Phân tích SEM với Dự án Ý Định hành vi khởi nghiệp, gói semPlot trong R
TRỊNH THỊ HƯỜNG -TMU
29/08/2021
Tài liệu phục vụ HỘI THẢO THỐNG KÊ TRONG KHOA HỌC XÃ HỘI VỚI PHẦN MỀM MÃ NGUỒN MỞ R Chi tiết tại: https://sites.google.com/view/tkud https://viasm.edu.vn/
GIỚI THIỆU
Dữ liệu sử dụng trong hướng dẫn này là của PGS.TS Trần Văn Trang (Đại học Thương mại).
Bài báo gốc lấy tại: http://tckhtm.tmu.edu.vn/vi/news/cac-so-tap-chi/tap-chi-khoa-hoc-thuong-mai-so-141-153.html
Dữ liệu tải tại google diver: https://drive.google.com/drive/folders/1Npip6h8WyZjI9JGf5wonnU_scBUYj4sP?usp=sharing
Load các gói cần thiết và nhập liệu
setwd("D:/Tap huan VIASM/HoiThao_KHXH_2021/Projects/Y_Dinh_Hanh_Vi")
library(foreign)
require(tidyverse)
require(lavaan)
require(semPlot)
d <- read.spss("Case study Behavior Intention.sav",
use.value.label=TRUE, to.data.frame=TRUE)1. Lọc các biến quan sát (items)
d1 <- d %>% select(-c("STT", "FAM", "Formation",
"Work", "Year",
"BI1", "BIRecode", "BI4"))2. Tìm hiểu các hàm sem trong gói semPlot tại:
?sem3.Fit model3 và xem các ước lượng, chỉ số mô hình
sem.model3 <- ' BEINTEN =~ BI2 + BI3 + BI5 + BI6
RELA =~ REL1 + REL2 + REL3+ REL4
EDUC =~ EDU1 + EDU2 + EDU3 + EDU4 + EDU5 + EDU6+EDU7+EDU8
GOVE =~ GOV1 + GOV2 + GOV3 + GOV4 + GOV5
ENDO =~ END2 + END3 + END4 + END5
# regression
BEINTEN ~ RELA + EDUC + GOVE + ENDO
# Covariance
BI2 ~~ BI3
REL4 ~~ REL1 + REL3
EDU2 ~~ EDU1 + EDU3 + EDU6
EDU5 ~~ EDU6
EDU7 ~~ EDU8
GOV4 ~~ GOV5
END5 ~~ END4 '
fitsem3 <- sem(sem.model3, data = d1)
summary(fitsem3)## lavaan 0.6-8 ended normally after 45 iterations
##
## Estimator ML
## Optimization method NLMINB
## Number of model parameters 70
##
## Number of observations 826
##
## Model Test User Model:
##
## Test statistic 756.302
## Degrees of freedom 255
## P-value (Chi-square) 0.000
##
## Parameter Estimates:
##
## Standard errors Standard
## Information Expected
## Information saturated (h1) model Structured
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|)
## BEINTEN =~
## BI2 1.000
## BI3 1.060 0.064 16.495 0.000
## BI5 1.307 0.081 16.225 0.000
## BI6 1.335 0.085 15.711 0.000
## RELA =~
## REL1 1.000
## REL2 0.987 0.043 23.040 0.000
## REL3 0.723 0.038 18.864 0.000
## REL4 0.626 0.042 14.854 0.000
## EDUC =~
## EDU1 1.000
## EDU2 0.994 0.044 22.763 0.000
## EDU3 1.081 0.054 19.941 0.000
## EDU4 1.154 0.057 20.186 0.000
## EDU5 1.026 0.051 20.014 0.000
## EDU6 1.005 0.058 17.269 0.000
## EDU7 1.032 0.052 19.782 0.000
## EDU8 0.875 0.049 17.707 0.000
## GOVE =~
## GOV1 1.000
## GOV2 1.099 0.052 21.291 0.000
## GOV3 1.099 0.052 21.326 0.000
## GOV4 0.988 0.052 19.126 0.000
## GOV5 0.813 0.051 15.936 0.000
## ENDO =~
## END2 1.000
## END3 1.044 0.051 20.636 0.000
## END4 0.789 0.045 17.409 0.000
## END5 0.891 0.055 16.309 0.000
##
## Regressions:
## Estimate Std.Err z-value P(>|z|)
## BEINTEN ~
## RELA 0.373 0.033 11.480 0.000
## EDUC 0.079 0.042 1.894 0.058
## GOVE 0.089 0.043 2.046 0.041
## ENDO -0.064 0.031 -2.039 0.041
##
## Covariances:
## Estimate Std.Err z-value P(>|z|)
## .BI2 ~~
## .BI3 0.243 0.046 5.320 0.000
## .REL1 ~~
## .REL4 -0.153 0.045 -3.415 0.001
## .REL3 ~~
## .REL4 0.445 0.056 7.969 0.000
## .EDU1 ~~
## .EDU2 0.238 0.030 7.949 0.000
## .EDU2 ~~
## .EDU3 0.230 0.027 8.577 0.000
## .EDU6 -0.052 0.025 -2.054 0.040
## .EDU5 ~~
## .EDU6 0.125 0.031 3.991 0.000
## .EDU7 ~~
## .EDU8 0.194 0.028 6.979 0.000
## .GOV4 ~~
## .GOV5 0.228 0.030 7.525 0.000
## .END4 ~~
## .END5 0.345 0.045 7.752 0.000
## RELA ~~
## EDUC 0.520 0.056 9.368 0.000
## GOVE 0.423 0.050 8.451 0.000
## ENDO -0.040 0.051 -0.787 0.432
## EDUC ~~
## GOVE 0.450 0.042 10.792 0.000
## ENDO 0.213 0.037 5.722 0.000
## GOVE ~~
## ENDO 0.188 0.035 5.406 0.000
##
## Variances:
## Estimate Std.Err z-value P(>|z|)
## .BI2 0.925 0.055 16.787 0.000
## .BI3 1.128 0.066 17.030 0.000
## .BI5 0.703 0.054 12.950 0.000
## .BI6 1.012 0.068 14.973 0.000
## .REL1 0.755 0.064 11.788 0.000
## .REL2 0.816 0.064 12.767 0.000
## .REL3 1.113 0.064 17.449 0.000
## .REL4 1.301 0.076 17.142 0.000
## .EDU1 0.998 0.054 18.406 0.000
## .EDU2 0.627 0.036 17.333 0.000
## .EDU3 0.594 0.036 16.519 0.000
## .EDU4 0.629 0.039 16.322 0.000
## .EDU5 0.517 0.032 16.241 0.000
## .EDU6 1.012 0.056 18.069 0.000
## .EDU7 0.565 0.034 16.627 0.000
## .EDU8 0.701 0.039 18.024 0.000
## .GOV1 0.743 0.042 17.561 0.000
## .GOV2 0.393 0.028 13.847 0.000
## .GOV3 0.388 0.028 13.754 0.000
## .GOV4 0.602 0.036 16.877 0.000
## .GOV5 0.814 0.044 18.468 0.000
## .END2 0.596 0.045 13.181 0.000
## .END3 0.351 0.041 8.561 0.000
## .END4 0.806 0.047 17.260 0.000
## .END5 1.254 0.071 17.781 0.000
## .BEINTEN 0.319 0.039 8.141 0.000
## RELA 1.661 0.124 13.350 0.000
## EDUC 0.834 0.079 10.555 0.000
## GOVE 0.721 0.066 10.983 0.000
## ENDO 0.914 0.076 11.955 0.000coef(fitsem3)## BEINTEN=~BI3 BEINTEN=~BI5 BEINTEN=~BI6 RELA=~REL2
## 1.060 1.307 1.335 0.987
## RELA=~REL3 RELA=~REL4 EDUC=~EDU2 EDUC=~EDU3
## 0.723 0.626 0.994 1.081
## EDUC=~EDU4 EDUC=~EDU5 EDUC=~EDU6 EDUC=~EDU7
## 1.154 1.026 1.005 1.032
## EDUC=~EDU8 GOVE=~GOV2 GOVE=~GOV3 GOVE=~GOV4
## 0.875 1.099 1.099 0.988
## GOVE=~GOV5 ENDO=~END3 ENDO=~END4 ENDO=~END5
## 0.813 1.044 0.789 0.891
## BEINTEN~RELA BEINTEN~EDUC BEINTEN~GOVE BEINTEN~ENDO
## 0.373 0.079 0.089 -0.064
## BI2~~BI3 REL1~~REL4 REL3~~REL4 EDU1~~EDU2
## 0.243 -0.153 0.445 0.238
## EDU2~~EDU3 EDU2~~EDU6 EDU5~~EDU6 EDU7~~EDU8
## 0.230 -0.052 0.125 0.194
## GOV4~~GOV5 END4~~END5 BI2~~BI2 BI3~~BI3
## 0.228 0.345 0.925 1.128
## BI5~~BI5 BI6~~BI6 REL1~~REL1 REL2~~REL2
## 0.703 1.012 0.755 0.816
## REL3~~REL3 REL4~~REL4 EDU1~~EDU1 EDU2~~EDU2
## 1.113 1.301 0.998 0.627
## EDU3~~EDU3 EDU4~~EDU4 EDU5~~EDU5 EDU6~~EDU6
## 0.594 0.629 0.517 1.012
## EDU7~~EDU7 EDU8~~EDU8 GOV1~~GOV1 GOV2~~GOV2
## 0.565 0.701 0.743 0.393
## GOV3~~GOV3 GOV4~~GOV4 GOV5~~GOV5 END2~~END2
## 0.388 0.602 0.814 0.596
## END3~~END3 END4~~END4 END5~~END5 BEINTEN~~BEINTEN
## 0.351 0.806 1.254 0.319
## RELA~~RELA EDUC~~EDUC GOVE~~GOVE ENDO~~ENDO
## 1.661 0.834 0.721 0.914
## RELA~~EDUC RELA~~GOVE RELA~~ENDO EDUC~~GOVE
## 0.520 0.423 -0.040 0.450
## EDUC~~ENDO GOVE~~ENDO
## 0.213 0.188fitted(fitsem3)## $cov
## BI2 BI3 BI5 BI6 REL1 REL2 REL3 REL4 EDU1 EDU2
## BI2 1.553
## BI3 0.908 1.833
## BI5 0.820 0.869 1.775
## BI6 0.838 0.888 1.095 2.130
## REL1 0.701 0.743 0.916 0.935 2.417
## REL2 0.692 0.733 0.904 0.923 1.639 2.434
## REL3 0.507 0.537 0.662 0.676 1.201 1.185 1.981
## REL4 0.439 0.465 0.573 0.586 0.887 1.026 1.197 1.952
## EDU1 0.286 0.303 0.374 0.382 0.520 0.513 0.376 0.326 1.831
## EDU2 0.284 0.301 0.371 0.379 0.517 0.510 0.374 0.324 1.067 1.451
## EDU3 0.309 0.328 0.404 0.412 0.562 0.555 0.406 0.352 0.901 1.125
## EDU4 0.330 0.350 0.431 0.440 0.600 0.592 0.434 0.376 0.962 0.956
## EDU5 0.293 0.311 0.383 0.392 0.534 0.527 0.386 0.334 0.856 0.850
## EDU6 0.287 0.305 0.375 0.384 0.523 0.516 0.378 0.327 0.838 0.781
## EDU7 0.295 0.313 0.385 0.394 0.537 0.530 0.388 0.336 0.860 0.855
## EDU8 0.250 0.265 0.327 0.334 0.455 0.449 0.329 0.285 0.730 0.725
## GOV1 0.245 0.260 0.320 0.327 0.423 0.418 0.306 0.265 0.450 0.447
## GOV2 0.269 0.285 0.352 0.359 0.465 0.459 0.336 0.291 0.495 0.491
## GOV3 0.269 0.286 0.352 0.360 0.465 0.459 0.336 0.291 0.495 0.492
## GOV4 0.242 0.257 0.316 0.323 0.418 0.412 0.302 0.262 0.445 0.442
## GOV5 0.199 0.211 0.260 0.266 0.344 0.340 0.249 0.215 0.366 0.364
## END2 -0.040 -0.042 -0.052 -0.053 -0.040 -0.040 -0.029 -0.025 0.213 0.212
## END3 -0.042 -0.044 -0.055 -0.056 -0.042 -0.041 -0.030 -0.026 0.223 0.221
## END4 -0.032 -0.033 -0.041 -0.042 -0.032 -0.031 -0.023 -0.020 0.168 0.167
## END5 -0.036 -0.038 -0.047 -0.048 -0.036 -0.035 -0.026 -0.022 0.190 0.189
## EDU3 EDU4 EDU5 EDU6 EDU7 EDU8 GOV1 GOV2 GOV3 GOV4
## BI2
## BI3
## BI5
## BI6
## REL1
## REL2
## REL3
## REL4
## EDU1
## EDU2
## EDU3 1.568
## EDU4 1.040 1.738
## EDU5 0.925 0.987 1.395
## EDU6 0.906 0.967 0.985 1.855
## EDU7 0.929 0.992 0.883 0.864 1.452
## EDU8 0.789 0.842 0.749 0.733 0.947 1.340
## GOV1 0.486 0.519 0.462 0.452 0.464 0.394 1.464
## GOV2 0.535 0.571 0.508 0.497 0.510 0.433 0.792 1.264
## GOV3 0.535 0.571 0.508 0.497 0.510 0.433 0.792 0.871 1.259
## GOV4 0.480 0.513 0.456 0.447 0.459 0.389 0.712 0.782 0.783 1.305
## GOV5 0.396 0.422 0.376 0.368 0.378 0.320 0.586 0.644 0.644 0.807
## END2 0.231 0.246 0.219 0.215 0.220 0.187 0.188 0.207 0.207 0.186
## END3 0.241 0.257 0.229 0.224 0.230 0.195 0.196 0.216 0.216 0.194
## END4 0.182 0.194 0.173 0.169 0.174 0.147 0.148 0.163 0.163 0.147
## END5 0.205 0.219 0.195 0.191 0.196 0.166 0.168 0.184 0.184 0.165
## GOV5 END2 END3 END4 END5
## BI2
## BI3
## BI5
## BI6
## REL1
## REL2
## REL3
## REL4
## EDU1
## EDU2
## EDU3
## EDU4
## EDU5
## EDU6
## EDU7
## EDU8
## GOV1
## GOV2
## GOV3
## GOV4
## GOV5 1.291
## END2 0.153 1.511
## END3 0.160 0.955 1.348
## END4 0.121 0.721 0.753 1.376
## END5 0.136 0.815 0.851 0.988 1.979show(fitsem3)## lavaan 0.6-8 ended normally after 45 iterations
##
## Estimator ML
## Optimization method NLMINB
## Number of model parameters 70
##
## Number of observations 826
##
## Model Test User Model:
##
## Test statistic 756.302
## Degrees of freedom 255
## P-value (Chi-square) 0.000logLik(fitsem3)## 'log Lik.' -29226.47 (df=70)fitMeasures(fitsem3, c("chisq", "df", "pvalue", "gfi","cfi","tli", "rmsea"))## chisq df pvalue gfi cfi tli rmsea
## 756.302 255.000 0.000 0.929 0.953 0.945 0.049fitMeasures(fitsem3,("chisq"))/fitMeasures(fitsem3,("df"))## chisq
## 2.966fitMeasures(fitsem3, fit.measures = "all")## npar fmin chisq df
## 70.000 0.458 756.302 255.000
## pvalue baseline.chisq baseline.df baseline.pvalue
## 0.000 10980.751 300.000 0.000
## cfi tli nnfi rfi
## 0.953 0.945 0.945 0.919
## nfi pnfi ifi rni
## 0.931 0.791 0.953 0.953
## logl unrestricted.logl aic bic
## -29226.467 -28848.316 58592.934 58923.096
## ntotal bic2 rmsea rmsea.ci.lower
## 826.000 58700.801 0.049 0.045
## rmsea.ci.upper rmsea.pvalue rmr rmr_nomean
## 0.053 0.684 0.087 0.087
## srmr srmr_bentler srmr_bentler_nomean crmr
## 0.052 0.052 0.052 0.054
## crmr_nomean srmr_mplus srmr_mplus_nomean cn_05
## 0.054 0.052 0.052 321.272
## cn_01 gfi agfi pgfi
## 340.068 0.929 0.910 0.729
## mfi ecvi
## 0.738 1.0854. Fit model4 và xem các ước lượng, chỉ số mô hình
#4.1.3 == SEM with latent + Control variables=================================
sem.model4 <- ' BEINTEN =~ BI2 + BI3 + BI5 + BI6
RELA =~ REL1 + REL2 + REL3+ REL4
EDUC =~ EDU1 + EDU2 + EDU3 + EDU4 + EDU5 + EDU6 + EDU7 + EDU8
GOVE =~ GOV1 + GOV2 + GOV3 + GOV4 + GOV5
ENDO =~ END2 + END3 + END4 + END5
# regression
BEINTEN ~ RELA + EDUC + GOVE + ENDO + FAM + Formation + Work + Year
# Covariance
BI2 ~~ BI3
REL4 ~~ REL1 + REL3
EDU2 ~~ EDU1 + EDU3 + EDU6
EDU5 ~~ EDU6
EDU7 ~~ EDU8
GOV4 ~~ GOV5
END5 ~~ END4 '
fitsem4 <- sem(sem.model4, data = d)
summary(fitsem4)## lavaan 0.6-8 ended normally after 45 iterations
##
## Estimator ML
## Optimization method NLMINB
## Number of model parameters 74
##
## Number of observations 826
##
## Model Test User Model:
##
## Test statistic 965.013
## Degrees of freedom 351
## P-value (Chi-square) 0.000
##
## Parameter Estimates:
##
## Standard errors Standard
## Information Expected
## Information saturated (h1) model Structured
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|)
## BEINTEN =~
## BI2 1.000
## BI3 1.060 0.065 16.357 0.000
## BI5 1.298 0.081 16.065 0.000
## BI6 1.326 0.085 15.552 0.000
## RELA =~
## REL1 1.000
## REL2 0.988 0.043 23.042 0.000
## REL3 0.723 0.038 18.853 0.000
## REL4 0.626 0.042 14.845 0.000
## EDUC =~
## EDU1 1.000
## EDU2 0.994 0.044 22.751 0.000
## EDU3 1.081 0.054 19.927 0.000
## EDU4 1.154 0.057 20.172 0.000
## EDU5 1.027 0.051 20.008 0.000
## EDU6 1.006 0.058 17.267 0.000
## EDU7 1.032 0.052 19.777 0.000
## EDU8 0.876 0.049 17.702 0.000
## GOVE =~
## GOV1 1.000
## GOV2 1.099 0.052 21.287 0.000
## GOV3 1.099 0.052 21.325 0.000
## GOV4 0.988 0.052 19.124 0.000
## GOV5 0.813 0.051 15.936 0.000
## ENDO =~
## END2 1.000
## END3 1.044 0.051 20.633 0.000
## END4 0.789 0.045 17.409 0.000
## END5 0.890 0.055 16.302 0.000
##
## Regressions:
## Estimate Std.Err z-value P(>|z|)
## BEINTEN ~
## RELA 0.372 0.032 11.465 0.000
## EDUC 0.057 0.041 1.372 0.170
## GOVE 0.095 0.043 2.206 0.027
## ENDO -0.060 0.031 -1.923 0.054
## FAM -0.003 0.054 -0.062 0.951
## Formation 0.106 0.051 2.088 0.037
## Work 0.104 0.042 2.500 0.012
## Year -0.083 0.051 -1.635 0.102
##
## Covariances:
## Estimate Std.Err z-value P(>|z|)
## .BI2 ~~
## .BI3 0.237 0.046 5.207 0.000
## .REL1 ~~
## .REL4 -0.152 0.045 -3.394 0.001
## .REL3 ~~
## .REL4 0.447 0.056 7.990 0.000
## .EDU1 ~~
## .EDU2 0.239 0.030 7.957 0.000
## .EDU2 ~~
## .EDU3 0.230 0.027 8.584 0.000
## .EDU6 -0.052 0.025 -2.052 0.040
## .EDU5 ~~
## .EDU6 0.124 0.031 3.976 0.000
## .EDU7 ~~
## .EDU8 0.194 0.028 6.969 0.000
## .GOV4 ~~
## .GOV5 0.228 0.030 7.525 0.000
## .END4 ~~
## .END5 0.345 0.045 7.758 0.000
## RELA ~~
## EDUC 0.520 0.056 9.364 0.000
## GOVE 0.423 0.050 8.448 0.000
## ENDO -0.040 0.051 -0.788 0.430
## EDUC ~~
## GOVE 0.450 0.042 10.790 0.000
## ENDO 0.213 0.037 5.722 0.000
## GOVE ~~
## ENDO 0.188 0.035 5.406 0.000
##
## Variances:
## Estimate Std.Err z-value P(>|z|)
## .BI2 0.919 0.055 16.737 0.000
## .BI3 1.122 0.066 16.984 0.000
## .BI5 0.709 0.054 13.069 0.000
## .BI6 1.016 0.068 15.039 0.000
## .REL1 0.757 0.064 11.804 0.000
## .REL2 0.814 0.064 12.731 0.000
## .REL3 1.114 0.064 17.454 0.000
## .REL4 1.302 0.076 17.155 0.000
## .EDU1 0.998 0.054 18.407 0.000
## .EDU2 0.628 0.036 17.336 0.000
## .EDU3 0.595 0.036 16.521 0.000
## .EDU4 0.629 0.039 16.323 0.000
## .EDU5 0.516 0.032 16.230 0.000
## .EDU6 1.012 0.056 18.064 0.000
## .EDU7 0.565 0.034 16.618 0.000
## .EDU8 0.701 0.039 18.020 0.000
## .GOV1 0.744 0.042 17.563 0.000
## .GOV2 0.393 0.028 13.850 0.000
## .GOV3 0.388 0.028 13.752 0.000
## .GOV4 0.602 0.036 16.878 0.000
## .GOV5 0.814 0.044 18.468 0.000
## .END2 0.596 0.045 13.173 0.000
## .END3 0.351 0.041 8.556 0.000
## .END4 0.806 0.047 17.261 0.000
## .END5 1.254 0.071 17.785 0.000
## .BEINTEN 0.315 0.039 8.085 0.000
## RELA 1.660 0.124 13.342 0.000
## EDUC 0.833 0.079 10.549 0.000
## GOVE 0.721 0.066 10.981 0.000
## ENDO 0.915 0.077 11.956 0.000coef(fitsem4)## BEINTEN=~BI3 BEINTEN=~BI5 BEINTEN=~BI6 RELA=~REL2
## 1.060 1.298 1.326 0.988
## RELA=~REL3 RELA=~REL4 EDUC=~EDU2 EDUC=~EDU3
## 0.723 0.626 0.994 1.081
## EDUC=~EDU4 EDUC=~EDU5 EDUC=~EDU6 EDUC=~EDU7
## 1.154 1.027 1.006 1.032
## EDUC=~EDU8 GOVE=~GOV2 GOVE=~GOV3 GOVE=~GOV4
## 0.876 1.099 1.099 0.988
## GOVE=~GOV5 ENDO=~END3 ENDO=~END4 ENDO=~END5
## 0.813 1.044 0.789 0.890
## BEINTEN~RELA BEINTEN~EDUC BEINTEN~GOVE BEINTEN~ENDO
## 0.372 0.057 0.095 -0.060
## BEINTEN~FAM BEINTEN~Formation BEINTEN~Work BEINTEN~Year
## -0.003 0.106 0.104 -0.083
## BI2~~BI3 REL1~~REL4 REL3~~REL4 EDU1~~EDU2
## 0.237 -0.152 0.447 0.239
## EDU2~~EDU3 EDU2~~EDU6 EDU5~~EDU6 EDU7~~EDU8
## 0.230 -0.052 0.124 0.194
## GOV4~~GOV5 END4~~END5 BI2~~BI2 BI3~~BI3
## 0.228 0.345 0.919 1.122
## BI5~~BI5 BI6~~BI6 REL1~~REL1 REL2~~REL2
## 0.709 1.016 0.757 0.814
## REL3~~REL3 REL4~~REL4 EDU1~~EDU1 EDU2~~EDU2
## 1.114 1.302 0.998 0.628
## EDU3~~EDU3 EDU4~~EDU4 EDU5~~EDU5 EDU6~~EDU6
## 0.595 0.629 0.516 1.012
## EDU7~~EDU7 EDU8~~EDU8 GOV1~~GOV1 GOV2~~GOV2
## 0.565 0.701 0.744 0.393
## GOV3~~GOV3 GOV4~~GOV4 GOV5~~GOV5 END2~~END2
## 0.388 0.602 0.814 0.596
## END3~~END3 END4~~END4 END5~~END5 BEINTEN~~BEINTEN
## 0.351 0.806 1.254 0.315
## RELA~~RELA EDUC~~EDUC GOVE~~GOVE ENDO~~ENDO
## 1.660 0.833 0.721 0.915
## RELA~~EDUC RELA~~GOVE RELA~~ENDO EDUC~~GOVE
## 0.520 0.423 -0.040 0.450
## EDUC~~ENDO GOVE~~ENDO
## 0.213 0.188fitted(fitsem4)## $cov
## BI2 BI3 BI5 BI6 REL1 REL2 REL3 REL4 EDU1 EDU2
## BI2 1.540
## BI3 0.895 1.819
## BI5 0.805 0.853 1.753
## BI6 0.823 0.872 1.068 2.107
## REL1 0.690 0.731 0.895 0.915 2.417
## REL2 0.681 0.722 0.884 0.904 1.640 2.434
## REL3 0.499 0.529 0.647 0.661 1.200 1.186 1.981
## REL4 0.432 0.457 0.560 0.572 0.886 1.026 1.197 1.952
## EDU1 0.271 0.287 0.351 0.359 0.520 0.513 0.376 0.325 1.831
## EDU2 0.269 0.285 0.349 0.357 0.516 0.510 0.373 0.323 1.067 1.450
## EDU3 0.293 0.310 0.380 0.388 0.562 0.555 0.406 0.352 0.901 1.125
## EDU4 0.312 0.331 0.405 0.414 0.600 0.593 0.434 0.375 0.961 0.955
## EDU5 0.278 0.295 0.361 0.369 0.534 0.527 0.386 0.334 0.856 0.850
## EDU6 0.272 0.288 0.353 0.361 0.523 0.516 0.378 0.327 0.838 0.781
## EDU7 0.279 0.296 0.362 0.370 0.537 0.530 0.388 0.336 0.860 0.855
## EDU8 0.237 0.251 0.308 0.314 0.455 0.450 0.329 0.285 0.730 0.725
## GOV1 0.240 0.255 0.312 0.319 0.423 0.418 0.306 0.265 0.450 0.447
## GOV2 0.264 0.280 0.343 0.350 0.465 0.459 0.336 0.291 0.494 0.491
## GOV3 0.264 0.280 0.343 0.350 0.465 0.459 0.336 0.291 0.495 0.491
## GOV4 0.237 0.251 0.308 0.315 0.418 0.413 0.302 0.261 0.444 0.442
## GOV5 0.195 0.207 0.253 0.259 0.344 0.340 0.249 0.215 0.366 0.364
## END2 -0.040 -0.042 -0.052 -0.053 -0.040 -0.040 -0.029 -0.025 0.213 0.212
## END3 -0.042 -0.044 -0.054 -0.055 -0.042 -0.041 -0.030 -0.026 0.223 0.221
## END4 -0.031 -0.033 -0.041 -0.042 -0.032 -0.031 -0.023 -0.020 0.168 0.167
## END5 -0.036 -0.038 -0.046 -0.047 -0.036 -0.035 -0.026 -0.022 0.190 0.189
## FAM 0.001 0.001 0.001 0.001 0.000 0.000 0.000 0.000 0.000 0.000
## Formation 0.027 0.029 0.035 0.036 0.000 0.000 0.000 0.000 0.000 0.000
## Work 0.038 0.040 0.050 0.051 0.000 0.000 0.000 0.000 0.000 0.000
## Year -0.017 -0.018 -0.022 -0.023 0.000 0.000 0.000 0.000 0.000 0.000
## EDU3 EDU4 EDU5 EDU6 EDU7 EDU8 GOV1 GOV2 GOV3 GOV4
## BI2
## BI3
## BI5
## BI6
## REL1
## REL2
## REL3
## REL4
## EDU1
## EDU2
## EDU3 1.568
## EDU4 1.039 1.738
## EDU5 0.925 0.987 1.395
## EDU6 0.906 0.967 0.985 1.855
## EDU7 0.930 0.992 0.883 0.865 1.452
## EDU8 0.789 0.842 0.749 0.734 0.947 1.340
## GOV1 0.486 0.519 0.462 0.452 0.464 0.394 1.464
## GOV2 0.534 0.571 0.508 0.497 0.510 0.433 0.792 1.264
## GOV3 0.535 0.571 0.508 0.497 0.511 0.433 0.792 0.871 1.259
## GOV4 0.480 0.513 0.456 0.447 0.459 0.389 0.712 0.782 0.783 1.305
## GOV5 0.396 0.422 0.376 0.368 0.378 0.320 0.586 0.644 0.644 0.807
## END2 0.231 0.246 0.219 0.215 0.220 0.187 0.188 0.207 0.207 0.186
## END3 0.241 0.257 0.229 0.224 0.230 0.195 0.196 0.216 0.216 0.194
## END4 0.182 0.194 0.173 0.169 0.174 0.147 0.148 0.163 0.163 0.147
## END5 0.205 0.219 0.195 0.191 0.196 0.166 0.167 0.184 0.184 0.165
## FAM 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
## Formation 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
## Work 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
## Year 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
## GOV5 END2 END3 END4 END5 FAM Formtn Work Year
## BI2
## BI3
## BI5
## BI6
## REL1
## REL2
## REL3
## REL4
## EDU1
## EDU2
## EDU3
## EDU4
## EDU5
## EDU6
## EDU7
## EDU8
## GOV1
## GOV2
## GOV3
## GOV4
## GOV5 1.291
## END2 0.153 1.511
## END3 0.160 0.955 1.348
## END4 0.121 0.721 0.753 1.376
## END5 0.136 0.814 0.850 0.988 1.979
## FAM 0.000 0.000 0.000 0.000 0.000 0.218
## Formation 0.000 0.000 0.000 0.000 0.000 0.022 0.249
## Work 0.000 0.000 0.000 0.000 0.000 -0.010 0.014 0.372
## Year 0.000 0.000 0.000 0.000 0.000 -0.006 0.009 0.026 0.250show(fitsem4)## lavaan 0.6-8 ended normally after 45 iterations
##
## Estimator ML
## Optimization method NLMINB
## Number of model parameters 74
##
## Number of observations 826
##
## Model Test User Model:
##
## Test statistic 965.013
## Degrees of freedom 351
## P-value (Chi-square) 0.000logLik(fitsem4)## 'log Lik.' -29220.27 (df=74)fitMeasures(fitsem4, c("chisq", "df", "pvalue", "gfi","cfi","tli", "rmsea"))## chisq df pvalue gfi cfi tli rmsea
## 965.013 351.000 0.000 0.912 0.943 0.935 0.046fitMeasures(fitsem4,("chisq"))/fitMeasures(fitsem4,("df"))## chisq
## 2.749fitMeasures(fitsem4, fit.measures = "all")## npar fmin chisq df
## 74.000 0.584 965.013 351.000
## pvalue baseline.chisq baseline.df baseline.pvalue
## 0.000 11201.862 400.000 0.000
## cfi tli nnfi rfi
## 0.943 0.935 0.935 0.902
## nfi pnfi ifi rni
## 0.914 0.802 0.943 0.943
## logl unrestricted.logl aic bic
## -29220.267 -28737.761 58588.535 58937.563
## ntotal bic2 rmsea rmsea.ci.lower
## 826.000 58702.566 0.046 0.043
## rmsea.ci.upper rmsea.pvalue rmr rmr_nomean
## 0.049 0.970 0.079 0.079
## srmr srmr_bentler srmr_bentler_nomean crmr
## 0.058 0.058 0.058 0.060
## crmr_nomean srmr_mplus srmr_mplus_nomean cn_05
## 0.060 0.058 0.058 339.688
## cn_01 gfi agfi pgfi
## 356.699 0.912 0.891 0.736
## mfi ecvi
## 0.690 1.3475. Ghép kết quả độ phù hợp model3 và model4
Tab5 <- data.frame(cbind(round(fitMeasures(fitsem3, c( "gfi","cfi", "rmsea")),3),
round(fitMeasures(fitsem4, c( "gfi","cfi", "rmsea")), 3)))
names(Tab5) <- c("Sem3", "Sem4")
Tab5["CMINDf", ] <- c(round(fitMeasures(fitsem3,("chisq"))/fitMeasures(fitsem3,("df")), 3),
round(fitMeasures(fitsem4,("chisq"))/fitMeasures(fitsem4,("df")), 3))
Tab5## Sem3 Sem4
## gfi 0.929 0.912
## cfi 0.953 0.943
## rmsea 0.049 0.046
## CMINDf 2.966 2.7496. Ghép bảng hệ số hồi quy của model3 và model4
Reg.coef.sem3 <- summary(fitsem3)## lavaan 0.6-8 ended normally after 45 iterations
##
## Estimator ML
## Optimization method NLMINB
## Number of model parameters 70
##
## Number of observations 826
##
## Model Test User Model:
##
## Test statistic 756.302
## Degrees of freedom 255
## P-value (Chi-square) 0.000
##
## Parameter Estimates:
##
## Standard errors Standard
## Information Expected
## Information saturated (h1) model Structured
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|)
## BEINTEN =~
## BI2 1.000
## BI3 1.060 0.064 16.495 0.000
## BI5 1.307 0.081 16.225 0.000
## BI6 1.335 0.085 15.711 0.000
## RELA =~
## REL1 1.000
## REL2 0.987 0.043 23.040 0.000
## REL3 0.723 0.038 18.864 0.000
## REL4 0.626 0.042 14.854 0.000
## EDUC =~
## EDU1 1.000
## EDU2 0.994 0.044 22.763 0.000
## EDU3 1.081 0.054 19.941 0.000
## EDU4 1.154 0.057 20.186 0.000
## EDU5 1.026 0.051 20.014 0.000
## EDU6 1.005 0.058 17.269 0.000
## EDU7 1.032 0.052 19.782 0.000
## EDU8 0.875 0.049 17.707 0.000
## GOVE =~
## GOV1 1.000
## GOV2 1.099 0.052 21.291 0.000
## GOV3 1.099 0.052 21.326 0.000
## GOV4 0.988 0.052 19.126 0.000
## GOV5 0.813 0.051 15.936 0.000
## ENDO =~
## END2 1.000
## END3 1.044 0.051 20.636 0.000
## END4 0.789 0.045 17.409 0.000
## END5 0.891 0.055 16.309 0.000
##
## Regressions:
## Estimate Std.Err z-value P(>|z|)
## BEINTEN ~
## RELA 0.373 0.033 11.480 0.000
## EDUC 0.079 0.042 1.894 0.058
## GOVE 0.089 0.043 2.046 0.041
## ENDO -0.064 0.031 -2.039 0.041
##
## Covariances:
## Estimate Std.Err z-value P(>|z|)
## .BI2 ~~
## .BI3 0.243 0.046 5.320 0.000
## .REL1 ~~
## .REL4 -0.153 0.045 -3.415 0.001
## .REL3 ~~
## .REL4 0.445 0.056 7.969 0.000
## .EDU1 ~~
## .EDU2 0.238 0.030 7.949 0.000
## .EDU2 ~~
## .EDU3 0.230 0.027 8.577 0.000
## .EDU6 -0.052 0.025 -2.054 0.040
## .EDU5 ~~
## .EDU6 0.125 0.031 3.991 0.000
## .EDU7 ~~
## .EDU8 0.194 0.028 6.979 0.000
## .GOV4 ~~
## .GOV5 0.228 0.030 7.525 0.000
## .END4 ~~
## .END5 0.345 0.045 7.752 0.000
## RELA ~~
## EDUC 0.520 0.056 9.368 0.000
## GOVE 0.423 0.050 8.451 0.000
## ENDO -0.040 0.051 -0.787 0.432
## EDUC ~~
## GOVE 0.450 0.042 10.792 0.000
## ENDO 0.213 0.037 5.722 0.000
## GOVE ~~
## ENDO 0.188 0.035 5.406 0.000
##
## Variances:
## Estimate Std.Err z-value P(>|z|)
## .BI2 0.925 0.055 16.787 0.000
## .BI3 1.128 0.066 17.030 0.000
## .BI5 0.703 0.054 12.950 0.000
## .BI6 1.012 0.068 14.973 0.000
## .REL1 0.755 0.064 11.788 0.000
## .REL2 0.816 0.064 12.767 0.000
## .REL3 1.113 0.064 17.449 0.000
## .REL4 1.301 0.076 17.142 0.000
## .EDU1 0.998 0.054 18.406 0.000
## .EDU2 0.627 0.036 17.333 0.000
## .EDU3 0.594 0.036 16.519 0.000
## .EDU4 0.629 0.039 16.322 0.000
## .EDU5 0.517 0.032 16.241 0.000
## .EDU6 1.012 0.056 18.069 0.000
## .EDU7 0.565 0.034 16.627 0.000
## .EDU8 0.701 0.039 18.024 0.000
## .GOV1 0.743 0.042 17.561 0.000
## .GOV2 0.393 0.028 13.847 0.000
## .GOV3 0.388 0.028 13.754 0.000
## .GOV4 0.602 0.036 16.877 0.000
## .GOV5 0.814 0.044 18.468 0.000
## .END2 0.596 0.045 13.181 0.000
## .END3 0.351 0.041 8.561 0.000
## .END4 0.806 0.047 17.260 0.000
## .END5 1.254 0.071 17.781 0.000
## .BEINTEN 0.319 0.039 8.141 0.000
## RELA 1.661 0.124 13.350 0.000
## EDUC 0.834 0.079 10.555 0.000
## GOVE 0.721 0.066 10.983 0.000
## ENDO 0.914 0.076 11.955 0.000Reg.coef.sem3 <- Reg.coef.sem3[[1]]%>% filter(op == "~") %>% select(lhs,op, rhs ,exo ,est, pvalue)
Reg.coef.sem3$star <- ifelse(Reg.coef.sem3$pvalue <= 0.001, "***",
ifelse(Reg.coef.sem3$pvalue > 0.001 &Reg.coef.sem3$pvalue<0.01, "**",
ifelse(Reg.coef.sem3$pvalue > 0.01 & Reg.coef.sem3$pvalue<0.05, "*", " ")))
Reg.coef.sem3 <- Reg.coef.sem3 %>%
mutate(Sem3 = paste0(round(est, 3),star ), Sem3Pvalue = round(pvalue,3)) %>%
select(rhs, Sem3, Sem3Pvalue)
Reg.coef.sem4 <- summary(fitsem4)## lavaan 0.6-8 ended normally after 45 iterations
##
## Estimator ML
## Optimization method NLMINB
## Number of model parameters 74
##
## Number of observations 826
##
## Model Test User Model:
##
## Test statistic 965.013
## Degrees of freedom 351
## P-value (Chi-square) 0.000
##
## Parameter Estimates:
##
## Standard errors Standard
## Information Expected
## Information saturated (h1) model Structured
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|)
## BEINTEN =~
## BI2 1.000
## BI3 1.060 0.065 16.357 0.000
## BI5 1.298 0.081 16.065 0.000
## BI6 1.326 0.085 15.552 0.000
## RELA =~
## REL1 1.000
## REL2 0.988 0.043 23.042 0.000
## REL3 0.723 0.038 18.853 0.000
## REL4 0.626 0.042 14.845 0.000
## EDUC =~
## EDU1 1.000
## EDU2 0.994 0.044 22.751 0.000
## EDU3 1.081 0.054 19.927 0.000
## EDU4 1.154 0.057 20.172 0.000
## EDU5 1.027 0.051 20.008 0.000
## EDU6 1.006 0.058 17.267 0.000
## EDU7 1.032 0.052 19.777 0.000
## EDU8 0.876 0.049 17.702 0.000
## GOVE =~
## GOV1 1.000
## GOV2 1.099 0.052 21.287 0.000
## GOV3 1.099 0.052 21.325 0.000
## GOV4 0.988 0.052 19.124 0.000
## GOV5 0.813 0.051 15.936 0.000
## ENDO =~
## END2 1.000
## END3 1.044 0.051 20.633 0.000
## END4 0.789 0.045 17.409 0.000
## END5 0.890 0.055 16.302 0.000
##
## Regressions:
## Estimate Std.Err z-value P(>|z|)
## BEINTEN ~
## RELA 0.372 0.032 11.465 0.000
## EDUC 0.057 0.041 1.372 0.170
## GOVE 0.095 0.043 2.206 0.027
## ENDO -0.060 0.031 -1.923 0.054
## FAM -0.003 0.054 -0.062 0.951
## Formation 0.106 0.051 2.088 0.037
## Work 0.104 0.042 2.500 0.012
## Year -0.083 0.051 -1.635 0.102
##
## Covariances:
## Estimate Std.Err z-value P(>|z|)
## .BI2 ~~
## .BI3 0.237 0.046 5.207 0.000
## .REL1 ~~
## .REL4 -0.152 0.045 -3.394 0.001
## .REL3 ~~
## .REL4 0.447 0.056 7.990 0.000
## .EDU1 ~~
## .EDU2 0.239 0.030 7.957 0.000
## .EDU2 ~~
## .EDU3 0.230 0.027 8.584 0.000
## .EDU6 -0.052 0.025 -2.052 0.040
## .EDU5 ~~
## .EDU6 0.124 0.031 3.976 0.000
## .EDU7 ~~
## .EDU8 0.194 0.028 6.969 0.000
## .GOV4 ~~
## .GOV5 0.228 0.030 7.525 0.000
## .END4 ~~
## .END5 0.345 0.045 7.758 0.000
## RELA ~~
## EDUC 0.520 0.056 9.364 0.000
## GOVE 0.423 0.050 8.448 0.000
## ENDO -0.040 0.051 -0.788 0.430
## EDUC ~~
## GOVE 0.450 0.042 10.790 0.000
## ENDO 0.213 0.037 5.722 0.000
## GOVE ~~
## ENDO 0.188 0.035 5.406 0.000
##
## Variances:
## Estimate Std.Err z-value P(>|z|)
## .BI2 0.919 0.055 16.737 0.000
## .BI3 1.122 0.066 16.984 0.000
## .BI5 0.709 0.054 13.069 0.000
## .BI6 1.016 0.068 15.039 0.000
## .REL1 0.757 0.064 11.804 0.000
## .REL2 0.814 0.064 12.731 0.000
## .REL3 1.114 0.064 17.454 0.000
## .REL4 1.302 0.076 17.155 0.000
## .EDU1 0.998 0.054 18.407 0.000
## .EDU2 0.628 0.036 17.336 0.000
## .EDU3 0.595 0.036 16.521 0.000
## .EDU4 0.629 0.039 16.323 0.000
## .EDU5 0.516 0.032 16.230 0.000
## .EDU6 1.012 0.056 18.064 0.000
## .EDU7 0.565 0.034 16.618 0.000
## .EDU8 0.701 0.039 18.020 0.000
## .GOV1 0.744 0.042 17.563 0.000
## .GOV2 0.393 0.028 13.850 0.000
## .GOV3 0.388 0.028 13.752 0.000
## .GOV4 0.602 0.036 16.878 0.000
## .GOV5 0.814 0.044 18.468 0.000
## .END2 0.596 0.045 13.173 0.000
## .END3 0.351 0.041 8.556 0.000
## .END4 0.806 0.047 17.261 0.000
## .END5 1.254 0.071 17.785 0.000
## .BEINTEN 0.315 0.039 8.085 0.000
## RELA 1.660 0.124 13.342 0.000
## EDUC 0.833 0.079 10.549 0.000
## GOVE 0.721 0.066 10.981 0.000
## ENDO 0.915 0.077 11.956 0.000Reg.coef.sem4 <- Reg.coef.sem4[[1]]%>%
filter(op == "~") %>%
select(lhs,op, rhs ,exo ,est, pvalue)
Reg.coef.sem4$star <- ifelse(Reg.coef.sem4$pvalue <= 0.001, "***",
ifelse(Reg.coef.sem4$pvalue > 0.001 &Reg.coef.sem4$pvalue<0.01, "**",
ifelse(Reg.coef.sem4$pvalue > 0.01 & Reg.coef.sem4$pvalue<0.05, "*", " ")))
Reg.coef.sem4 <- Reg.coef.sem4 %>%
mutate(Sem4 = paste0(round(est, 3),star ), Sem4Pvalue = round(pvalue,3)) %>%
select(rhs, Sem4, Sem4Pvalue)
Reg.coef.ALL <- left_join(Reg.coef.sem4, Reg.coef.sem3)## Joining, by = "rhs"Reg.coef.ALL ## rhs Sem4 Sem4Pvalue Sem3 Sem3Pvalue
## 1 RELA 0.372*** 0.000 0.373*** 0.000
## 2 EDUC 0.057 0.170 0.079 0.058
## 3 GOVE 0.095* 0.027 0.089* 0.041
## 4 ENDO -0.06 0.054 -0.064* 0.041
## 5 FAM -0.003 0.951 <NA> NA
## 6 Formation 0.106* 0.037 <NA> NA
## 7 Work 0.104* 0.012 <NA> NA
## 8 Year -0.083 0.102 <NA> NA7. Ghi bảng kết quả dưới file .csv
write.csv(Tab5, file= "Tab5.csv")
write.csv(Reg.coef.ALL, file= "Reg.coef.ALL.csv")8. Vẽ Pathdiagram
semPaths(fitsem3, "std", edge.label.cex = 0.7,
curvePivot = FALSE, rotation = 2,
intercepts = FALSE,
fade = FALSE,
sizeMan2 = 1.5, edge.width = 0.6,# width size of edge.x
covAtResiduals = FALSE,
esize = 2, residScale = 1) 
semPaths(fitsem4, "est", edge.label.cex = 0.5,
curvePivot = FALSE, rotation = 2,
intercepts = FALSE,
fade = FALSE,
sizeMan2 = 1.5, edge.width = 0.6,# width size of edge.x
covAtResiduals = FALSE,
esize = 2, residScale = 1) 
Trân trọng cảm ơn!