Chi tiêu giáo dục Đồng bằng sông Hồng_TNU JST
Hường Trịnh
3/27/2021
Đọc số liệu vào
Kết quả đăng trong Bài báo: "Determinant of household expenditure on education in Red River Delta region, TNU Journal of Science and Technology, 226(04): 53 - 61, 2021, Ngo Thi Ngoan*, Nguyen Thi Tuyet Mai, Dam Thi Thu Trang", link tại: http://jst.tnu.edu.vn/jst/ft/view/81
Dữ liệu download tại: https://drive.google.com/drive/folders/1EYAG-uYIFbuP6exA6PE2sywmMdxavWSS?usp=sharing
setwd("D:/Public paper/Education expenditure Vietnam/Data_DBSH_TNU")
require(reshape2)
require(tidyverse)
require(gridExtra)
source("FunctionEduc.R")
#---
load("GIAODUC18.RData")
load("CCCHIGD.RData")
GIAODUC18 <- GIAODUC18[which(GIAODUC18$THUBQ >=0), ]
source("DescriptionTableFunction.R")
GIAODUC18 <- GIAODUC18 %>% mutate(songuoidihoc = SONAMDANGHOC + SONUDANGHOC)
GIAODUC18 <- GIAODUC18 %>% mutate(NGHENGHIEP_CH = ifelse(LAMCONGANLUONG == "Yes", "LAMCONGANLUONG",
ifelse(NONGLAMTHUYSAN == "Yes", "NONGLAMTHUYSAN", "KINHDOANHDICHVU")))
GIAODUC18 <- GIAODUC18 %>% mutate(songuoidihocFactor = as.factor(songuoidihoc ))
levels(GIAODUC18$HONNHAN_CH) <- c("Other", "Married" ,
"Other", "Other",
"Other", "Other" )
#--add HOCTHEM variables
CCCHIGD_HH <- CCCHIGD %>% group_by(ID) %>% summarise(HOCTHEM = sum(HOCTHEM))
CCCHIGD_HH <- CCCHIGD_HH %>% mutate(HOCTHEM = ifelse(HOCTHEM > 0 , "Yes", "No"))
#GIAODUC18
GIAODUC18 <- GIAODUC18 [which(GIAODUC18 $CHIGD < 60000), ] #delete some outlier
GIAODUC18 <- left_join(GIAODUC18, CCCHIGD_HH)
GIAODUC18$HOCTHEM[is.na(GIAODUC18$HOCTHEM)] <- "No"
factor.name <- c("NOISONG", "DANTOC_CH", "HONNHAN_CH", "GIOITINH_CH",
"NGHENGHIEP_CH", "songuoidihocFactor",
"BANGCAP_CH", "HOCTHEM", "TROCAP")
conti.name <- c( "TUOI_CH", "TSNGUOI", "songuoidihoc" )
Des1 <- fun.DescriptionData(GIAODUC18, factor.name, conti.name)
Des1 <- fun.DescriptionData(GIAODUC18, factor.name, conti.name)
#by Province
Des1 <- fun.DescriptionData(GIAODUC18, factor.name, conti.name)
DesHanoi <- fun.DescriptionData(GIAODUC18 %>% filter(PROVINCE == "Hanoi"), factor.name, conti.name)
DesHaiPhong <- fun.DescriptionData(GIAODUC18 %>% filter(PROVINCE == "HaiPhong"), factor.name, conti.name)
DesVinhPhuc <- fun.DescriptionData(GIAODUC18 %>% filter(PROVINCE == "VinhPhuc"), factor.name, conti.name)
DesBacNinh <- fun.DescriptionData(GIAODUC18 %>% filter(PROVINCE == "BacNinh"), factor.name, conti.name)
DesQuangNinh <- fun.DescriptionData(GIAODUC18 %>% filter(PROVINCE == "QuangNinh"), factor.name, conti.name)
DesHaiDuong <- fun.DescriptionData(GIAODUC18 %>% filter(PROVINCE == "HaiDuong"), factor.name, conti.name)
DesHungYen <- fun.DescriptionData(GIAODUC18 %>% filter(PROVINCE == "HungYen"), factor.name, conti.name)
DesThaiBinh <- fun.DescriptionData(GIAODUC18 %>% filter(PROVINCE == "ThaiBinh"), factor.name, conti.name)
DesHaNam <- fun.DescriptionData(GIAODUC18 %>% filter(PROVINCE == "HaNam"), factor.name, conti.name)
DesNamDinh<- fun.DescriptionData(GIAODUC18 %>% filter(PROVINCE == "NamDinh"), factor.name, conti.name)
DesNinhBinh<- fun.DescriptionData(GIAODUC18 %>% filter(PROVINCE == "NinhBinh"), factor.name, conti.name)
setdiff( row.names(DesHaiPhong), row.names(DesHanoi))## character(0)DesHaiPhong <- DesHaiPhong[row.names(DesHanoi), ]
DesVinhPhuc <- DesVinhPhuc[row.names(DesHanoi), ]
DesBacNinh <- DesBacNinh[row.names(DesHanoi), ]
DesQuangNinh <- DesQuangNinh[row.names(DesHanoi), ]
DesHaiDuong<- DesHaiDuong[row.names(DesHanoi), ]
DesHungYen <- DesHungYen[row.names(DesHanoi), ]
DesThaiBinh <- DesThaiBinh[row.names(DesHanoi), ]
DesHaNam <- DesHaNam [row.names(DesHanoi), ]
DesNamDinh <- DesNamDinh[row.names(DesHanoi), ]
DesNinhBinh <- DesNinhBinh[row.names(DesHanoi), ]
DesALL <- data.frame(Hanoi = DesHanoi, VinhPhuc = DesVinhPhuc, BacNinh = DesBacNinh,
QuangNinh = DesQuangNinh, HaiDuong = DesHaiDuong,
HungYen = DesHungYen, ThaiBinh = DesThaiBinh,
Hanam = DesHaNam, Haiphong = DesHaiPhong ,
NamDinh = DesNamDinh, NinhBinh = DesNinhBinh)Bảng 2. Thông tin chung về đặc điểm hộ và chế độ học tập
DesALL## percen.factor VinhPhuc BacNinh
## So luong quan sat 306 134 137
## TUOI_CH 56.2 (13.1) 50.2 (12.7) 51.3 (11.3)
## TSNGUOI 3.8 (1.6) 4.1 (1.6) 4.2 (1.7)
## songuoidihoc 1 (1) 1.2 (1) 1.2 (1)
## NOISONG_URBAN 43.8 26.1 27.7
## NOISONG_RURAL 56.2 73.9 72.3
## DANTOC_CH_Kinh 99.7 94.8 100
## DANTOC_CH_Minority 0.3 5.2 <NA>
## HONNHAN_CH_Other 18 15.7 13.9
## HONNHAN_CH_Married 82 84.3 86.1
## GIOITINH_CH_Male 69.3 74.6 80.3
## GIOITINH_CH_Female 30.7 25.4 19.7
## NGHENGHIEP_CH_KINHDOANHDICHVU 50.7 25.4 34.3
## NGHENGHIEP_CH_LAMCONGANLUONG 32.7 42.5 42.3
## NGHENGHIEP_CH_NONGLAMTHUYSAN 16.7 32.1 23.4
## songuoidihocFactor_0 43.8 32.8 33.6
## songuoidihocFactor_1 22.9 22.4 23.4
## songuoidihocFactor_2 26.5 34.3 32.8
## songuoidihocFactor_3 5.9 10.4 10.2
## songuoidihocFactor_4 1 0 0
## BANGCAP_CH_No qualification 7.2 7.5 10.2
## BANGCAP_CH_Primary school 17 19.4 24.1
## BANGCAP_CH_Secondary-high school 57.2 62.7 58.4
## BANGCAP_CH_University 18.6 10.4 7.3
## HOCTHEM_No 63.1 45.5 69.3
## HOCTHEM_Yes 36.9 54.5 30.7
## TROCAP_Yes 21.2 44 25.5
## TROCAP_No 78.8 56 74.5
## QuangNinh HaiDuong HungYen ThaiBinh
## So luong quan sat 147 180 145 188
## TUOI_CH 53 (13.1) 55.1 (13.7) 56.6 (14.3) 57.2 (14.4)
## TSNGUOI 3.6 (1.4) 3.6 (1.6) 3.3 (1.6) 3 (1.5)
## songuoidihoc 0.9 (0.9) 0.8 (0.9) 0.8 (0.9) 0.7 (0.9)
## NOISONG_URBAN 59.2 21.7 11.7 9.6
## NOISONG_RURAL 40.8 78.3 88.3 90.4
## DANTOC_CH_Kinh 90.5 100 100 100
## DANTOC_CH_Minority 9.5 <NA> <NA> <NA>
## HONNHAN_CH_Other 16.3 16.7 22.8 19.1
## HONNHAN_CH_Married 83.7 83.3 77.2 80.9
## GIOITINH_CH_Male 83 81.1 78.6 81.9
## GIOITINH_CH_Female 17 18.9 21.4 18.1
## NGHENGHIEP_CH_KINHDOANHDICHVU 36.7 26.1 26.2 25.5
## NGHENGHIEP_CH_LAMCONGANLUONG 40.1 40.6 42.1 39.9
## NGHENGHIEP_CH_NONGLAMTHUYSAN 23.1 33.3 31.7 34.6
## songuoidihocFactor_0 45.6 51.1 53.1 56.4
## songuoidihocFactor_1 27.2 20.6 18.6 22.3
## songuoidihocFactor_2 23.8 23.9 26.2 17.6
## songuoidihocFactor_3 3.4 4.4 1.4 3.7
## songuoidihocFactor_4 0 0 0.7 0
## BANGCAP_CH_No qualification 12.9 5.6 11 4.3
## BANGCAP_CH_Primary school 18.4 15 11 10.6
## BANGCAP_CH_Secondary-high school 57.8 75 71.7 81.4
## BANGCAP_CH_University 10.9 4.4 6.2 3.7
## HOCTHEM_No 84.4 65 68.3 71.8
## HOCTHEM_Yes 15.6 35 31.7 28.2
## TROCAP_Yes 27.2 23.3 24.1 20.7
## TROCAP_No 72.8 76.7 75.9 79.3
## Hanam Haiphong NamDinh
## So luong quan sat 123 184 193
## TUOI_CH 56.4 (14.4) 53.1 (14.4) 57.4 (13.1)
## TSNGUOI 3.1 (1.3) 3.5 (1.5) 3.1 (1.5)
## songuoidihoc 0.8 (0.9) 1 (0.9) 0.7 (1)
## NOISONG_URBAN 12.2 46.2 18.7
## NOISONG_RURAL 87.8 53.8 81.3
## DANTOC_CH_Kinh 100 99.5 100
## DANTOC_CH_Minority <NA> 0.5 <NA>
## HONNHAN_CH_Other 22 21.2 23.8
## HONNHAN_CH_Married 78 78.8 76.2
## GIOITINH_CH_Male 79.7 72.8 77.2
## GIOITINH_CH_Female 20.3 27.2 22.8
## NGHENGHIEP_CH_KINHDOANHDICHVU 30.1 33.7 24.4
## NGHENGHIEP_CH_LAMCONGANLUONG 35 46.7 37.3
## NGHENGHIEP_CH_NONGLAMTHUYSAN 35 19.6 38.3
## songuoidihocFactor_0 48 41.3 58
## songuoidihocFactor_1 26.8 24.5 15
## songuoidihocFactor_2 21.1 32.1 21.2
## songuoidihocFactor_3 4.1 2.2 5.7
## songuoidihocFactor_4 0 0 0
## BANGCAP_CH_No qualification 12.2 6 11.4
## BANGCAP_CH_Primary school 13 19.6 16.6
## BANGCAP_CH_Secondary-high school 72.4 62 67.4
## BANGCAP_CH_University 2.4 12.5 4.7
## HOCTHEM_No 58.5 53.3 69.9
## HOCTHEM_Yes 41.5 46.7 30.1
## TROCAP_Yes 22.8 31.5 17.6
## TROCAP_No 77.2 68.5 82.4
## NinhBinh
## So luong quan sat 129
## TUOI_CH 51.1 (15.3)
## TSNGUOI 3.4 (1.4)
## songuoidihoc 1 (1)
## NOISONG_URBAN 20.9
## NOISONG_RURAL 79.1
## DANTOC_CH_Kinh 98.4
## DANTOC_CH_Minority 1.6
## HONNHAN_CH_Other 11.6
## HONNHAN_CH_Married 88.4
## GIOITINH_CH_Male 80.6
## GIOITINH_CH_Female 19.4
## NGHENGHIEP_CH_KINHDOANHDICHVU 19.4
## NGHENGHIEP_CH_LAMCONGANLUONG 50.4
## NGHENGHIEP_CH_NONGLAMTHUYSAN 30.2
## songuoidihocFactor_0 39.5
## songuoidihocFactor_1 27.9
## songuoidihocFactor_2 25.6
## songuoidihocFactor_3 7
## songuoidihocFactor_4 0
## BANGCAP_CH_No qualification 7
## BANGCAP_CH_Primary school 17.1
## BANGCAP_CH_Secondary-high school 70.5
## BANGCAP_CH_University 5.4
## HOCTHEM_No 56.6
## HOCTHEM_Yes 43.4
## TROCAP_Yes 34.1
## TROCAP_No 65.9Hình 1. Biểu đồ hộp về tỉ lệ chi tiêu giáo dục (%) và chi tiêu (nghìn đồng) của các hộ gia đình ở đồng bằng sông Hồng
p1 <- ggplot(GIAODUC18, aes(y = TONGCHIGD, x = PROVINCE)) +
geom_boxplot() +ylim (c( 0, 20))+
ylab("Phantram")+
stat_summary(fun.y=mean, colour="darkred", geom="point",
shape=18, size=3,show_guide = FALSE)
means <- aggregate(CHIGD ~ PROVINCE, GIAODUC18, mean)
p2 <- ggplot(GIAODUC18, aes(y = CHIGD, x = PROVINCE)) +
geom_boxplot() +ylim (c( 0, 36000)) + ylab("nghin dong")+
stat_summary(fun.y=mean, colour="darkred", geom="point",
shape=18, size=3,show_guide = FALSE)
grid.arrange(p1, p2, ncol = 1)
p2 <- ggplot( CCCHIGD, aes(x = PROVINCE, y = CHIALL, fill = m2xc6)) +
geom_boxplot(outlier.shape = NA) + ylim(0, 35000)+
ggtitle("Chi phi 1 nam hoc theo cap hoc va theo tinh")
p2 
sử dụng hình với màu đen và trắng
library(patternplot)
pattern.type <- c('hdashes', 'blank', 'crosshatch','hdashes', 'blank', 'crosshatch')
ggplot( CCCHIGD, aes(x = PROVINCE, y = CHIALL)) +
geom_boxplot(aes(fill=m2xc6, outlier.shape = NA)) + ylim(0, 35000)+
scale_fill_brewer(palette="OrRd")
#CCCHIGD$PROVINCE%in%c( "QuangNinh", "VinhPhuc", "BacNinh","HaiDuong")
tgc <- summarySE(CCCHIGD, measurevar="CHIALL", groupvars=c("PROVINCE","m2xc6"))
#tgc
# The errorbars overlapped, so use position_dodge to move them horizontally
pd <- position_dodge(0.5) # move them .05 to the left and right
p2N <- ggplot(tgc, aes(x = PROVINCE, y = CHIALL, shape=m2xc6)) + # colour = m2xc6,
geom_errorbar(aes(ymin=CHIALL-ci, ymax=CHIALL+ci), width=0.3, position=pd) +
geom_line(position=pd, size = 2) +
geom_point(position=pd, size = 3)
p2N
Tiếp tục xử lý chi tiêu theo từng khoản mục
#CCCHIGD %>% filter(TRAITUYEN >0 )
Muc2X.Chitieu2 <- CCCHIGD %>% select(ID, PROVINCE, HOCPHI, DONGGOP, QUANAO,
SGK, DUNGCU, HOCTHEM, CHIGDKHAC )
Muc2X.Chitieu2 <- melt(Muc2X.Chitieu2, id.vars = c("ID", "PROVINCE"))
ggplot(Muc2X.Chitieu2, aes(x = variable, y = value, fill = PROVINCE)) +
geom_boxplot(outlier.shape = NA) + ylim(0, 2000)
#----- Tinh ti le % cua chia tieu
Muc2X.Pct <- CCCHIGD %>% mutate( PHOCPHI = HOCPHI*100/CHIALL,
PDONGGOP= DONGGOP*100/CHIALL,
PQUANAO = QUANAO*100/CHIALL,
PSGK = SGK*100/CHIALL,
PDUNGCU = DUNGCU*100/CHIALL,
PHOCTHEM = HOCTHEM*100/CHIALL ) %>%
select(ID, PROVINCE, m2xc6, PHOCPHI ,PDONGGOP ,PQUANAO ,PSGK ,PDUNGCU, PHOCTHEM )
Muc2X.Pct <- Muc2X.Pct %>%
mutate(PCHIKHAC = 100 - PHOCPHI - PDONGGOP - PQUANAO - PSGK - PDUNGCU -PHOCTHEM)
Muc2X.Pct$PROVINCE <- as.factor(Muc2X.Pct$PROVINCE)
Muc2X.Pct$m2xc6 <- as.factor(Muc2X.Pct$m2xc6)
Muc2X.Pct <- Muc2X.Pct %>%dplyr::group_by( m2xc6, PROVINCE) %>%
dplyr::summarise(PHOCPHI = mean(PHOCPHI, na.rm = TRUE),
PDONGGOP = mean(PDONGGOP, na.rm = TRUE),
PQUANAO = mean(PQUANAO, na.rm = TRUE),
PSGK = mean(PSGK, na.rm = TRUE),
PDUNGCU = mean(PDUNGCU, na.rm = TRUE),
PHOCTHEM = mean(PHOCTHEM, na.rm = TRUE))
Muc2X.Pct <- melt(Muc2X.Pct, id.vars = c("m2xc6", "PROVINCE"))Hình với đen và trắng
p3 <- ggplot(data=Muc2X.Pct, aes(x = m2xc6, y = value, fill = variable)) +
geom_bar(stat="identity")+
facet_grid(cols = vars(PROVINCE))+
theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))+
xlab("Levels") + ylab("Percentage")+
# scale_fill_brewer(palette="Sequential")+
ggtitle("Ty le phan tram chi tieu cho tung khoan muc va cap hoc")
p3
#------ Vẽ lại và không sử dụng màu
# plotting settings -------------------------------------------------------
p3N <- ggplot(data = Muc2X.Pct, mapping = aes(x = PROVINCE, y = value, group = m2xc6, shape = variable)) +
# scale_x_continuous(breaks = seq(1960, 2000, by = 20)) +
theme_bw() +
#geom_line(aes( fill = m2xc6 ))+
# colored dots
geom_point( size = 1) +
facet_grid(cols = vars(PROVINCE))
# colored lines
#geom_line(data = Muc2X.Pct, aes(color = variable, group = variable)) +
#facet_grid(cols = vars(PROVINCE))
# weighted average black line
# geom_line(data = Muc2X.Pct, aes(x = m2xc6, y = value, size = 1)) +
#geom_point(data = Muc2X.Pct, aes(x = m2xc6, y = value, size = 1000))
p3N
Kết quả hồi quy
require(AER)
#DONE!!
GIAODUC18$HOCTHEM <- as.factor(GIAODUC18$HOCTHEM)
GIAODUC18$PROVINCE <- factor(GIAODUC18$PROVINCE, levels = c("Hanoi", "QuangNinh", "VinhPhuc", "BacNinh", "HaiDuong",
"HaiPhong", "HungYen", "ThaiBinh", "HaNam", "NamDinh",
"NinhBinh" ))
to6 <- tobit(CHIGD ~ THUBQ + BANGCAP_CH +GIOITINH_CH+TROCAP + songuoidihoc + HOCTHEM+NOISONG +PROVINCE ,
data = GIAODUC18, left = 0, dist = "gaussian") # Chạy Tobit Model.
#summary(to6)
to60 <- update(to6, . ~ 1)
#1 - as.vector(logLik(to6)/logLik(to60)) # Pseudo R2
Coef6 <- gen.model1(to6)
Coef6## return.Coef V2
## (Intercept) -16800.12*** (1839.22)
## THUBQ 0.28*** (0.1)
## BANGCAP_CHPrimary school 1320.77 (1415.12)
## BANGCAP_CHSecondary-high school 3041.38** (1277.59)
## BANGCAP_CHUniversity 5285.94*** (1595.82)
## GIOITINH_CHFemale -1920.18** (763.81)
## TROCAPNo 6632.19*** (739.53)
## songuoidihoc 12853.68*** (463.4)
## HOCTHEMYes 3832.87*** (765.42)
## NOISONGRURAL -4433.44*** (741.33)
## PROVINCEQuangNinh -2758.79** (1316.23)
## PROVINCEVinhPhuc -2648.55** (1288.13)
## PROVINCEBacNinh -104.9 (1265.32)
## PROVINCEHaiDuong -2604.71** (1253.63)
## PROVINCEHaiPhong 686.48 (1156.66)
## PROVINCEHungYen -953.59 (1344.4)
## PROVINCEThaiBinh -3815.28*** (1294.3)
## PROVINCEHaNam -2465.54* (1422.78)
## PROVINCENamDinh -1869.25 (1254.92)
## PROVINCENinhBinh -3610.89*** (1357.19)
## Log(scale) 9.28*** (0.02)