Analysis with processR package
Moderated Mediation(model no 7)
library(processR)
library(lavaan)
library(MPsychoR)
data(Paskvan)
labels=list(X="wintense",M="cogapp",Y="emotion",W="pclimate")
pmacroModel(7,labels=labels)
moderator=list(name="pclimate",site=list("a"))
model1=tripleEquation(labels=labels,mode=1)
cat(model1)cogapp~wintense
emotion~cogapp+wintensefit=eq2fit(model1,data=Paskvan)
summary(fit[[1]])
Call:
lm(formula = as.formula(eq[i]), data = data)
Residuals:
Min 1Q Median 3Q Max
-2.92667 -0.36259 0.05572 0.45502 2.63811
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 1.81612 0.08611 21.09 <2e-16 ***
wintense 0.54577 0.02373 23.00 <2e-16 ***
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Residual standard error: 0.6421 on 801 degrees of freedom
Multiple R-squared: 0.3976, Adjusted R-squared: 0.3969
F-statistic: 528.8 on 1 and 801 DF, p-value: < 2.2e-16summary(fit[[2]])
Call:
lm(formula = as.formula(eq[i]), data = data)
Residuals:
Min 1Q Median 3Q Max
-3.3928 -0.6739 0.0562 0.6594 4.0171
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 0.08933 0.16690 0.535 0.593
cogapp 0.61143 0.05491 11.135 <2e-16 ***
wintense 0.24926 0.04753 5.245 2e-07 ***
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Residual standard error: 0.998 on 800 degrees of freedom
Multiple R-squared: 0.3184, Adjusted R-squared: 0.3167
F-statistic: 186.9 on 2 and 800 DF, p-value: < 2.2e-16modelsSummaryTable(fit,labels=labels)Consequent | |||||||||
M(cogapp) | Y(emotion) | ||||||||
Antecedent | Coef | SE | t | p | Coef | SE | t | p | |
X(wintense) | 0.546 | 0.024 | 22.995 | <.001 | 0.249 | 0.048 | 5.245 | <.001 | |
M(cogapp) | 0.611 | 0.055 | 11.135 | <.001 | |||||
Constant | 1.816 | 0.086 | 21.090 | <.001 | 0.089 | 0.167 | 0.535 | .593 | |
Observations | 803 | 803 | |||||||
R2 | 0.398 | 0.318 | |||||||
Adjusted R2 | 0.397 | 0.317 | |||||||
Residual SE | 0.642 ( df = 801) | 0.998 ( df = 800) | |||||||
F statistic | F(1,801) = 528.787, p < .001 | F(2,800) = 186.888, p < .001 | |||||||
model=tripleEquation(labels=labels,moderator=moderator)
cat(model)cogapp~a1*wintense+a2*pclimate+a3*wintense:pclimate
emotion~b1*cogapp+c1*wintense
pclimate ~ pclimate.mean*1
pclimate ~~ pclimate.var*pclimate
indirect :=(a1+a3*pclimate.mean)*(b1)
direct :=c1
total := direct + indirect
prop.mediated := indirect / total
indirect.below :=(a1+a3*(pclimate.mean-sqrt(pclimate.var)))*(b1)
indirect.above :=(a1+a3*(pclimate.mean+sqrt(pclimate.var)))*(b1)
direct.below:=c1
direct.above:=c1
total.below := direct.below + indirect.below
total.above := direct.above + indirect.above
prop.mediated.below := indirect.below / total.below
prop.mediated.above := indirect.above / total.abovesemfit=sem(model=model,data=Paskvan,se="boot",bootstrap=10)
summary(semfit,ci=TRUE)lavaan 0.6-3 ended normally after 26 iterations
Optimization method NLMINB
Number of free parameters 16
Number of observations 803
Estimator ML
Model Fit Test Statistic 2031.202
Degrees of freedom 4
P-value (Chi-square) 0.000
Parameter Estimates:
Standard Errors Bootstrap
Number of requested bootstrap draws 10
Number of successful bootstrap draws 10
Regressions:
Estimate Std.Err z-value P(>|z|) ci.lower ci.upper
cogapp ~
wintense (a1) 0.685 0.063 10.950 0.000 0.613 0.796
pclimate (a2) 0.053 0.080 0.670 0.503 -0.062 0.211
wntns:pcl (a3) -0.066 0.020 -3.303 0.001 -0.110 -0.043
emotion ~
cogapp (b1) 0.611 0.068 9.055 0.000 0.476 0.712
wintense (c1) 0.249 0.058 4.274 0.000 0.139 0.341
Covariances:
Estimate Std.Err z-value P(>|z|) ci.lower ci.upper
wintense ~~
wintense:pclmt 1.709 0.132 12.985 0.000 1.516 1.966
Intercepts:
Estimate Std.Err z-value P(>|z|) ci.lower ci.upper
pclimat (pcl.) 2.783 0.019 146.068 0.000 2.771 2.828
.cogapp 1.810 0.265 6.841 0.000 1.341 2.175
.emotion 0.089 0.076 1.172 0.241 0.031 0.249
wintens 3.500 0.017 207.625 0.000 3.439 3.503
wntns:p 9.490 0.092 102.950 0.000 9.294 9.582
Variances:
Estimate Std.Err z-value P(>|z|) ci.lower ci.upper
pclimat (pcl.) 0.804 0.034 23.696 0.000 0.750 0.855
.cogapp 0.385 0.021 17.912 0.000 0.356 0.427
.emotion 0.992 0.053 18.595 0.000 0.925 1.136
wintens 0.912 0.033 27.814 0.000 0.897 1.007
wntns:p 12.814 0.463 27.700 0.000 11.990 13.716
Defined Parameters:
Estimate Std.Err z-value P(>|z|) ci.lower ci.upper
indirect 0.306 0.043 7.043 0.000 0.229 0.374
direct 0.249 0.061 4.055 0.000 0.139 0.341
total 0.555 0.029 18.856 0.000 0.490 0.571
prop.mediated 0.551 0.098 5.602 0.000 0.401 0.725
indirect.below 0.342 0.044 7.720 0.000 0.275 0.421
indirect.above 0.270 0.045 6.055 0.000 0.182 0.327
direct.below 0.249 0.061 4.055 0.000 0.139 0.341
direct.above 0.249 0.061 4.055 0.000 0.139 0.341
total.below 0.591 0.033 17.855 0.000 0.518 0.617
total.above 0.519 0.028 18.229 0.000 0.462 0.546
prop.medtd.blw 0.579 0.092 6.302 0.000 0.447 0.748
prop.meditd.bv 0.520 0.106 4.881 0.000 0.348 0.698statisticalDiagram(7,labels=labels,fit=semfit,whatLabel = "est",radx=0.12)
modmedSummaryTable(semfit,mod="pclimate")Indirect Effect | Direct Effect | ||||||
pclimate(W) | estimate | 95% Bootstrap CI | p | estimate | 95% Bootstrap CI | p | |
1.887 | 0.342 | (0.275 to 0.421) | <.001 | 0.249 | (0.139 to 0.341) | <.001 | |
2.783 | 0.306 | (0.229 to 0.374) | <.001 | 0.249 | (0.139 to 0.341) | <.001 | |
3.680 | 0.270 | (0.182 to 0.327) | <.001 | 0.249 | (0.139 to 0.341) | <.001 | |
boot.ci.type = bca.simple | |||||||
conditionalEffectPlot(semfit,data=Paskvan,mod="pclimate")