Stroop Women vs. Men
Dissertation Analyses - Stroop with WOMEN only
load(file="/Users/meganwilliams/Desktop/Dissertation/StroopWomen.rdata")
load(file="/Users/meganwilliams/Desktop/Dissertation/StroopMen.rdata")
load(file="/Users/meganwilliams/Desktop/Dissertation/Allvars.rdata")
library(effects)
library(interactions)
library(rcompanion)
library(car)Stroop Color-Word Test - Psychological Aggression
Model 1
SCWTWlog1 <- glm(PsychAggress ~ StroopMixed + WRATtotal, data=StroopWomen,family = "binomial")
summary(SCWTWlog1)##
## Call:
## glm(formula = PsychAggress ~ StroopMixed + WRATtotal, family = "binomial",
## data = StroopWomen)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.4587 0.3579 0.4536 0.5493 0.9890
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.59845 1.13926 -0.525 0.5994
## StroopMixed 0.04615 0.02024 2.281 0.0226 *
## WRATtotal 0.02608 0.02845 0.917 0.3593
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 218.34 on 292 degrees of freedom
## Residual deviance: 208.90 on 290 degrees of freedom
## AIC: 214.9
##
## Number of Fisher Scoring iterations: 5confint(SCWTWlog1)## 2.5 % 97.5 %
## (Intercept) -2.806714567 1.69120772
## StroopMixed 0.006976643 0.08668136
## WRATtotal -0.030979348 0.08115080exp(cbind(OR = coef(SCWTWlog1), confint(SCWTWlog1)))## OR 2.5 % 97.5 %
## (Intercept) 0.5496647 0.06040312 5.426030
## StroopMixed 1.0472334 1.00700104 1.090549
## WRATtotal 1.0264190 0.96949560 1.084534#Plots
plot(predictorEffect("StroopMixed",SCWTWlog1))
########Compare to null model
#Difference in Deviance
with(SCWTWlog1,null.deviance - deviance)## [1] 9.443564#Degrees of freedom for the difference between two models
with(SCWTWlog1,df.null - df.residual)## [1] 2#p-value
with(SCWTWlog1,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))## [1] 0.008899305#Pseudo R-Squared
nagelkerke(SCWTWlog1)## $Models
##
## Model: "glm, PsychAggress ~ StroopMixed + WRATtotal, binomial, StroopWomen"
## Null: "glm, PsychAggress ~ 1, binomial, StroopWomen"
##
## $Pseudo.R.squared.for.model.vs.null
## Pseudo.R.squared
## McFadden 0.0432511
## Cox and Snell (ML) 0.0317167
## Nagelkerke (Cragg and Uhler) 0.0603715
##
## $Likelihood.ratio.test
## Df.diff LogLik.diff Chisq p.value
## -2 -4.7218 9.4436 0.0088993
##
## $Number.of.observations
##
## Model: 293
## Null: 293
##
## $Messages
## [1] "Note: For models fit with REML, these statistics are based on refitting with ML"
##
## $Warnings
## [1] "None"Model 3
SCWTWlog3 <- glm(PsychAggress ~ (StroopMixed + PovStat)^2 + Age + WRATtotal, data = StroopWomen, family = "binomial")
summary(SCWTWlog3)##
## Call:
## glm(formula = PsychAggress ~ (StroopMixed + PovStat)^2 + Age +
## WRATtotal, family = "binomial", data = StroopWomen)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.4410 0.3572 0.4561 0.5539 1.0493
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.857442 1.675271 -0.512 0.6088
## StroopMixed 0.049350 0.024241 2.036 0.0418 *
## PovStatBelow 0.553900 1.340516 0.413 0.6795
## Age 0.000767 0.022247 0.034 0.9725
## WRATtotal 0.027754 0.028966 0.958 0.3380
## StroopMixed:PovStatBelow -0.012855 0.043599 -0.295 0.7681
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 218.34 on 292 degrees of freedom
## Residual deviance: 208.61 on 287 degrees of freedom
## AIC: 220.61
##
## Number of Fisher Scoring iterations: 5confint(SCWTWlog3)## 2.5 % 97.5 %
## (Intercept) -4.152852815 2.44910644
## StroopMixed 0.002609916 0.09823365
## PovStatBelow -2.048140525 3.25280280
## Age -0.042882544 0.04473689
## WRATtotal -0.030313416 0.08385346
## StroopMixed:PovStatBelow -0.097430787 0.07493664exp(cbind(OR = coef(SCWTWlog3), confint(SCWTWlog3)))## OR 2.5 % 97.5 %
## (Intercept) 0.4242459 0.01571951 11.577996
## StroopMixed 1.0505882 1.00261333 1.103221
## PovStatBelow 1.7400252 0.12897451 25.862727
## Age 1.0007673 0.95802391 1.045753
## WRATtotal 1.0281425 0.97014143 1.087470
## StroopMixed:PovStatBelow 0.9872269 0.90716513 1.077816##Plots
plot(predictorEffect("StroopMixed",SCWTWlog3))
########Compare to null model
#Difference in Deviance
with(SCWTWlog3,null.deviance - deviance)## [1] 9.732659#Degrees of freedom for the difference between two models
with(SCWTWlog3,df.null - df.residual)## [1] 5#p-value
with(SCWTWlog3,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))## [1] 0.08317448#Pseudo R-Squared
nagelkerke(SCWTWlog3)## $Models
##
## Model: "glm, PsychAggress ~ (StroopMixed + PovStat)^2 + Age + WRATtotal, binomial, StroopWomen"
## Null: "glm, PsychAggress ~ 1, binomial, StroopWomen"
##
## $Pseudo.R.squared.for.model.vs.null
## Pseudo.R.squared
## McFadden 0.0445752
## Cox and Snell (ML) 0.0326716
## Nagelkerke (Cragg and Uhler) 0.0621891
##
## $Likelihood.ratio.test
## Df.diff LogLik.diff Chisq p.value
## -5 -4.8663 9.7327 0.083174
##
## $Number.of.observations
##
## Model: 293
## Null: 293
##
## $Messages
## [1] "Note: For models fit with REML, these statistics are based on refitting with ML"
##
## $Warnings
## [1] "None"Stroop Color-Word Test - Physical Assault
Model 1
SCWTWlog1 <- glm(PhysAssault ~ StroopMixed + WRATtotal, data=StroopWomen,family = "binomial")
summary(SCWTWlog1)##
## Call:
## glm(formula = PhysAssault ~ StroopMixed + WRATtotal, family = "binomial",
## data = StroopWomen)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -0.7963 -0.5591 -0.5101 -0.4547 2.2336
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -3.121549 1.197104 -2.608 0.00912 **
## StroopMixed 0.025776 0.018955 1.360 0.17386
## WRATtotal 0.008196 0.029639 0.277 0.78214
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 229.86 on 292 degrees of freedom
## Residual deviance: 226.97 on 290 degrees of freedom
## AIC: 232.97
##
## Number of Fisher Scoring iterations: 4confint(SCWTWlog1)## 2.5 % 97.5 %
## (Intercept) -5.59887873 -0.87892919
## StroopMixed -0.01116633 0.06340302
## WRATtotal -0.04826455 0.06857784exp(cbind(OR = coef(SCWTWlog1), confint(SCWTWlog1)))## OR 2.5 % 97.5 %
## (Intercept) 0.0440888 0.003702012 0.4152273
## StroopMixed 1.0261116 0.988895785 1.0654562
## WRATtotal 1.0082299 0.952881667 1.0709840########Compare to null model
#Difference in Deviance
with(SCWTWlog1,null.deviance - deviance)## [1] 2.888132#Degrees of freedom for the difference between two models
with(SCWTWlog1,df.null - df.residual)## [1] 2#p-value
with(SCWTWlog1,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))## [1] 0.2359663#Pseudo R-Squared
nagelkerke(SCWTWlog1)## $Models
##
## Model: "glm, PhysAssault ~ StroopMixed + WRATtotal, binomial, StroopWomen"
## Null: "glm, PhysAssault ~ 1, binomial, StroopWomen"
##
## $Pseudo.R.squared.for.model.vs.null
## Pseudo.R.squared
## McFadden 0.01256490
## Cox and Snell (ML) 0.00980868
## Nagelkerke (Cragg and Uhler) 0.01804230
##
## $Likelihood.ratio.test
## Df.diff LogLik.diff Chisq p.value
## -2 -1.4441 2.8881 0.23597
##
## $Number.of.observations
##
## Model: 293
## Null: 293
##
## $Messages
## [1] "Note: For models fit with REML, these statistics are based on refitting with ML"
##
## $Warnings
## [1] "None"#Plots
plot(predictorEffect("StroopMixed",SCWTWlog1))
Model 3
SCWTWlog3 <- glm(PhysAssault ~ (StroopMixed + PovStat)^2 + Age + WRATtotal, data = StroopWomen, family = "binomial")
summary(SCWTWlog3)##
## Call:
## glm(formula = PhysAssault ~ (StroopMixed + PovStat)^2 + Age +
## WRATtotal, family = "binomial", data = StroopWomen)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -0.8211 -0.5915 -0.4818 -0.3849 2.4741
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -2.40914 1.72686 -1.395 0.163
## StroopMixed 0.02365 0.02407 0.983 0.326
## PovStatBelow 1.24257 1.35168 0.919 0.358
## Age -0.02976 0.02177 -1.367 0.172
## WRATtotal 0.01958 0.03082 0.635 0.525
## StroopMixed:PovStatBelow -0.02257 0.03823 -0.590 0.555
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 229.86 on 292 degrees of freedom
## Residual deviance: 222.22 on 287 degrees of freedom
## AIC: 234.22
##
## Number of Fisher Scoring iterations: 5confint(SCWTWlog3)## 2.5 % 97.5 %
## (Intercept) -5.87854988 0.92002169
## StroopMixed -0.02331525 0.07145934
## PovStatBelow -1.46774799 3.87005788
## Age -0.07352351 0.01221587
## WRATtotal -0.03895401 0.08245625
## StroopMixed:PovStatBelow -0.09775810 0.05304793exp(cbind(OR = coef(SCWTWlog3), confint(SCWTWlog3)))## OR 2.5 % 97.5 %
## (Intercept) 0.08989286 0.002798841 2.509345
## StroopMixed 1.02393384 0.976954453 1.074074
## PovStatBelow 3.46450070 0.230443862 47.945161
## Age 0.97068064 0.929114304 1.012291
## WRATtotal 1.01977669 0.961794943 1.085951
## StroopMixed:PovStatBelow 0.97768583 0.906868247 1.054480########Compare to null model
#Difference in Deviance
with(SCWTWlog3,null.deviance - deviance)## [1] 7.63734#Degrees of freedom for the difference between two models
with(SCWTWlog3,df.null - df.residual)## [1] 5#p-value
with(SCWTWlog3,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))## [1] 0.1773877#Pseudo R-Squared
nagelkerke(SCWTWlog3)## $Models
##
## Model: "glm, PhysAssault ~ (StroopMixed + PovStat)^2 + Age + WRATtotal, binomial, StroopWomen"
## Null: "glm, PhysAssault ~ 1, binomial, StroopWomen"
##
## $Pseudo.R.squared.for.model.vs.null
## Pseudo.R.squared
## McFadden 0.0332264
## Cox and Snell (ML) 0.0257292
## Nagelkerke (Cragg and Uhler) 0.0473268
##
## $Likelihood.ratio.test
## Df.diff LogLik.diff Chisq p.value
## -5 -3.8187 7.6373 0.17739
##
## $Number.of.observations
##
## Model: 293
## Null: 293
##
## $Messages
## [1] "Note: For models fit with REML, these statistics are based on refitting with ML"
##
## $Warnings
## [1] "None"#Plots
plot(predictorEffect("StroopMixed",SCWTWlog3))
Dissertation Analyses - Stroop with MEN only
Stroop Color-Word Test - Psychological Aggression
Model 1
SCWTMlog1 <- glm(PsychAggress ~ StroopMixed + WRATtotal, data=StroopMen,family = "binomial")
summary(SCWTMlog1 )##
## Call:
## glm(formula = PsychAggress ~ StroopMixed + WRATtotal, family = "binomial",
## data = StroopMen)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.1330 0.5059 0.5778 0.6351 0.8038
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.573355 0.957586 0.599 0.549
## StroopMixed 0.025346 0.017682 1.433 0.152
## WRATtotal 0.005201 0.023224 0.224 0.823
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 231.39 on 254 degrees of freedom
## Residual deviance: 228.56 on 252 degrees of freedom
## AIC: 234.56
##
## Number of Fisher Scoring iterations: 4confint(SCWTMlog1 )## 2.5 % 97.5 %
## (Intercept) -1.254688757 2.52463462
## StroopMixed -0.009261462 0.06035586
## WRATtotal -0.041660411 0.04990716exp(cbind(OR = coef(SCWTMlog1 ), confint(SCWTMlog1 )))## OR 2.5 % 97.5 %
## (Intercept) 1.774209 0.2851646 12.486332
## StroopMixed 1.025670 0.9907813 1.062214
## WRATtotal 1.005215 0.9591955 1.051174#Plots
plot(predictorEffect("StroopMixed",SCWTMlog1 ))
########Compare to null model
#Difference in Deviance
with(SCWTMlog1 ,null.deviance - deviance)## [1] 2.831976#Degrees of freedom for the difference between two models
with(SCWTMlog1 ,df.null - df.residual)## [1] 2#p-value
with(SCWTMlog1 ,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))## [1] 0.2426857#Pseudo R-Squared
nagelkerke(SCWTMlog1 )## $Models
##
## Model: "glm, PsychAggress ~ StroopMixed + WRATtotal, binomial, StroopMen"
## Null: "glm, PsychAggress ~ 1, binomial, StroopMen"
##
## $Pseudo.R.squared.for.model.vs.null
## Pseudo.R.squared
## McFadden 0.0122390
## Cox and Snell (ML) 0.0110443
## Nagelkerke (Cragg and Uhler) 0.0185174
##
## $Likelihood.ratio.test
## Df.diff LogLik.diff Chisq p.value
## -2 -1.416 2.832 0.24269
##
## $Number.of.observations
##
## Model: 255
## Null: 255
##
## $Messages
## [1] "Note: For models fit with REML, these statistics are based on refitting with ML"
##
## $Warnings
## [1] "None"Model 3
SCWTMlog3 <- glm(PsychAggress ~ (StroopMixed + PovStat)^2 + Age + WRATtotal, data = StroopMen, family = "binomial")
summary(SCWTMlog3)##
## Call:
## glm(formula = PsychAggress ~ (StroopMixed + PovStat)^2 + Age +
## WRATtotal, family = "binomial", data = StroopMen)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.3281 0.4320 0.5611 0.6493 0.9211
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.136e+00 1.575e+00 1.356 0.1751
## StroopMixed 1.796e-02 2.120e-02 0.847 0.3967
## PovStatBelow 3.589e-01 1.164e+00 0.308 0.7578
## Age -3.449e-02 2.093e-02 -1.648 0.0994 .
## WRATtotal 1.101e-02 2.385e-02 0.462 0.6444
## StroopMixed:PovStatBelow -5.567e-05 3.932e-02 -0.001 0.9989
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 231.39 on 254 degrees of freedom
## Residual deviance: 224.19 on 249 degrees of freedom
## AIC: 236.19
##
## Number of Fisher Scoring iterations: 4confint(SCWTMlog3)## 2.5 % 97.5 %
## (Intercept) -0.91913096 5.280075330
## StroopMixed -0.02344319 0.060112456
## PovStatBelow -1.87688251 2.743036977
## Age -0.07645394 0.005980495
## WRATtotal -0.03684554 0.057212143
## StroopMixed:PovStatBelow -0.07738540 0.078190865exp(cbind(OR = coef(SCWTMlog3), confint(SCWTMlog3)))## OR 2.5 % 97.5 %
## (Intercept) 8.4613002 0.3988655 196.384668
## StroopMixed 1.0181265 0.9768295 1.061956
## PovStatBelow 1.4318083 0.1530665 15.534090
## Age 0.9661016 0.9263956 1.005998
## WRATtotal 1.0110700 0.9638250 1.058880
## StroopMixed:PovStatBelow 0.9999443 0.9255331 1.081329##Plots
plot(predictorEffect("StroopMixed",SCWTMlog3))
########Compare to null model
#Difference in Deviance
with(SCWTMlog3,null.deviance - deviance)## [1] 7.194528#Degrees of freedom for the difference between two models
with(SCWTMlog3,df.null - df.residual)## [1] 5#p-value
with(SCWTMlog3,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))## [1] 0.2065703#Pseudo R-Squared
nagelkerke(SCWTMlog3)## $Models
##
## Model: "glm, PsychAggress ~ (StroopMixed + PovStat)^2 + Age + WRATtotal, binomial, StroopMen"
## Null: "glm, PsychAggress ~ 1, binomial, StroopMen"
##
## $Pseudo.R.squared.for.model.vs.null
## Pseudo.R.squared
## McFadden 0.0310928
## Cox and Snell (ML) 0.0278195
## Nagelkerke (Cragg and Uhler) 0.0466434
##
## $Likelihood.ratio.test
## Df.diff LogLik.diff Chisq p.value
## -5 -3.5973 7.1945 0.20657
##
## $Number.of.observations
##
## Model: 255
## Null: 255
##
## $Messages
## [1] "Note: For models fit with REML, these statistics are based on refitting with ML"
##
## $Warnings
## [1] "None"Stroop Color-Word Test - Physical Assault
Model 1
SCWTMlog1 <- glm(PhysAssault ~ StroopMixed + WRATtotal, data=StroopMen,family = "binomial")
summary(SCWTMlog1 )##
## Call:
## glm(formula = PhysAssault ~ StroopMixed + WRATtotal, family = "binomial",
## data = StroopMen)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -0.6180 -0.4673 -0.4326 -0.3917 2.3280
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -2.53047 1.32925 -1.904 0.057 .
## StroopMixed -0.02395 0.02234 -1.072 0.284
## WRATtotal 0.02287 0.03173 0.721 0.471
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 159.10 on 254 degrees of freedom
## Residual deviance: 157.85 on 252 degrees of freedom
## AIC: 163.85
##
## Number of Fisher Scoring iterations: 5confint(SCWTMlog1 )## 2.5 % 97.5 %
## (Intercept) -5.34789298 -0.09157939
## StroopMixed -0.06797701 0.02003670
## WRATtotal -0.03653533 0.08860834exp(cbind(OR = coef(SCWTMlog1 ), confint(SCWTMlog1 )))## OR 2.5 % 97.5 %
## (Intercept) 0.07962193 0.004758166 0.9124889
## StroopMixed 0.97633854 0.934281952 1.0202388
## WRATtotal 1.02313465 0.964124031 1.0926526########Compare to null model
#Difference in Deviance
with(SCWTMlog1 ,null.deviance - deviance)## [1] 1.252079#Degrees of freedom for the difference between two models
with(SCWTMlog1 ,df.null - df.residual)## [1] 2#p-value
with(SCWTMlog1 ,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))## [1] 0.5347053#Pseudo R-Squared
nagelkerke(SCWTMlog1 )## $Models
##
## Model: "glm, PhysAssault ~ StroopMixed + WRATtotal, binomial, StroopMen"
## Null: "glm, PhysAssault ~ 1, binomial, StroopMen"
##
## $Pseudo.R.squared.for.model.vs.null
## Pseudo.R.squared
## McFadden 0.00786972
## Cox and Snell (ML) 0.00489808
## Nagelkerke (Cragg and Uhler) 0.01055250
##
## $Likelihood.ratio.test
## Df.diff LogLik.diff Chisq p.value
## -2 -0.62604 1.2521 0.53471
##
## $Number.of.observations
##
## Model: 255
## Null: 255
##
## $Messages
## [1] "Note: For models fit with REML, these statistics are based on refitting with ML"
##
## $Warnings
## [1] "None"#Plots
plot(predictorEffect("StroopMixed",SCWTMlog1 ))
Model 3
SCWTMlog3 <- glm(PhysAssault ~ (StroopMixed + PovStat)^2 + Age + WRATtotal, data = StroopMen, family = "binomial")
summary(SCWTMlog3)##
## Call:
## glm(formula = PhysAssault ~ (StroopMixed + PovStat)^2 + Age +
## WRATtotal, family = "binomial", data = StroopMen)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -0.8454 -0.4846 -0.3720 -0.2871 2.5519
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.47757 2.13956 -0.691 0.4898
## StroopMixed -0.05393 0.03160 -1.707 0.0879 .
## PovStatBelow -0.78810 1.41002 -0.559 0.5762
## Age -0.02713 0.02610 -1.039 0.2986
## WRATtotal 0.04040 0.03291 1.228 0.2195
## StroopMixed:PovStatBelow 0.05824 0.04408 1.321 0.1864
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 159.10 on 254 degrees of freedom
## Residual deviance: 149.02 on 249 degrees of freedom
## AIC: 161.02
##
## Number of Fisher Scoring iterations: 5confint(SCWTMlog3)## 2.5 % 97.5 %
## (Intercept) -5.80337887 2.645971863
## StroopMixed -0.11749160 0.007246323
## PovStatBelow -3.63935483 1.949785577
## Age -0.07929506 0.023735879
## WRATtotal -0.02102793 0.108802697
## StroopMixed:PovStatBelow -0.02701156 0.146779894exp(cbind(OR = coef(SCWTMlog3), confint(SCWTMlog3)))## OR 2.5 % 97.5 %
## (Intercept) 0.2281912 0.003017342 14.097139
## StroopMixed 0.9475000 0.889147981 1.007273
## PovStatBelow 0.4547100 0.026269287 7.027181
## Age 0.9732390 0.923767317 1.024020
## WRATtotal 1.0412302 0.979191619 1.114942
## StroopMixed:PovStatBelow 1.0599724 0.973349991 1.158099########Compare to null model
#Difference in Deviance
with(SCWTMlog3,null.deviance - deviance)## [1] 10.0804#Degrees of freedom for the difference between two models
with(SCWTMlog3,df.null - df.residual)## [1] 5#p-value
with(SCWTMlog3,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))## [1] 0.07298889#Pseudo R-Squared
nagelkerke(SCWTMlog3)## $Models
##
## Model: "glm, PhysAssault ~ (StroopMixed + PovStat)^2 + Age + WRATtotal, binomial, StroopMen"
## Null: "glm, PhysAssault ~ 1, binomial, StroopMen"
##
## $Pseudo.R.squared.for.model.vs.null
## Pseudo.R.squared
## McFadden 0.0633586
## Cox and Snell (ML) 0.0387598
## Nagelkerke (Cragg and Uhler) 0.0835048
##
## $Likelihood.ratio.test
## Df.diff LogLik.diff Chisq p.value
## -5 -5.0402 10.08 0.072989
##
## $Number.of.observations
##
## Model: 255
## Null: 255
##
## $Messages
## [1] "Note: For models fit with REML, these statistics are based on refitting with ML"
##
## $Warnings
## [1] "None"#Plots
plot(predictorEffect("StroopMixed",SCWTMlog3))