Code
knitr::opts_chunk$set(warning = FALSE, message = FALSE)Exit Models
Code
library(sjPlot)
library(sjmisc)
library(sjlabelled)
library(lme4)
library(regclass)
library(stargazer)
library(dplyr)
library(tidyr)
library(effects)
library(lavaan)
library(ggplot2)
library(stringr)
library(labelled)Code
data <- read.csv("data_for_analysis_extended_17.4.csv", header = TRUE)
new_data <- read.csv("new_data.csv", header = TRUE)A Distribution of Perceptions of Civil Service Exit
Code
data <- read.csv("data_for_analysis_extended_17.4.csv", header = TRUE)
new_data <- read.csv("new_data.csv", header = TRUE)A Distribution of Perceptions of Civil Service Exit
A Distribution of Perceptions of Influence (Future)
Code
Q5 <- as.data.frame(
new_data %>%
group_by(Q5) %>%
filter(!is.na(Q5))%>%
summarize(N = n()))
Q5_merged <-
Q5 %>%
mutate(Q5 = case_when(Q5 %in% c("1","2") ~ "1",
Q5 %in% "3" ~ "2",
Q5 %in% c("4","5") ~ "3"))
Q5_new <- Q5_merged %>%
type.convert(as.is=TRUE) %>%
group_by(Q5) %>%
summarise(N=sum(N, na.rm = TRUE))%>%
mutate(freq = N / sum(N))%>% round(2)%>%
mutate(freq.lab=str_c(100*freq,"%"))
ggplot(Q5_new) +
geom_bar(aes(x=factor(Q5), y=freq), stat="identity")+
geom_text(aes(x=factor(Q5), y=freq, label = freq.lab),vjust=-0.3,size=4)+
scale_x_discrete(labels = c("החלשות והחלשות משמעותית", "ללא שינוי" ,"התחזקות והתחזקות משמעותית"))+
scale_y_continuous(labels = function(x) paste0(x*100, "%"), limits=c(0,1))+
theme(
text = element_text(family="Optima"),
strip.text = element_text(size=24),
axis.text = element_text(size =12),
axis.line = element_line(colour = "black"),
panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
legend.title=element_blank(),
legend.text=element_text(size=14),
legend.position='bottom',
legend.direction ='horizontal',
legend.key=element_blank(),
legend.margin = unit(0.2, "line"),
legend.key.height=unit(0.6,"line"),
plot.title=element_text(size=12,face="bold", hjust = 0.5),
axis.title.x=element_text(size=24),
axis.title.y=element_text(size=24))+
xlab("")+
ylab("")+
ggtitle("במבט קדימה על חמש השנים הבאות, באיזו מידה אתה צופה שתהיה התחזקות או החלשות \n ?במידת ההשפעה של הדרג המקצועי ביחידה שלך על מדיניות המשרד בתחום הפעילות של היחידה")
A Distribution of Perceptions of Meritiocracy (Future)
Code
Q11 <- as.data.frame(
new_data %>%
group_by(Q11) %>%
filter(!is.na(Q11)) %>%
summarize(N = n()))
Q11_merged <-
Q11 %>%
mutate(Q11 = case_when(Q11 %in% c("1","2") ~ "1",
Q11 %in% "3" ~ "2",
Q11 %in% c("4","5") ~ "3"))
Q11_new <- Q11_merged %>%
type.convert(as.is=TRUE) %>%
group_by(Q11) %>%
summarise(N=sum(N, na.rm = TRUE))%>%
mutate(freq = N / sum(N))%>% round(2)%>%
mutate(freq.lab=str_c(100*freq,"%"))
ggplot(Q11_new) +
geom_bar(aes(factor(x=Q11), y=freq), stat="identity")+
geom_text(aes(x=factor(Q11), y=freq, label = freq.lab),vjust=-0.3,size=4)+
scale_x_discrete(labels = c("שינוי לרעה ושינוי משמעותי לרעה", "ללא שינוי" ,"שינוי לטובה ושינוי משמעותי לטובה"))+
scale_y_continuous(labels = function(x) paste0(x*100, "%"), limits=c(0,1))+
theme(strip.text = element_text(size=24),
axis.text = element_text(size =12),
axis.line = element_line(colour = "black"),
panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
legend.title=element_blank(),
legend.text=element_text(size=20),
legend.position='bottom',
legend.direction ='horizontal',
legend.key=element_blank(),
legend.margin = unit(0.2, "line"),
legend.key.height=unit(0.6,"line"),
plot.title=element_text(size=12,face="bold", hjust = 0.5),
axis.title.x=element_text(size=24),
axis.title.y=element_text(size=24))+
xlab("")+
ylab("")+
ggtitle("במבט קדימה על חמש השנים הבאות, באיזו מידה אתה צופה שיהיה שינוי לטובה\n ?או לרעה במידה שבה קידומים במשרד ייעשו בפועל על בסיס ניסיון רלוונטי, יכולת ועבודה קשה")
Code
t1<-new_data %>% group_by(Q11) %>%
drop_na(Q11) %>%
summarize(n=n())%>%
mutate(freq=(n/sum(n))%>%round(2))%>%
mutate(freq.lab=str_c(100*freq,"%"))
t1%>%
#drop_na(trust_in_gov_lab)%>%
ggplot(aes(x=Q11 %>% as_factor(),y=n))+geom_col(width=0.75,alpha=0.8,fill="dodgerblue2")+
geom_text(aes(label=freq.lab),colour="gray20",hjust=-0.1,size=3)+
theme_classic()+
labs(title=var_label(new_data$Q11) %>% stringr::str_wrap(),x="",y="")+
coord_flip() +
theme_classic()+
theme(
plot.title = element_text(hjust = 1),
plot.subtitle = element_text(hjust = 1),
plot.caption = element_text(hjust = 0)
)
Code
ggplot(data = data,aes(Q26))+geom_histogram(binwidth = 1)+
scale_x_continuous(breaks = 1:5)+
labs(x="רצון לעזוב את השירות הציבורי לו היתה מוצעת משרה מחוץ לשירות הציבורי בשכר דומה (1 = נכונות נמוכה; 5 = נכונות גבוהה)",
y = "כמות משיבים")
Omit NA
Code
data_1 <- select(data, c(PAST_INFLUENCE, PAST_MERITOCRATIC,
PROJECT_MERITOCRATIC,BACKSLIDING_2,PROJECT_INFLUENCE,
INTENT_EXIT_2,position_type,ranking,tenure,ministry,age,gender,religiosity,education))
data_2 <- na.omit(data_1)Descriptive Statistics
# A tibble: 2 x 2
gender N
<chr> <int>
1 female 206
2 male 155# A tibble: 4 x 2
ranking N
<chr> <int>
1 junior 56
2 middle 180
3 senior 107
4 very senior 18# A tibble: 5 x 2
education N
<chr> <int>
1 bachelor 51
2 high-school 3
3 master 274
4 other 6
5 phd 27# A tibble: 5 x 2
tenure N
<chr> <int>
1 1- 10
2 1-5 103
3 11-20 91
4 20+ 61
5 6-10 96# A tibble: 4 x 2
position_type N
<chr> <int>
1 competitive tender 310
2 other 28
3 replacement 12
4 trust based 11Models
Code
m1 <- lme4::lmer(INTENT_EXIT_2 ~
tenure +
position_type +
ranking +
gender +
age +
factor(education)+
factor(religiosity) +
(1|ministry),
data_2)
m2 <- lme4::lmer(INTENT_EXIT_2 ~
BACKSLIDING_2+
tenure +
position_type +
ranking +
gender +
age +
factor(education)+
factor(religiosity) +
(1|ministry),
data_2)
m3 <- lme4::lmer(INTENT_EXIT_2 ~
PROJECT_INFLUENCE +
PAST_INFLUENCE+
tenure +
position_type +
ranking +
gender +
age +
factor(education)+
factor(religiosity) +
(1|ministry),
data_2)
m4 <- lme4::lmer(INTENT_EXIT_2 ~
PROJECT_MERITOCRATIC +
PAST_MERITOCRATIC +
tenure +
position_type +
ranking +
gender +
age +
factor(education)+
factor(religiosity) +
(1|ministry),
data_2)
m5 <- lme4::lmer(INTENT_EXIT_2 ~
PROJECT_INFLUENCE +
PROJECT_MERITOCRATIC +
PAST_MERITOCRATIC +
PAST_INFLUENCE+
tenure +
position_type +
ranking +
gender +
age +
factor(education)+
factor(religiosity) +
(1|ministry),
data_2)
m6 <- lme4::lmer(INTENT_EXIT_2 ~
BACKSLIDING_2+
PROJECT_INFLUENCE +
PROJECT_MERITOCRATIC+
PAST_INFLUENCE +
PAST_MERITOCRATIC +
tenure +
position_type +
ranking +
gender +
age +
factor(education)+
factor(religiosity) +
(1|ministry),
data_2)
m7 <- lme4::lmer(INTENT_EXIT_2 ~
BACKSLIDING_2+
PROJECT_MERITOCRATIC+
PAST_MERITOCRATIC +
tenure +
position_type +
ranking +
gender +
age +
factor(education)+
factor(religiosity) +
(1|ministry),
data_2)
#Predictions from model 6
data_2$religiosity <- as.factor(data_2$religiosity)
data_2$education <- as.factor(data_2$education)
m6 <- lme4::lmer(INTENT_EXIT_2 ~
BACKSLIDING_2+
PROJECT_INFLUENCE +
PROJECT_MERITOCRATIC+
PAST_INFLUENCE +
PAST_MERITOCRATIC +
tenure +
position_type +
ranking +
gender +
age +
education+
religiosity +
(1|ministry),
data_2)Plot intention to leave
Code
predict <- as.data.frame(effect("BACKSLIDING_2",m6))
predict <- rename(predict,Intention_to_leave ="fit")
predict <- rename(predict,democratic_backsliding ="BACKSLIDING_2")
ggplot(data=predict, aes(x=democratic_backsliding, y=Intention_to_leave))+
geom_pointrange(aes(ymin=Intention_to_leave-se, ymax=Intention_to_leave+se))+
scale_x_continuous(breaks = c(0,0.2, 0.5, 0.8,1))
Plot intention to leave
Code
predict <- as.data.frame(effect("PROJECT_MERITOCRATIC",m6))
predict <- rename(predict,Intention_to_leave ="fit")
predict <- rename(predict,expectation_of_future_meritocracy ="PROJECT_MERITOCRATIC")
ggplot(data=predict, aes(x=expectation_of_future_meritocracy, y=Intention_to_leave))+
geom_pointrange(aes(ymin=Intention_to_leave-se, ymax=Intention_to_leave+se))+
scale_x_continuous(breaks = c(0,0.2, 0.5, 0.8,1))
Regression table of all 7 models as html
#FE modelling without demographics and ministry affilliations (including regression table as html)