# List of packages
packages <- c("tidyverse", "infer", "fst", "modelsummary", "broom") # add any you need here
# Install packages if they aren't installed already
new_packages <- packages[!(packages %in% installed.packages()[,"Package"])]
if(length(new_packages)) install.packages(new_packages)
# Load the packages
lapply(packages, library, character.only = TRUE)## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr 1.1.3 ✔ readr 2.1.4
## ✔ forcats 1.0.0 ✔ stringr 1.5.0
## ✔ ggplot2 3.4.3 ✔ tibble 3.2.1
## ✔ lubridate 1.9.2 ✔ tidyr 1.3.0
## ✔ purrr 1.0.2
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors## [[1]]
## [1] "lubridate" "forcats" "stringr" "dplyr" "purrr" "readr"
## [7] "tidyr" "tibble" "ggplot2" "tidyverse" "stats" "graphics"
## [13] "grDevices" "utils" "datasets" "methods" "base"
##
## [[2]]
## [1] "infer" "lubridate" "forcats" "stringr" "dplyr" "purrr"
## [7] "readr" "tidyr" "tibble" "ggplot2" "tidyverse" "stats"
## [13] "graphics" "grDevices" "utils" "datasets" "methods" "base"
##
## [[3]]
## [1] "fst" "infer" "lubridate" "forcats" "stringr" "dplyr"
## [7] "purrr" "readr" "tidyr" "tibble" "ggplot2" "tidyverse"
## [13] "stats" "graphics" "grDevices" "utils" "datasets" "methods"
## [19] "base"
##
## [[4]]
## [1] "modelsummary" "fst" "infer" "lubridate" "forcats"
## [6] "stringr" "dplyr" "purrr" "readr" "tidyr"
## [11] "tibble" "ggplot2" "tidyverse" "stats" "graphics"
## [16] "grDevices" "utils" "datasets" "methods" "base"
##
## [[5]]
## [1] "broom" "modelsummary" "fst" "infer" "lubridate"
## [6] "forcats" "stringr" "dplyr" "purrr" "readr"
## [11] "tidyr" "tibble" "ggplot2" "tidyverse" "stats"
## [16] "graphics" "grDevices" "utils" "datasets" "methods"
## [21] "base"ess <- read_fst("All-ESS-Data.fst")Task 1：
belgium_data <- ess %>%
filter(cntry == "BE") %>%
mutate(trstep = ifelse(trstep %in% c(77, 88, 99), NA, trstep),
)unique(belgium_data$trstep)## [1] 0 7 8 5 6 NA 4 9 3 1 10 2belgium_data <- belgium_data %>% filter(!is.na(trstep))belgium_data <- ess %>%
filter(cntry == "BE") %>%
mutate(wrkprty = ifelse(wrkprty %in% c(7, 8, 9), NA, wrkprty),
)unique(belgium_data$wrkprty)## [1] 2 1 NAbelgium_data <- belgium_data %>% filter(!is.na(wrkprty))belgium_data <- belgium_data %>%
mutate(lrscale = ifelse(lrscale %in% c(77, 88, 99), NA, lrscale)) %>%
mutate(polintr = case_when(
polintr %in% c(1, 2) ~ "Interested",
polintr %in% c(3, 4) ~ "Not Interested",
polintr %in% c(7, 8, 9) ~ NA_character_,
TRUE ~ as.character(polintr)
)) %>%
mutate(wrkprty = case_when(
wrkprty == 1 ~ "Yes",
wrkprty == 2 ~ "No",
wrkprty %in% c(7, 8, 9) ~ NA_character_,
TRUE ~ as.character(wrkprty)
)) %>%
mutate(workcivil = case_when(
wrkorg == 1 ~ "Yes",
wrkorg == 2 ~ "No",
wrkorg %in% c(7, 8, 9) ~ NA_character_,
TRUE ~ as.character(wrkorg)
)) %>%
mutate(protest = case_when(
pbldmn == 1 ~ "Yes",
pbldmn == 2 ~ "No",
pbldmn %in% c(7, 8, 9) ~ NA_character_,
TRUE ~ as.character(pbldmn)
))
table(belgium_data$wrkprty)##
## No Yes
## 15354 738model1 <- lm(trstep ~ wrkprty, data = belgium_data)coefficients <- coef(model1)
print(coefficients)## (Intercept) wrkprtyYes
## 7.5491077 -0.9610318remotes::install_github("datalorax/equatiomatic")## Skipping install of 'equatiomatic' from a github remote, the SHA1 (29ff168f) has not changed since last install.
## Use `force = TRUE` to force installationequatiomatic::extract_eq(model1, use_coefs = TRUE)\[ \operatorname{\widehat{trstep}} = 7.55 - 0.96(\operatorname{wrkprty}_{\operatorname{Yes}}) \]
Interpretation: The intercept represents the estimated value of trust in the European Parliament, ranging from 0 to 10, when the explanatory variable is zero or, for categorical variables, at their reference level. In other words, when considering individuals who have not “worked in a political party” (thus indicating a response of “No”), the anticipated value of trstplt is approximately 7.5491077 (or rounded to 7.55).
wrkprtyYes -0.9610318： This coefficient represents the discrepancy in the predicted trust value towards the European Parliament between individuals who have worked in a political party (“Yes” category) and those who have not (“No” category). Specifically, being categorized as “Yes” (i.e., having prior experience in a political party) is associated with an approximate decrease of 0.9610318 (rounded to 0.96) in the predicted trust value towards the European Parliament compared to being categorized as “No” (i.e., lacking experience in a political party). Therefore, the estimated value for people who have had worked in a political party before will be 7.55 - 0.96, which after calculation is 6.59.
Regression Equation as follows: trstepˆ=7.5491077 - 0.9610318(wrkprtyYes)
y(hat) = intercept + B(hat)0 + B(hat)1 * x1 + epsilon
trstep(hat) = 7.5491077 - 0.9610318 (wrkprtyYes) + error term
Task 2：
bulgaria_data <- ess %>%
filter(cntry == "BG") %>%
mutate(stfdem = ifelse(stfdem %in% c(77, 88, 99), NA, stfdem),
)unique(bulgaria_data$stfdem)## [1] 0 1 NA 2 3 5 4 6 8 7 10 9bulgaria_data <- bulgaria_data %>% filter(!is.na(stfdem))bulgaria_data <- bulgaria_data %>%
mutate(
geo = recode(as.character(domicil),
'1' = "Urban",
'2' = "Peri-Urban",
'3' = "Rural",
'4' = "Rural",
'5' = "Rural",
'7' = NA_character_,
'8' = NA_character_,
'9' = NA_character_),
native = recode(brncntr,
`1` = "Yes",
`2` = "No",
`7` = NA_character_,
`8` = NA_character_,
`9` = NA_character_)
)
table(bulgaria_data$native)##
## No Yes
## 97 12239model2 <- lm(stfdem ~ native, data = bulgaria_data)coefficients <- coef(model2)
print(coefficients)## (Intercept) nativeYes
## 2.8041237 0.1189909summary(model2)##
## Call:
## lm(formula = stfdem ~ native, data = bulgaria_data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.9231 -1.9231 0.0769 2.0769 7.1959
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.8041 0.2265 12.382 <2e-16 ***
## nativeYes 0.1190 0.2274 0.523 0.601
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 2.231 on 12334 degrees of freedom
## (2 observations deleted due to missingness)
## Multiple R-squared: 2.22e-05, Adjusted R-squared: -5.887e-05
## F-statistic: 0.2739 on 1 and 12334 DF, p-value: 0.6007The expected average satisfaction with democracy for respondents that were not born in Bulgaria is 2.8041 ≈ 2.80
Task 3:
germany_data <- ess %>%
filter(cntry == "DE") %>%
mutate(trstlgl = ifelse(trstlgl %in% c(77, 88, 99), NA, trstlgl),
)unique(germany_data$trstlgl)## [1] 6 7 9 8 4 10 5 3 NA 2 0 1germany_data <- germany_data %>% filter(!is.na(trstlgl))germany_data <- germany_data %>%
mutate(
edulvla = case_when(
essround < 5 & edulvla == 55 ~ NA_real_,
TRUE ~ edulvla
),
edulvlb = case_when(
essround >= 5 & edulvlb == 5555 ~ NA_real_,
TRUE ~ edulvlb
),
edulvlfa = case_when(
essround < 5 & edulvlfa == 55 ~ NA_real_,
TRUE ~ edulvlfa
),
edulvlfb = case_when(
essround >= 5 & edulvlfb == 5555 ~ NA_real_,
TRUE ~ edulvlfb
),
edulvlma = case_when(
essround < 5 & edulvlma == 55 ~ NA_real_,
TRUE ~ edulvlma
),
edulvlmb = case_when(
essround >= 5 & edulvlmb == 5555 ~ NA_real_,
TRUE ~ edulvlmb
),
educ_level = case_when(
essround < 5 & edulvla == 5 ~ "BA",
essround >= 5 & edulvlb > 600 ~ "BA",
TRUE ~ "No BA"
),
educ_level_father = case_when(
essround < 5 & edulvlfa == 5 ~ "BA",
essround >= 5 & edulvlfb > 600 ~ "BA",
TRUE ~ "No BA"
),
educ_level_mother = case_when(
essround < 5 & edulvlma == 5 ~ "BA",
essround >= 5 & edulvlmb > 600 ~ "BA",
TRUE ~ "No BA"
),
eisced = case_when(
eisced == 0 | eisced > 7 ~ as.character(NA),
eisced %in% 1:5 ~ "No BA",
eisced == 6 ~ "BA",
eisced == 7 ~ "MA",
TRUE ~ as.character(eisced)
),
)
table(germany_data$educ_level)##
## BA No BA
## 8728 25239model3 <- lm(trstlgl ~ educ_level, data = germany_data)coefficients <- coef(model3)
print(coefficients)## (Intercept) educ_levelNo BA
## 6.415330 -0.813761remotes::install_github("datalorax/equatiomatic")## Skipping install of 'equatiomatic' from a github remote, the SHA1 (29ff168f) has not changed since last install.
## Use `force = TRUE` to force installationequatiomatic::extract_eq(model3, use_coefs = TRUE)\[ \operatorname{\widehat{trstlgl}} = 6.42 - 0.81(\operatorname{educ\_level}_{\operatorname{No\ BA}}) \]
Regression Equation as follows: trstlglˆ=6.42 - 0.81(educ_levelNoBA)
y(hat) = intercept + B(hat)0 + B(hat)1 * x1 + epsilon
trstlgl(hat) = 6.42 - 0.81 (educ_levelNoBA) + error term
Interpretation:
The intercept represents the estimated value of Trust in the legal system, specifically in the country of Germany, ranging from 0 to 10, when the explanatory variable is zero or, for categorical variables, at their reference level. In other words, when considering individuals who have “BA” (thus indicating that group has a BA degree), the anticipated value of trstlgl is approximately 6.415330 (or rounded to 6.42).
educ_levelNoBA -0.813761： This coefficient represents the discrepancy in the predicted trust value towards legal system between individuals who have attained a BA degree (“BA” category) and those who have not attained a BA degree (“No BA” category). Specifically, being categorized as “NO BA” (i.e., who have not attained a BA degree) is associated with an approximate decrease of 0.813761 (rounded to 0.81) in the predicted trust value towards the legal system in Germany compared to being categorized as “BA” (i.e., who have attained a BA degree). Therefore, the estimated value of Trust in the legal system for people who do not have a BA degree would be 6.42 - 0.81, which after calculation is 5.61.