Trail Making Test Part A - Psychological Aggression
Model 1
TMTAlog1 <- glm(PsychAggress ~ TrailsA, data=Allvars,family = "binomial")
summary(TMTAlog1)
##
## Call:
## glm(formula = PsychAggress ~ TrailsA, family = "binomial", data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.0612 0.5078 0.5190 0.5338 1.1300
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.117145 0.158671 13.343 <2e-16 ***
## TrailsA -0.006682 0.002709 -2.466 0.0136 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 505.40 on 640 degrees of freedom
## Residual deviance: 500.02 on 639 degrees of freedom
## AIC: 504.02
##
## Number of Fisher Scoring iterations: 4
confint(TMTAlog1)
## 2.5 % 97.5 %
## (Intercept) 1.81004605 2.43478018
## TrailsA -0.01211072 -0.00112323
exp(cbind(OR = coef(TMTAlog1), confint(TMTAlog1)))
## OR 2.5 % 97.5 %
## (Intercept) 8.3073870 6.1107288 11.4133096
## TrailsA 0.9933406 0.9879623 0.9988774
plot(predictorEffect("TrailsA",TMTAlog1))
########Compare to null model
#Difference in Deviance
with(TMTAlog1,null.deviance - deviance)
## [1] 5.381549
#Degrees of freedom for the difference between two models
with(TMTAlog1,df.null - df.residual)
## [1] 1
#p-value
with(TMTAlog1,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))
## [1] 0.0203508
Model 2
TMTAlog2 <- glm(PsychAggress ~ TrailsA + Age0 + Sex + PovStat + WRATtotal, data = Allvars, family = "binomial")
summary(TMTAlog2)
##
## Call:
## glm(formula = PsychAggress ~ TrailsA + Age0 + Sex + PovStat +
## WRATtotal, family = "binomial", data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.3328 0.4179 0.4977 0.5692 1.0358
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 1.351564 0.685237 1.972 0.0486 *
## TrailsA -0.005442 0.002800 -1.944 0.0519 .
## Age0 -0.019052 0.013150 -1.449 0.1474
## SexMen -0.414508 0.237194 -1.748 0.0805 .
## PovStatBelow 0.287057 0.272031 1.055 0.2913
## WRATtotal 0.020193 0.014685 1.375 0.1691
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 505.40 on 640 degrees of freedom
## Residual deviance: 491.19 on 635 degrees of freedom
## AIC: 503.19
##
## Number of Fisher Scoring iterations: 4
confint(TMTAlog2)
## 2.5 % 97.5 %
## (Intercept) 0.03581675 2.7310660748
## TrailsA -0.01092206 0.0004472425
## Age0 -0.04502990 0.0066403168
## SexMen -0.88490865 0.0476851129
## PovStatBelow -0.23280077 0.8382117419
## WRATtotal -0.00913341 0.0486223838
exp(cbind(OR = coef(TMTAlog2), confint(TMTAlog2)))
## OR 2.5 % 97.5 %
## (Intercept) 3.8634619 1.0364659 15.349242
## TrailsA 0.9945728 0.9891374 1.000447
## Age0 0.9811286 0.9559689 1.006662
## SexMen 0.6606650 0.4127519 1.048840
## PovStatBelow 1.3324997 0.7923114 2.312228
## WRATtotal 1.0203987 0.9909082 1.049824
plot(predictorEffect("TrailsA",TMTAlog2))
## Warning in Analyze.model(focal.predictors, mod, xlevels, default.levels, : the
## predictor Age0 is a one-column matrix that was converted to a vector
########Compare to null model
#Difference in Deviance
with(TMTAlog2,null.deviance - deviance)
## [1] 14.20926
#Degrees of freedom for the difference between two models
with(TMTAlog2,df.null - df.residual)
## [1] 5
#p-value
with(TMTAlog2,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))
## [1] 0.01433339
Model 3
TMTAlog3 <- glm(PsychAggress ~ (TrailsA + Sex + PovStat)^3 + Age + WRATtotal, data = Allvars, family = "binomial")
summary(TMTAlog3)
##
## Call:
## glm(formula = PsychAggress ~ (TrailsA + Sex + PovStat)^3 + Age +
## WRATtotal, family = "binomial", data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.3050 0.4145 0.4876 0.5662 0.9371
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.3232837 0.9521720 2.440 0.0147 *
## TrailsA -0.0046246 0.0044733 -1.034 0.3012
## SexMen -0.5352512 0.3757272 -1.425 0.1543
## PovStatBelow -0.3211181 0.9165205 -0.350 0.7261
## Age -0.0211628 0.0133563 -1.584 0.1131
## WRATtotal 0.0213502 0.0147965 1.443 0.1490
## TrailsA:SexMen 0.0002853 0.0061339 0.047 0.9629
## TrailsA:PovStatBelow 0.0121948 0.0262325 0.465 0.6420
## SexMen:PovStatBelow 1.1973680 1.0734939 1.115 0.2647
## TrailsA:SexMen:PovStatBelow -0.0212556 0.0279137 -0.761 0.4464
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 505.40 on 640 degrees of freedom
## Residual deviance: 489.18 on 631 degrees of freedom
## AIC: 509.18
##
## Number of Fisher Scoring iterations: 5
confint(TMTAlog3)
## 2.5 % 97.5 %
## (Intercept) 0.481332779 4.221863340
## TrailsA -0.013299827 0.005857359
## SexMen -1.282198611 0.200258901
## PovStatBelow -2.259676235 1.362523862
## Age -0.047546389 0.004928338
## WRATtotal -0.008162031 0.050025754
## TrailsA:SexMen -0.012576295 0.012711360
## TrailsA:PovStatBelow -0.030481680 0.074099544
## SexMen:PovStatBelow -0.795772727 3.446700474
## TrailsA:SexMen:PovStatBelow -0.085997696 0.025273803
exp(cbind(OR = coef(TMTAlog3), confint(TMTAlog3)))
## OR 2.5 % 97.5 %
## (Intercept) 10.2091427 1.6182297 68.160372
## TrailsA 0.9953861 0.9867882 1.005875
## SexMen 0.5855222 0.2774267 1.221719
## PovStatBelow 0.7253376 0.1043843 3.906039
## Age 0.9790596 0.9535662 1.004941
## WRATtotal 1.0215797 0.9918712 1.051298
## TrailsA:SexMen 1.0002853 0.9875025 1.012792
## TrailsA:PovStatBelow 1.0122695 0.9699782 1.076914
## SexMen:PovStatBelow 3.3113900 0.4512324 31.396627
## TrailsA:SexMen:PovStatBelow 0.9789687 0.9175963 1.025596
########Compare to null model
#Difference in Deviance
with(TMTAlog3,null.deviance - deviance)
## [1] 16.21973
#Degrees of freedom for the difference between two models
with(TMTAlog3,df.null - df.residual)
## [1] 9
#p-value
with(TMTAlog3,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))
## [1] 0.06243246
Compare Models 1,2, & 3
anova(TMTAlog1,TMTAlog2,test = "LR")
## Analysis of Deviance Table
##
## Model 1: PsychAggress ~ TrailsA
## Model 2: PsychAggress ~ TrailsA + Age0 + Sex + PovStat + WRATtotal
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 639 500.02
## 2 635 491.19 4 8.8277 0.06555 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
anova(TMTAlog2,TMTAlog3,test = "LR")
## Analysis of Deviance Table
##
## Model 1: PsychAggress ~ TrailsA + Age0 + Sex + PovStat + WRATtotal
## Model 2: PsychAggress ~ (TrailsA + Sex + PovStat)^3 + Age + WRATtotal
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 635 491.19
## 2 631 489.18 4 2.0105 0.7338
anova(TMTAlog1,TMTAlog3,test = "LR")
## Analysis of Deviance Table
##
## Model 1: PsychAggress ~ TrailsA
## Model 2: PsychAggress ~ (TrailsA + Sex + PovStat)^3 + Age + WRATtotal
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 639 500.02
## 2 631 489.18 8 10.838 0.211
Suggested Model by predictors
anova(TMTAlog3, test="Chisq")
## Analysis of Deviance Table
##
## Model: binomial, link: logit
##
## Response: PsychAggress
##
## Terms added sequentially (first to last)
##
##
## Df Deviance Resid. Df Resid. Dev Pr(>Chi)
## NULL 640 505.40
## TrailsA 1 5.3815 639 500.02 0.02035 *
## Sex 1 3.3913 638 496.63 0.06554 .
## PovStat 1 1.3103 637 495.32 0.25234
## Age 1 2.2815 636 493.04 0.13092
## WRATtotal 1 1.8446 635 491.19 0.17441
## TrailsA:Sex 1 0.1610 634 491.03 0.68825
## TrailsA:PovStat 1 0.4359 633 490.60 0.50911
## Sex:PovStat 1 0.7356 632 489.86 0.39108
## TrailsA:Sex:PovStat 1 0.6780 631 489.18 0.41027
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
TMTAlog4 <- glm(PsychAggress ~ TrailsA + Sex, data = Allvars, family = "binomial")
summary(TMTAlog4)
##
## Call:
## glm(formula = PsychAggress ~ TrailsA + Sex, family = "binomial",
## data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.1553 0.4598 0.4829 0.5732 1.1918
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.335826 0.205687 11.356 <2e-16 ***
## TrailsA -0.006459 0.002729 -2.367 0.0179 *
## SexMen -0.431945 0.235930 -1.831 0.0671 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 505.40 on 640 degrees of freedom
## Residual deviance: 496.63 on 638 degrees of freedom
## AIC: 502.63
##
## Number of Fisher Scoring iterations: 4
plot(predictorEffect("TrailsA",TMTAlog3))
Trail Making Test Part A - Physical Assault
Model 1
TMTAlog1 <- glm(PhysAssault ~ TrailsA, data=Allvars,family = "binomial")
summary(TMTAlog1)
##
## Call:
## glm(formula = PhysAssault ~ TrailsA, family = "binomial", data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -0.5662 -0.5393 -0.5186 -0.4840 2.2518
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.602571 0.283815 -5.647 1.64e-08 ***
## TrailsA -0.010496 0.008213 -1.278 0.201
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 482.53 on 640 degrees of freedom
## Residual deviance: 479.87 on 639 degrees of freedom
## AIC: 483.87
##
## Number of Fisher Scoring iterations: 6
exp(cbind(OR = coef(TMTAlog1), confint(TMTAlog1)))
## OR 2.5 % 97.5 %
## (Intercept) 0.2013782 0.124961 0.3715058
## TrailsA 0.9895584 0.971045 1.0014763
########Compare to null model
#Difference in Deviance
with(TMTAlog1,null.deviance - deviance)
## [1] 2.664886
#Degrees of freedom for the difference between two models
with(TMTAlog1,df.null - df.residual)
## [1] 1
#p-value
with(TMTAlog1,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))
## [1] 0.1025852
Model 2
TMTAlog2 <- glm(PhysAssault ~ TrailsA + Age + Sex + PovStat + WRATtotal, data = Allvars, family = "binomial")
summary(TMTAlog2)
##
## Call:
## glm(formula = PhysAssault ~ TrailsA + Age + Sex + PovStat + WRATtotal,
## family = "binomial", data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -0.8280 -0.5544 -0.4735 -0.3948 2.4131
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.106171 1.056618 -1.047 0.2951
## TrailsA -0.005757 0.006888 -0.836 0.4033
## Age -0.031146 0.014013 -2.223 0.0262 *
## SexMen -0.133250 0.245234 -0.543 0.5869
## PovStatBelow 0.497935 0.253546 1.964 0.0495 *
## WRATtotal 0.014741 0.017431 0.846 0.3977
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 482.53 on 640 degrees of freedom
## Residual deviance: 468.44 on 635 degrees of freedom
## AIC: 480.44
##
## Number of Fisher Scoring iterations: 5
confint(TMTAlog2)
## 2.5 % 97.5 %
## (Intercept) -3.204878260 0.950549388
## TrailsA -0.023671777 0.004114868
## Age -0.058973070 -0.003903141
## SexMen -0.618366886 0.345925262
## PovStatBelow -0.004036407 0.992778921
## WRATtotal -0.018681019 0.049876074
exp(cbind(OR = coef(TMTAlog2), confint(TMTAlog2)))
## OR 2.5 % 97.5 %
## (Intercept) 0.3308234 0.04056384 2.5871306
## TrailsA 0.9942597 0.97660620 1.0041233
## Age 0.9693345 0.94273216 0.9961045
## SexMen 0.8752466 0.53882368 1.4132970
## PovStatBelow 1.6453197 0.99597173 2.6987236
## WRATtotal 1.0148506 0.98149239 1.0511408
########Compare to null model
#Difference in Deviance
with(TMTAlog2,null.deviance - deviance)
## [1] 14.09208
#Degrees of freedom for the difference between two models
with(TMTAlog2,df.null - df.residual)
## [1] 5
#p-value
with(TMTAlog2,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))
## [1] 0.01503496
Model 3
TMTAlog3 <- glm(PhysAssault ~ (TrailsA + Sex + PovStat)^3 + Age + WRATtotal, data = Allvars, family = "binomial")
summary(TMTAlog3)
##
## Call:
## glm(formula = PhysAssault ~ (TrailsA + Sex + PovStat)^3 + Age +
## WRATtotal, family = "binomial", data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -0.9896 -0.5444 -0.4687 -0.3914 2.4456
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.363e+00 1.058e+00 -1.289 0.1973
## TrailsA 7.282e-05 7.277e-03 0.010 0.9920
## SexMen 7.630e-02 5.733e-01 0.133 0.8941
## PovStatBelow 1.662e+00 8.945e-01 1.858 0.0632 .
## Age -2.951e-02 1.412e-02 -2.090 0.0366 *
## WRATtotal 1.521e-02 1.745e-02 0.872 0.3832
## TrailsA:SexMen -8.029e-03 1.526e-02 -0.526 0.5987
## TrailsA:PovStatBelow -4.104e-02 2.948e-02 -1.392 0.1638
## SexMen:PovStatBelow -1.192e+00 1.159e+00 -1.028 0.3040
## TrailsA:SexMen:PovStatBelow 4.372e-02 3.509e-02 1.246 0.2129
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 482.53 on 640 degrees of freedom
## Residual deviance: 466.00 on 631 degrees of freedom
## AIC: 486
##
## Number of Fisher Scoring iterations: 6
confint(TMTAlog3)
## 2.5 % 97.5 %
## (Intercept) -3.466733133 0.698550735
## TrailsA -0.022462644 0.010741157
## SexMen -0.973440421 1.399852489
## PovStatBelow 0.006201422 3.512932999
## Age -0.057593946 -0.002109772
## WRATtotal -0.018221155 0.050392858
## TrailsA:SexMen -0.048493721 0.019944784
## TrailsA:PovStatBelow -0.105757832 0.010049154
## SexMen:PovStatBelow -3.556398613 1.153705945
## TrailsA:SexMen:PovStatBelow -0.024980614 0.119436934
exp(cbind(OR = coef(TMTAlog3), confint(TMTAlog3)))
## OR 2.5 % 97.5 %
## (Intercept) 0.2557689 0.03121885 2.0108364
## TrailsA 1.0000728 0.97778776 1.0107991
## SexMen 1.0792863 0.37778107 4.0546018
## PovStatBelow 5.2680898 1.00622069 33.5465155
## Age 0.9709182 0.94403320 0.9978925
## WRATtotal 1.0153311 0.98194385 1.0516842
## TrailsA:SexMen 0.9920033 0.95266332 1.0201450
## TrailsA:PovStatBelow 0.9597892 0.89964249 1.0100998
## SexMen:PovStatBelow 0.3037608 0.02854143 3.1699187
## TrailsA:SexMen:PovStatBelow 1.0446857 0.97532882 1.1268622
########Compare to null model
#Difference in Deviance
with(TMTAlog3,null.deviance - deviance)
## [1] 16.52871
#Degrees of freedom for the difference between two models
with(TMTAlog3,df.null - df.residual)
## [1] 9
#p-value
with(TMTAlog3,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))
## [1] 0.05662862
Compare Models 1,2, & 3
anova(TMTAlog1,TMTAlog2,test = "LR")
## Analysis of Deviance Table
##
## Model 1: PhysAssault ~ TrailsA
## Model 2: PhysAssault ~ TrailsA + Age + Sex + PovStat + WRATtotal
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 639 479.87
## 2 635 468.44 4 11.427 0.02216 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
anova(TMTAlog2,TMTAlog3,test = "LR")
## Analysis of Deviance Table
##
## Model 1: PhysAssault ~ TrailsA + Age + Sex + PovStat + WRATtotal
## Model 2: PhysAssault ~ (TrailsA + Sex + PovStat)^3 + Age + WRATtotal
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 635 468.44
## 2 631 466.00 4 2.4366 0.656
anova(TMTAlog1,TMTAlog3,test = "LR")
## Analysis of Deviance Table
##
## Model 1: PhysAssault ~ TrailsA
## Model 2: PhysAssault ~ (TrailsA + Sex + PovStat)^3 + Age + WRATtotal
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 639 479.87
## 2 631 466.00 8 13.864 0.08538 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Suggested Model by predictors
anova(TMTAlog3, test="Chisq")
## Analysis of Deviance Table
##
## Model: binomial, link: logit
##
## Response: PhysAssault
##
## Terms added sequentially (first to last)
##
##
## Df Deviance Resid. Df Resid. Dev Pr(>Chi)
## NULL 640 482.53
## TrailsA 1 2.6649 639 479.87 0.10259
## Sex 1 0.5758 638 479.29 0.44795
## PovStat 1 5.0877 637 474.20 0.02410 *
## Age 1 5.0333 636 469.17 0.02486 *
## WRATtotal 1 0.7303 635 468.44 0.39278
## TrailsA:Sex 1 0.0169 634 468.42 0.89666
## TrailsA:PovStat 1 0.7216 633 467.70 0.39561
## Sex:PovStat 1 0.0962 632 467.61 0.75639
## TrailsA:Sex:PovStat 1 1.6019 631 466.00 0.20563
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
TMTAlog4 <- glm(PhysAssault ~ TrailsA + PovStat + Age, data = Allvars, family = "binomial")
summary(TMTAlog4)
##
## Call:
## glm(formula = PhysAssault ~ TrailsA + PovStat + Age, family = "binomial",
## data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -0.7820 -0.5546 -0.4799 -0.3925 2.3219
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.435029 0.650356 -0.669 0.5036
## TrailsA -0.007343 0.007395 -0.993 0.3208
## PovStatBelow 0.471055 0.249746 1.886 0.0593 .
## Age -0.031701 0.014071 -2.253 0.0243 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 482.53 on 640 degrees of freedom
## Residual deviance: 469.43 on 637 degrees of freedom
## AIC: 477.43
##
## Number of Fisher Scoring iterations: 5
Trail Making Test Part B - Psychological Aggression
Model 1
TMTBlog1 <- glm(PsychAggress ~ TrailsB, data=Allvars,family = "binomial")
summary(TMTBlog1)
##
## Call:
## glm(formula = PsychAggress ~ TrailsB, family = "binomial", data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.0693 0.5050 0.5163 0.5371 0.6409
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.096241 0.206390 10.157 <2e-16 ***
## TrailsB -0.002060 0.001462 -1.409 0.159
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 505.40 on 640 degrees of freedom
## Residual deviance: 503.51 on 639 degrees of freedom
## AIC: 507.51
##
## Number of Fisher Scoring iterations: 4
confint(TMTBlog1)
## 2.5 % 97.5 %
## (Intercept) 1.698395018 2.5090041923
## TrailsB -0.004850198 0.0009051254
exp(cbind(OR = coef(TMTBlog1), confint(TMTBlog1)))
## OR 2.5 % 97.5 %
## (Intercept) 8.1355301 5.4651689 12.292683
## TrailsB 0.9979425 0.9951615 1.000906
########Compare to null model
#Difference in Deviance
with(TMTBlog1,null.deviance - deviance)
## [1] 1.895159
#Degrees of freedom for the difference between two models
with(TMTBlog1,df.null - df.residual)
## [1] 1
#p-value
with(TMTBlog1,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))
## [1] 0.1686212
Model 2
TMTBlog2 <- glm(PsychAggress ~ TrailsB + Age + Sex + PovStat + WRATtotal, data = Allvars, family = "binomial")
summary(TMTBlog2)
##
## Call:
## glm(formula = PsychAggress ~ TrailsB + Age + Sex + PovStat +
## WRATtotal, family = "binomial", data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.3531 0.4211 0.5030 0.5782 0.8264
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.0933884 1.0094916 2.074 0.0381 *
## TrailsB -0.0003588 0.0017599 -0.204 0.8385
## Age -0.0211512 0.0134906 -1.568 0.1169
## SexMen -0.4252520 0.2365446 -1.798 0.0722 .
## PovStatBelow 0.2968811 0.2712465 1.095 0.2737
## WRATtotal 0.0220975 0.0166967 1.323 0.1857
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 505.40 on 640 degrees of freedom
## Residual deviance: 494.48 on 635 degrees of freedom
## AIC: 506.48
##
## Number of Fisher Scoring iterations: 4
confint(TMTBlog2)
## 2.5 % 97.5 %
## (Intercept) 0.133869024 4.098898653
## TrailsB -0.003714358 0.003210407
## Age -0.047809751 0.005190886
## SexMen -0.894452306 0.035563355
## PovStatBelow -0.221287309 0.846636446
## WRATtotal -0.010959994 0.054668008
exp(cbind(OR = coef(TMTBlog2), confint(TMTBlog2)))
## OR 2.5 % 97.5 %
## (Intercept) 8.1123569 1.1432431 60.273869
## TrailsB 0.9996413 0.9962925 1.003216
## Age 0.9790709 0.9533151 1.005204
## SexMen 0.6536051 0.4088315 1.036203
## PovStatBelow 1.3456552 0.8014864 2.331791
## WRATtotal 1.0223434 0.9890998 1.056190
########Compare to null model
#Difference in Deviance
with(TMTBlog2,null.deviance - deviance)
## [1] 10.92303
#Degrees of freedom for the difference between two models
with(TMTBlog2,df.null - df.residual)
## [1] 5
#p-value
with(TMTBlog2,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))
## [1] 0.05292765
Model 3
TMTBlog3 <- glm(PsychAggress ~ (TrailsB + Sex + PovStat)^3 + Age + WRATtotal, data = Allvars, family = "binomial")
summary(TMTBlog3)
##
## Call:
## glm(formula = PsychAggress ~ (TrailsB + Sex + PovStat)^3 + Age +
## WRATtotal, family = "binomial", data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.3166 0.4161 0.4923 0.5692 0.9530
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.119e+00 1.070e+00 1.980 0.0477 *
## TrailsB 6.311e-05 2.892e-03 0.022 0.9826
## SexMen -2.864e-01 4.901e-01 -0.584 0.5590
## PovStatBelow 2.764e-01 7.131e-01 0.388 0.6983
## Age -2.183e-02 1.356e-02 -1.610 0.1073
## WRATtotal 2.252e-02 1.719e-02 1.310 0.1900
## TrailsB:SexMen -2.024e-03 3.473e-03 -0.583 0.5600
## TrailsB:PovStatBelow -1.571e-03 5.167e-03 -0.304 0.7611
## SexMen:PovStatBelow -8.859e-01 1.024e+00 -0.865 0.3868
## TrailsB:SexMen:PovStatBelow 1.100e-02 8.062e-03 1.364 0.1725
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 505.40 on 640 degrees of freedom
## Residual deviance: 490.99 on 631 degrees of freedom
## AIC: 510.99
##
## Number of Fisher Scoring iterations: 5
confint(TMTBlog3)
## 2.5 % 97.5 %
## (Intercept) 0.040995977 4.243386118
## TrailsB -0.005314178 0.006180063
## SexMen -1.257432994 0.671448445
## PovStatBelow -1.102935668 1.720782202
## Age -0.048621607 0.004639968
## WRATtotal -0.011510819 0.056032243
## TrailsB:SexMen -0.009085005 0.004661670
## TrailsB:PovStatBelow -0.011454410 0.009293020
## SexMen:PovStatBelow -2.939374247 1.102186380
## TrailsB:SexMen:PovStatBelow -0.004275290 0.028250524
exp(cbind(OR = coef(TMTBlog3), confint(TMTBlog3)))
## OR 2.5 % 97.5 %
## (Intercept) 8.3202488 1.04184791 69.643273
## TrailsB 1.0000631 0.99469992 1.006199
## SexMen 0.7509982 0.28438310 1.957070
## PovStatBelow 1.3184107 0.33189532 5.588898
## Age 0.9784054 0.95254150 1.004651
## WRATtotal 1.0227769 0.98855518 1.057632
## TrailsB:SexMen 0.9979777 0.99095614 1.004673
## TrailsB:PovStatBelow 0.9984302 0.98861094 1.009336
## SexMen:PovStatBelow 0.4123557 0.05289882 3.010741
## TrailsB:SexMen:PovStatBelow 1.0110596 0.99573384 1.028653
########Compare to null model
#Difference in Deviance
with(TMTBlog3,null.deviance - deviance)
## [1] 14.41466
#Degrees of freedom for the difference between two models
with(TMTBlog3,df.null - df.residual)
## [1] 9
#p-value
with(TMTBlog3,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))
## [1] 0.1083207
Compare Models 1,2, & 3
anova(TMTBlog1,TMTBlog2,test = "LR")
## Analysis of Deviance Table
##
## Model 1: PsychAggress ~ TrailsB
## Model 2: PsychAggress ~ TrailsB + Age + Sex + PovStat + WRATtotal
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 639 503.51
## 2 635 494.48 4 9.0279 0.06041 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
anova(TMTBlog2,TMTBlog3,test = "LR")
## Analysis of Deviance Table
##
## Model 1: PsychAggress ~ TrailsB + Age + Sex + PovStat + WRATtotal
## Model 2: PsychAggress ~ (TrailsB + Sex + PovStat)^3 + Age + WRATtotal
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 635 494.48
## 2 631 490.99 4 3.4916 0.4792
anova(TMTBlog1,TMTBlog3,test = "LR")
## Analysis of Deviance Table
##
## Model 1: PsychAggress ~ TrailsB
## Model 2: PsychAggress ~ (TrailsB + Sex + PovStat)^3 + Age + WRATtotal
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 639 503.51
## 2 631 490.99 8 12.52 0.1295
Suggested model by predictors
anova(TMTBlog3, test="Chisq")
## Analysis of Deviance Table
##
## Model: binomial, link: logit
##
## Response: PsychAggress
##
## Terms added sequentially (first to last)
##
##
## Df Deviance Resid. Df Resid. Dev Pr(>Chi)
## NULL 640 505.40
## TrailsB 1 1.8952 639 503.51 0.16862
## Sex 1 3.5987 638 499.91 0.05783 .
## PovStat 1 1.5013 637 498.41 0.22048
## Age 1 2.1980 636 496.21 0.13819
## WRATtotal 1 1.7299 635 494.48 0.18842
## TrailsB:Sex 1 0.0003 634 494.48 0.98510
## TrailsB:PovStat 1 1.1518 633 493.33 0.28317
## Sex:PovStat 1 0.3637 632 492.96 0.54648
## TrailsB:Sex:PovStat 1 1.9758 631 490.99 0.15983
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
TMTBlog4 <- glm(PsychAggress ~ Sex + Age + WRATtotal, data = Allvars, family = "binomial")
summary(TMTBlog4)
##
## Call:
## glm(formula = PsychAggress ~ Sex + Age + WRATtotal, family = "binomial",
## data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.2850 0.4317 0.4986 0.5734 0.8093
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.31934 0.89316 2.597 0.00941 **
## SexMen -0.43214 0.23609 -1.830 0.06718 .
## Age -0.02420 0.01290 -1.875 0.06073 .
## WRATtotal 0.02134 0.01444 1.478 0.13953
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 505.40 on 640 degrees of freedom
## Residual deviance: 495.76 on 637 degrees of freedom
## AIC: 503.76
##
## Number of Fisher Scoring iterations: 4
Trail Making Test Part B - Physical Assault
Model 1
TMTBlog1 <- glm(PhysAssault ~ TrailsB, data=Allvars,family = "binomial")
summary(TMTBlog1)
##
## Call:
## glm(formula = PhysAssault ~ TrailsB, family = "binomial", data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -0.5441 -0.5312 -0.5209 -0.4882 2.1550
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.800563 0.214371 -8.399 <2e-16 ***
## TrailsB -0.001394 0.001730 -0.806 0.42
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 482.53 on 640 degrees of freedom
## Residual deviance: 481.85 on 639 degrees of freedom
## AIC: 485.85
##
## Number of Fisher Scoring iterations: 4
confint(TMTBlog1)
## 2.5 % 97.5 %
## (Intercept) -2.225644745 -1.383091814
## TrailsB -0.004987308 0.001835261
exp(cbind(OR = coef(TMTBlog1), confint(TMTBlog1)))
## OR 2.5 % 97.5 %
## (Intercept) 0.1652059 0.1079978 0.2508019
## TrailsB 0.9986066 0.9950251 1.0018369
########Compare to null model
#Difference in Deviance
with(TMTBlog1,null.deviance - deviance)
## [1] 0.6782001
#Degrees of freedom for the difference between two models
with(TMTBlog1,df.null - df.residual)
## [1] 1
#p-value
with(TMTBlog1,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))
## [1] 0.4102072
Model 2
TMTBlog2 <- glm(PhysAssault ~ TrailsB + Age + Sex + PovStat + WRATtotal, data = Allvars, family = "binomial")
summary(TMTBlog2)
##
## Call:
## glm(formula = PhysAssault ~ TrailsB + Age + Sex + PovStat + WRATtotal,
## family = "binomial", data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -0.8230 -0.5492 -0.4750 -0.3924 2.3795
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.3836639 1.0890102 -1.271 0.2039
## TrailsB 0.0004079 0.0019634 0.208 0.8354
## Age -0.0340712 0.0142096 -2.398 0.0165 *
## SexMen -0.1582465 0.2442859 -0.648 0.5171
## PovStatBelow 0.4860101 0.2528681 1.922 0.0546 .
## WRATtotal 0.0191430 0.0186090 1.029 0.3036
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 482.53 on 640 degrees of freedom
## Residual deviance: 469.38 on 635 degrees of freedom
## AIC: 481.38
##
## Number of Fisher Scoring iterations: 5
confint(TMTBlog2)
## 2.5 % 97.5 %
## (Intercept) -3.550585477 0.728040054
## TrailsB -0.003625984 0.004115663
## Age -0.062358059 -0.006533071
## SexMen -0.641711303 0.318865256
## PovStatBelow -0.014762189 0.979371610
## WRATtotal -0.016588773 0.056560198
exp(cbind(OR = coef(TMTBlog2), confint(TMTBlog2)))
## OR 2.5 % 97.5 %
## (Intercept) 0.2506585 0.02870783 2.0710175
## TrailsB 1.0004079 0.99638058 1.0041241
## Age 0.9665027 0.93954641 0.9934882
## SexMen 0.8536393 0.52639084 1.3755660
## PovStatBelow 1.6258164 0.98534624 2.6627825
## WRATtotal 1.0193274 0.98354806 1.0581903
########Compare to null model
#Difference in Deviance
with(TMTBlog2,null.deviance - deviance)
## [1] 13.14969
#Degrees of freedom for the difference between two models
with(TMTBlog2,df.null - df.residual)
## [1] 5
#p-value
with(TMTBlog2,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))
## [1] 0.02201567
Model 3
TMTBlog3 <- glm(PhysAssault ~ (TrailsB + Sex + PovStat)^3 + Age + WRATtotal, data = Allvars, family = "binomial")
summary(TMTBlog3)
##
## Call:
## glm(formula = PhysAssault ~ (TrailsB + Sex + PovStat)^3 + Age +
## WRATtotal, family = "binomial", data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -0.8076 -0.5530 -0.4680 -0.3874 2.3898
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.3070879 1.1474161 -1.139 0.255
## TrailsB -0.0003952 0.0033321 -0.119 0.906
## SexMen -0.4164733 0.5404079 -0.771 0.441
## PovStatBelow 0.3301475 0.5989564 0.551 0.581
## Age -0.0339108 0.0142097 -2.386 0.017 *
## WRATtotal 0.0193090 0.0188812 1.023 0.306
## TrailsB:SexMen 0.0022476 0.0043351 0.518 0.604
## TrailsB:PovStatBelow 0.0012567 0.0049259 0.255 0.799
## SexMen:PovStatBelow 0.5698502 0.9182035 0.621 0.535
## TrailsB:SexMen:PovStatBelow -0.0046844 0.0073815 -0.635 0.526
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 482.53 on 640 degrees of freedom
## Residual deviance: 468.86 on 631 degrees of freedom
## AIC: 488.86
##
## Number of Fisher Scoring iterations: 5
confint(TMTBlog3)
## 2.5 % 97.5 %
## (Intercept) -3.586519885 0.922804562
## TrailsB -0.007675075 0.005630066
## SexMen -1.488985092 0.639884285
## PovStatBelow -0.853851868 1.506410081
## Age -0.062201434 -0.006373086
## WRATtotal -0.016975165 0.057248566
## TrailsB:SexMen -0.006319224 0.011010130
## TrailsB:PovStatBelow -0.008693041 0.011011262
## SexMen:PovStatBelow -1.229855510 2.384454883
## TrailsB:SexMen:PovStatBelow -0.019760664 0.009613506
########Compare to null model
#Difference in Deviance
with(TMTBlog3,null.deviance - deviance)
## [1] 13.67086
#Degrees of freedom for the difference between two models
with(TMTBlog3,df.null - df.residual)
## [1] 9
#p-value
with(TMTBlog3,pchisq(null.deviance-deviance,df.null-df.residual,lower.tail=FALSE))
## [1] 0.1345244
Compare Models 1,2, & 3
anova(TMTBlog1,TMTBlog2,test = "LR")
## Analysis of Deviance Table
##
## Model 1: PhysAssault ~ TrailsB
## Model 2: PhysAssault ~ TrailsB + Age + Sex + PovStat + WRATtotal
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 639 481.85
## 2 635 469.38 4 12.472 0.01417 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
anova(TMTBlog2,TMTBlog3,test = "LR")
## Analysis of Deviance Table
##
## Model 1: PhysAssault ~ TrailsB + Age + Sex + PovStat + WRATtotal
## Model 2: PhysAssault ~ (TrailsB + Sex + PovStat)^3 + Age + WRATtotal
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 635 469.38
## 2 631 468.86 4 0.52117 0.9714
anova(TMTBlog1,TMTBlog3,test = "LR")
## Analysis of Deviance Table
##
## Model 1: PhysAssault ~ TrailsB
## Model 2: PhysAssault ~ (TrailsB + Sex + PovStat)^3 + Age + WRATtotal
## Resid. Df Resid. Dev Df Deviance Pr(>Chi)
## 1 639 481.85
## 2 631 468.86 8 12.993 0.1121
Sugested Model by predictors
anova(TMTBlog3, test="Chisq")
## Analysis of Deviance Table
##
## Model: binomial, link: logit
##
## Response: PhysAssault
##
## Terms added sequentially (first to last)
##
##
## Df Deviance Resid. Df Resid. Dev Pr(>Chi)
## NULL 640 482.53
## TrailsB 1 0.6782 639 481.85 0.41021
## Sex 1 0.7212 638 481.13 0.39574
## PovStat 1 5.0344 637 476.10 0.02485 *
## Age 1 5.6329 636 470.47 0.01763 *
## WRATtotal 1 1.0830 635 469.38 0.29803
## TrailsB:Sex 1 0.0358 634 469.35 0.84987
## TrailsB:PovStat 1 0.0523 633 469.29 0.81916
## Sex:PovStat 1 0.0247 632 469.27 0.87506
## TrailsB:Sex:PovStat 1 0.4084 631 468.86 0.52281
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
TMTBlog4 <- glm(PhysAssault ~ Age + PovStat, data = Allvars, family = "binomial")
summary(TMTBlog4)
##
## Call:
## glm(formula = PhysAssault ~ Age + PovStat, family = "binomial",
## data = Allvars)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -0.7492 -0.5539 -0.4770 -0.3956 2.3402
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.53355 0.64202 -0.831 0.4059
## Age -0.03471 0.01372 -2.530 0.0114 *
## PovStatBelow 0.45113 0.24862 1.815 0.0696 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 482.53 on 640 degrees of freedom
## Residual deviance: 470.90 on 638 degrees of freedom
## AIC: 476.9
##
## Number of Fisher Scoring iterations: 5