用R进行路径分析
来自OBHRM百科
示意图
R脚本与注释
library(lavaan) # 调用lavaan包，如果没有安装，需要先安装
curwd=setwd("c:\\program files\\mplus\\mplus examples\\user's guide examples") # 设定数据文件所在的目录为工作目录，如果脚本与数据文件在同一目录，则不需要设定
pathdata <- read.table("ex3.11.dat") # 读取数据文件到pathdata，pathdata是自己命名，可以随便定
names(pathdata) <-c(paste("y", 1:3, sep=""),paste("x", 1:3,sep="")) # 给变量命名
pathmodel <- ' y1 ~ x1 +x2 +x3 # 设置模型，pathmodel还是自己根据情况命名。因为是路径分析，用的~
y2 ~ x1 +x2 +x3
y3 ~ x2 +y1 +y2'
pathfit <- sem(pathmodel, pathdata) # 进行SEM，pathfit为自己的名称，方便后面调用
summary(pathfit,fit.measures="TRUE") # 显示SEM的总体结果
fitMeasures(pathfit,fit.measures="all", baseline.model=NULL) # 显示所有拟合指数
standardizedSolution(pathfit) # 显示标准化的结果
结果
> summary(pathfit,fit.measures="TRUE") # 显示SEM的总体结果
lavaan (0.5-23.1097) converged normally after 44 iterations
Number of observations 500
Estimator ML
Minimum Function Test Statistic 0.757
Degrees of freedom 3
P-value (Chi-square) 0.860
Model test baseline model:
Minimum Function Test Statistic 4107.449
Degrees of freedom 12
P-value 0.000
User model versus baseline model:
Comparative Fit Index (CFI) 1.000
Tucker-Lewis Index (TLI) 1.002
Loglikelihood and Information Criteria:
Loglikelihood user model (H0) -4556.552
Loglikelihood unrestricted model (H1) -4556.174
Number of free parameters 12
Akaike (AIC) 9137.105
Bayesian (BIC) 9187.680
Sample-size adjusted Bayesian (BIC) 9149.591
Root Mean Square Error of Approximation:
RMSEA 0.000
90 Percent Confidence Interval 0.000 0.040
P-value RMSEA <= 0.05 0.972
Standardized Root Mean Square Residual:
SRMR 0.001
Parameter Estimates:
Information Expected
Standard Errors Standard
Regressions:
Estimate Std.Err z-value P(>|z|)
y1 ~
x1 0.992 0.043 22.979 0.000
x2 2.001 0.045 44.618 0.000
x3 3.052 0.045 68.274 0.000
y2 ~
x1 2.935 0.050 59.002 0.000
x2 1.992 0.052 38.556 0.000
x3 1.023 0.051 19.869 0.000
y3 ~
x2 1.046 0.072 14.539 0.000
y1 0.507 0.020 25.494 0.000
y2 0.746 0.020 37.918 0.000
Variances:
Estimate Std.Err z-value P(>|z|)
.y1 1.061 0.067 15.811 0.000
.y2 1.408 0.089 15.811 0.000
.y3 1.717 0.109 15.811 0.000
> fitMeasures(pathfit,fit.measures="all", baseline.model=NULL) # 显示所有拟合指数
npar fmin chisq df pvalue
12.000 0.001 0.757 3.000 0.860
baseline.chisq baseline.df baseline.pvalue cfi tli
4107.449 12.000 0.000 1.000 1.002
nnfi rfi nfi pnfi ifi
1.002 0.999 1.000 0.250 1.001
rni logl unrestricted.logl aic bic
1.001 -4556.552 -4556.174 9137.105 9187.680
ntotal bic2 rmsea rmsea.ci.lower rmsea.ci.upper
500.000 9149.591 0.000 0.000 0.040
rmsea.pvalue rmr rmr_nomean srmr srmr_bentler
0.972 0.005 0.005 0.001 0.001
srmr_bentler_nomean srmr_bollen srmr_bollen_nomean srmr_mplus srmr_mplus_nomean
0.001 0.001 0.001 0.001 0.001
cn_05 cn_01 gfi agfi pgfi
5164.205 7496.574 1.000 1.000 0.143
mfi ecvi
1.002 0.050
> standardizedSolution(pathfit) # 显示标准化的结果
lhs op rhs est.std se z pvalue
1 y1 ~ x1 0.254 0.011 23.780 0
2 y1 ~ x2 0.495 0.010 51.513 0
3 y1 ~ x3 0.758 0.007 105.969 0
4 y2 ~ x1 0.759 0.008 92.913 0
5 y2 ~ x2 0.497 0.011 44.664 0
6 y2 ~ x3 0.256 0.012 20.578 0
7 y3 ~ x2 0.192 0.013 14.616 0
8 y3 ~ y1 0.375 0.015 25.561 0
9 y3 ~ y2 0.547 0.014 39.362 0
10 y1 ~~ y1 0.061 0.004 15.841 0
11 y2 ~~ y2 0.082 0.005 15.865 0
12 y3 ~~ y3 0.054 0.004 15.369 0
13 x1 ~~ x1 1.000 0.000 NA NA
14 x1 ~~ x2 0.035 0.000 NA NA
15 x1 ~~ x3 -0.052 0.000 NA NA
16 x2 ~~ x2 1.000 0.000 NA NA
17 x2 ~~ x3 0.089 0.000 NA NA
18 x3 ~~ x3 1.000 0.000 NA NA
