercept <- (eb2017asel$ccpercept -1) * (4/9) + 1

eb2017asel$cc_boost_growth <- as.numeric(eb2017asel$qc4_1)
eb2017asel <- eb2017asel %>% 
     mutate_at(c("cc_boost_growth"), funs(recode(., `1` = 4, `2` = 3, `3` = 2, `4` = 1)))

# Independent variables 
table(eb2017asel$d1, useNA = "always")
eb2017asel$lrplace <- as.numeric(eb2017asel$d1)
table(eb2017asel$d8, useNA = "always")
eb2017asel$d8 <- revalue(eb2017asel$d8, c("2 years"="2"))
eb2017asel$eduyrs <- as.numeric(as.character(eb2017asel$d8)) #same as above
eb2017asel$eduyrs <- eb2017asel$eduyrs - 6
mean(eb2017asel$eduyrs, na.rm=T)
eb2017asel$eduyrs[is.na(eb2017asel$eduyrs)] <- 15.160
eb2017asel$isced[eb2017asel$eduyrs <=4] <- 0
eb2017asel$isced[eb2017asel$eduyrs > 4 & eb2017asel$eduyrs <= 6] <- 1
eb2017asel$isced[eb2017asel$eduyrs > 6 & eb2017asel$eduyrs <= 10] <- 2
eb2017asel$isced[eb2017asel$eduyrs > 10 & eb2017asel$eduyrs <= 13] <- 3
eb2017asel$isced[eb2017asel$eduyrs > 13 & eb2017asel$eduyrs <= 15] <- 4
eb2017asel$isced[eb2017asel$eduyrs > 15 & eb2017asel$eduyrs <= 18] <- 5
eb2017asel$isced[eb2017asel$eduyrs > 18] <- 6
table(eb2017asel$isced, useNA = "always")

table(eb2017asel$d10, useNA = "always") 
eb2017asel$sex <- revalue(eb2017asel$d10, c("Man"="1", "Woman"="2"))
table(eb2017asel$d11, useNA = "always") 
eb2017asel$d11 <- revalue(eb2017asel$d11, c("15 years"="15"))
eb2017asel$age <- as.numeric(as.character(eb2017asel$d11))
table(eb2017asel$d25, useNA = "always") 
eb2017asel$urban <- revalue(eb2017asel$d25, c("Rural area or village"= "Low urbanity", "Small or middle sized town"="Medium urbanity", "Large town" = "High urbanity"))

table(eb2017asel$d7r1, useNA = "always")
eb2017asel$d7r1 <- as.numeric(eb2017asel$d7r1)
eb2017asel$marstat[eb2017asel$d7r1 <=3] <- 1 #Living together
eb2017asel$marstat[eb2017asel$d7r1 > 3] <- 2 #Not living together
eb2017asel$marstat <- as.factor(eb2017asel$marstat)
table(eb2017asel$marstat, useNA = "always" )

# Missings
lapply(eb2017asel, table, useNA = "always")

mis_vars <- c("lrplace") 

for (var in mis_vars) {
  eb2017asel[is.na(eb2017asel[,var]), var] <- mean(eb2017asel[,var], na.rm = TRUE)
}


eb2017asel <- eb2017asel %>% select(studyno1, w1, cchange, cchange2, cchangetot, prsaction, ccpercept, eduyrs, isced, sex, age, lrplace, urban, marstat) 
#%>% 
  #rename (v1 = studyno1, v8 = w1)

save(eb2017asel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2017asel.Rdata")

# 2nd wave of 2017 
eb2017_2 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb2017_2.sav", use.value.labels = T,  to.data.frame = T)

eb2017bsel <- eb2017_2 %>% filter(country=="NL - The Netherlands") %>%
  select(studyno1, w1, qd7_1:qd7_6, qd5_1:qd5_3, d1, d7r1, d8, d10, d11, d25)

# Dependent variables 
eb2017bsel$doprot_comp <- as.numeric(eb2017bsel$qd7_1)
eb2017bsel$doprot_citiz <- as.numeric(eb2017bsel$qd7_2)
eb2017bsel$doprot_city <- as.numeric(eb2017bsel$qd7_3)
eb2017bsel$doprot_region <- as.numeric(eb2017bsel$qd7_4)
eb2017bsel$doprot_natgov <- as.numeric(eb2017bsel$qd7_5)
eb2017bsel$doprot_eu <- as.numeric(eb2017bsel$qd7_6)

eb2017bsel <- eb2017bsel %>% mutate_at(vars("doprot_comp":"doprot_eu"), funs((. - 1)*(4/2) + 1))

eb2017bsel$role_ind <- as.numeric(eb2017bsel$qd5_1)
eb2017bsel$big_pol <- as.numeric(eb2017bsel$qd5_2)
eb2017bsel$eff_daily <- as.numeric(eb2017bsel$qd5_3)

eb2017bsel <- eb2017bsel %>% 
     mutate_at(c("role_ind", "big_pol", "eff_daily"), funs(recode(., `1` = 4, `2` = 3, `3` = 2, `4` = 1)))

eb2017bsel <- eb2017bsel %>% mutate_at(vars("role_ind", "big_pol", "eff_daily"), funs((. - 1)*(4/3) + 1))

# Independent variables 
table(eb2017bsel$d1, useNA = "always")
eb2017bsel$lrplace <- as.numeric(eb2017bsel$d1)
table(eb2017bsel$d8, useNA = "always")
eb2017bsel$eduyrs <- as.numeric(as.character(eb2017bsel$d8)) #same as above
eb2017bsel$eduyrs <- eb2017bsel$eduyrs - 6
mean(eb2017bsel$eduyrs, na.rm=T)
eb2017bsel$eduyrs[is.na(eb2017bsel$eduyrs)] <- 14.994
eb2017bsel$isced[eb2017bsel$eduyrs <=4] <- 0
eb2017bsel$isced[eb2017bsel$eduyrs > 4 & eb2017bsel$eduyrs <= 6] <- 1
eb2017bsel$isced[eb2017bsel$eduyrs > 6 & eb2017bsel$eduyrs <= 10] <- 2
eb2017bsel$isced[eb2017bsel$eduyrs > 10 & eb2017bsel$eduyrs <= 13] <- 3
eb2017bsel$isced[eb2017bsel$eduyrs > 13 & eb2017bsel$eduyrs <= 15] <- 4
eb2017bsel$isced[eb2017bsel$eduyrs > 15 & eb2017bsel$eduyrs <= 18] <- 5
eb2017bsel$isced[eb2017bsel$eduyrs > 18] <- 6
table(eb2017bsel$isced, useNA = "always")

table(eb2017bsel$d10, useNA = "always") 
eb2017bsel$sex <- revalue(eb2017bsel$d10, c("Man"="1", "Woman"="2"))
table(eb2017bsel$d11, useNA = "always") 
eb2017bsel$d11 <- revalue(eb2017bsel$d11, c("15 years"="15"))
eb2017bsel$age <- as.numeric(as.character(eb2017bsel$d11))
table(eb2017bsel$d25, useNA = "always") 
eb2017bsel$urban <- revalue(eb2017bsel$d25, c("Rural area or village"= "Low urbanity", "Small/middle town"="Medium urbanity", "Large town" = "High urbanity"))

table(eb2017bsel$d7r1, useNA = "always")
eb2017bsel$d7r1 <- as.numeric(eb2017bsel$d7r1)
eb2017bsel$marstat[eb2017bsel$d7r1 <=3] <- 1 #Living together
eb2017bsel$marstat[eb2017bsel$d7r1 > 3] <- 2 #Not living together
eb2017bsel$marstat <- as.factor(eb2017bsel$marstat)
table(eb2017bsel$marstat, useNA = "always" )

# Missings
lapply(eb2017bsel, table, useNA = "always")

mis_vars <- c("lrplace") 

for (var in mis_vars) {
  eb2017bsel[is.na(eb2017bsel[,var]), var] <- mean(eb2017bsel[,var], na.rm = TRUE)
}


eb2017bsel <- eb2017bsel %>% select(studyno1, w1, doprot_comp:doprot_eu, role_ind:eff_daily, eduyrs, isced, sex, age, urban, lrplace, marstat) 
#%>% 
 # rename(v1 = studyno1, v8 = w1)

save(eb2017bsel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2017bsel.Rdata")

2021

#Lastly, 2021

eb2021 <- foreign::read.spss("/Users/anuschka/Documents/gesis_dir/eurobarometer/eb2021.sav", use.value.labels = T,  to.data.frame = T)

eb2021sel <- eb2021 %>% filter(country=="NL - The Netherlands") %>%
  select(studyno1, w1, qb2, qb1a, qb1b.1, qb1t.1, d1, d7r, d8, d10, d11, d25 )

# Dependent variables 
eb2021sel$qb1a <- as.numeric(eb2021sel$qb1a)
eb2021sel$cchange[eb2021sel$qb1a==1] <- 1
eb2021sel$cchange [eb2021sel$qb1a!=1] <- 0

eb2021sel$qb1b.1 <- as.numeric(eb2021sel$qb1b.1)
eb2021sel$cchange2[eb2021sel$qb1b.1==1] <- 1
eb2021sel$cchange2[eb2021sel$qb1b.1!=1] <- 0

eb2021sel$qb1t.1 <- as.numeric(eb2021sel$qb1t.1)
eb2021sel$cchangetot[eb2021sel$qb1t.1==1] <- 1
eb2021sel$cchangetot[eb2021sel$qb1t.1!=1] <- 0

eb2021sel <- eb2021sel %>% mutate_at(vars("cchange", "cchange2", "cchangetot"), funs((.)*(4/1) + 1))

eb2021sel$ccpercept <- as.numeric(eb2021sel$qb2)
eb2021sel$ccpercept <- (eb2021sel$ccpercept -1) * (4/9) + 1


# Independent variables 
table(eb2021sel$d1, useNA = "always")
eb2021sel$lrplace <- as.numeric(eb2021sel$d1)
eb2021sel$lrplace[eb2021sel$lrplace==11 | eb2021sel$lrplace ==12] <- NA
table(eb2021sel$d8, useNA = "always")
eb2021sel$eduyrs <- as.numeric(as.character(eb2021sel$d8)) #same as above
eb2021sel$eduyrs <- eb2021sel$eduyrs - 6
mean(eb2021sel$eduyrs, na.rm=T)
eb2021sel$eduyrs[is.na(eb2021sel$eduyrs)] <- 14.994
eb2021sel$isced[eb2021sel$eduyrs <=4] <- 0
eb2021sel$isced[eb2021sel$eduyrs > 4 & eb2021sel$eduyrs <= 6] <- 1
eb2021sel$isced[eb2021sel$eduyrs > 6 & eb2021sel$eduyrs <= 10] <- 2
eb2021sel$isced[eb2021sel$eduyrs > 10 & eb2021sel$eduyrs <= 13] <- 3
eb2021sel$isced[eb2021sel$eduyrs > 13 & eb2021sel$eduyrs <= 15] <- 4
eb2021sel$isced[eb2021sel$eduyrs > 15 & eb2021sel$eduyrs <= 18] <- 5
eb2021sel$isced[eb2021sel$eduyrs > 18] <- 6
table(eb2021sel$isced, useNA = "always")

table(eb2021sel$d10, useNA = "always") 
eb2021sel$sex <- revalue(eb2021sel$d10, c("Man"="1", "Woman"="2"))
table(eb2021sel$d11, useNA = "always") 
eb2021sel$d11 <- revalue(eb2021sel$d11, c("15 years"="15"))
eb2021sel$age <- as.numeric(as.character(eb2021sel$d11))
table(eb2021sel$d25, useNA = "always") 
eb2021sel$urban <- revalue(eb2021sel$d25, c("Rural area or village"= "Low urbanity", "Small/middle town"="Medium urbanity", "Large town" = "High urbanity"))

table(eb2021sel$d7r, useNA = "always")
eb2021sel$d7r <- as.numeric(eb2021sel$d7r)
eb2021sel$marstat[eb2021sel$d7r <=3] <- 1 #Living together
eb2021sel$marstat[eb2021sel$d7r > 3] <- 2 #Not living together
eb2021sel$marstat <- as.factor(eb2021sel$marstat)
table(eb2021sel$marstat, useNA = "always" )

# Missings
lapply(eb2021sel, table, useNA = "always")

mis_vars <- c("lrplace") 

for (var in mis_vars) {
  eb2021sel[is.na(eb2021sel[,var]), var] <- mean(eb2021sel[,var], na.rm = TRUE)
}


eb2021sel <- eb2021sel %>% select(studyno1, w1, cchange, cchange2, cchangetot, ccpercept, eduyrs, isced, sex, age, urban, lrplace, marstat ) 
#%>% rename(v1 = studyno1, v8 = w1)

save(eb2021sel, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2021sel.Rdata")

Merge all the data into one dataset

#Create one large dataset in smaller steps. 
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb1986sel.Rdata")
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb1992sel.Rdata")
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb1995sel.Rdata")

eb1986sel$surveyyear <- 1986
eb1992sel$surveyyear <- 1992
eb1995sel$surveyyear <- 1995

ebu2000 <- plyr::rbind.fill(eb1986sel, eb1992sel, eb1995sel)

save(ebu2000, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/ebu2000.Rdata")

#Now the years under 2010
rm(list = ls())
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2004sel.Rdata")
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2007sel.Rdata")
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2008sel.Rdata")
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2009asel.Rdata")
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2009bsel.Rdata")

eb2004sel$surveyyear <- 2004
eb2007sel$surveyyear <- 2007
eb2008sel$surveyyear <- 2008
eb2009asel$surveyyear<- 2009
eb2009bsel$surveyyear <- 2009

eb0010 <- plyr::rbind.fill(eb2004sel, eb2007sel, eb2008sel, eb2009asel, eb2009bsel) #Looks how it should look. 

save(eb0010, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/ebu0010.Rdata")

rm(list=ls())

load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2011asel.Rdata")
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2011bsel.Rdata")
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2013sel.Rdata")
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2014sel.Rdata")
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2015sel.Rdata")

eb2011asel$surveyyear <- 2011
eb2011bsel$surveyyear <- 2011
eb2013sel$surveyyear <- 2013
eb2014sel$surveyyear <- 2014
eb2015sel$surveyyear <- 2015

ebtm15 <- plyr::rbind.fill(eb2011asel, eb2011bsel, eb2013sel, eb2014sel, eb2015sel)

save(ebtm15, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/ebutm15.Rdata")

#Now the last waves
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2017asel.Rdata")
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2017bsel.Rdata")
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb2021sel.Rdata")

eb2017asel$surveyyear <- 2017
eb2017bsel$surveyyear <- 2017 
eb2021sel$surveyyear <- 2021

eb17up <- plyr::rbind.fill(eb2017asel, eb2017bsel, eb2021sel)

save(eb17up, file="/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb17up.Rdata")

# Final merge 
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/ebu2000.Rdata")
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/ebu0010.Rdata")
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/ebutm15.Rdata")
load("/Users/anuschka/Documents/climatechange/climatechange/data/all_waves/eb17up.Rdata")

table(eb0010$surveyyear)
table(ebu2000$surveyyear)
table(ebtm15$surveyyear)
table(eb17up$surveyyear)

eb_tot <- plyr::rbind.fill(ebu2000, eb0010, ebtm15, eb17up)

table(eb_tot$surveyyear)

save(eb_tot, file="/Users/anuschka/Documents/climatechange/climatechange/data/final_data/eb_tot.Rdata")

library(plyr)
eb_tot$urban <- revalue(eb_tot$urban, c("1"= "Low urbanity", "2"="Medium urbanity", "3" = "High urbanity"))
eb_tot$urban <- factor(eb_tot$urban, levels=c("Low urbanity", "Medium urbanity", "High urbanity"), ordered=TRUE)
table(eb_tot$urban, useNA = "always")

# Change education/isced into three categories 
load("./data/final_data/eb_tot.Rdata")


table(eb_tot$isced)
eb_tot$isced_cat[eb_tot$isced <=2] <- "Basic"
eb_tot$isced_cat[eb_tot$isced == 3 | eb_tot$isced == 4] <- "Intermediate"
eb_tot$isced_cat[eb_tot$isced >=5] <- "Advanced"
eb_tot$isced_cat <- factor(eb_tot$isced_cat, levels=c("Basic", "Intermediate", "Advanced"), ordered=TRUE)
table(eb_tot$isced_cat)

#Code sex as binary
eb_tot$sex <- eb_tot$sex - 1
table(eb_tot$sex)

save(eb_tot, file="./data/final_data/eb_tot.Rdata")