Testosterone Data
knitr::purl("Testosterone.Rmd",output="Test_pilot.R")##
##
## processing file: Testosterone.Rmd##
|
| | 0%
|
|.. | 3%
|
|... | 5%
|
|..... | 8%
|
|....... | 11%
|
|......... | 13%
|
|.......... | 16%
|
|............ | 18%
|
|.............. | 21%
|
|............... | 24%
|
|................. | 26%
|
|................... | 29%
|
|..................... | 32%
|
|...................... | 34%
|
|........................ | 37%
|
|.......................... | 39%
|
|........................... | 42%
|
|............................. | 45%
|
|............................... | 47%
|
|................................ | 50%
|
|.................................. | 53%
|
|.................................... | 55%
|
|...................................... | 58%
|
|....................................... | 61%
|
|......................................... | 63%
|
|........................................... | 66%
|
|............................................ | 68%
|
|.............................................. | 71%
|
|................................................ | 74%
|
|.................................................. | 76%
|
|................................................... | 79%
|
|..................................................... | 82%
|
|....................................................... | 84%
|
|........................................................ | 87%
|
|.......................................................... | 89%
|
|............................................................ | 92%
|
|.............................................................. | 95%
|
|............................................................... | 97%
|
|.................................................................| 100%## output file: Test_pilot.R## [1] "Test_pilot.R"GM.center <- function(x) {
if(is.numeric(x)==FALSE){x <- as.numeric(x)}
if(is.integer(x)==TRUE){x <- as.numeric(x)}
m <- mean(x, na.rm = TRUE)
out <- x-m
out
}
r.t <- function(data,type="none",reflect=FALSE,offset=0) {
x <- data
# x <- A1$ADRSQOPP
# reflect <- TRUE
# offset <- 1
# type <- "log"
if(reflect==TRUE) {
dir <- 1
offset <- 1
off2 <- 0
m.x <- max(x,na.rm=TRUE)+offset
} else {
if(offset==0 & min(x,na.rm=TRUE)<=0) {offset <- 1}
m.x <- 0
dir <- -1
off2 <- offset
}
out <- NULL
switch(type,
sqrt={
for (i in 1:length(x)) {
if(is.na(x[i])==FALSE) {
x[i] <- (dir*(m.x-x[i]))+off2
out[i] <- sqrt(x[i])
} else {
out[i] <- x[i]
}
}
},
log={
for (i in 1:length(x)) {
if(is.na(x[i])==FALSE) {
x[i] <- (dir*(m.x-x[i]))+off2
out[i] <- log(x[i])
} else {
out[i] <- x[i]
}
}
},
inverse={
for (i in 1:length(x)) {
if(is.na(x[i])==FALSE) {
x[i] <- (dir*(m.x-x[i]))+off2
out[i] <- 1/x[i]
} else {
out[i] <- x[i]
}
}
},
none={
for (i in 1:length(x)) {
if(is.na(x[i])==FALSE) {
out[i] <- (dir*(m.x-x[i]))+off2
}
}
}
)
return(out)
}
## Load Packages
library(nlme)
library(lsmeans)## Loading required package: estimabilitylibrary(readr)## Warning: package 'readr' was built under R version 3.3.2library(ggplot2)
## Load Data
T_DF <- read_csv("~/R/DataHouse/PROJECTS/ChanceStrenth/T_Data_Pilot.csv", # <- Enter your own Path to the data here.
col_types = cols(Anger = col_double(),
Empathy = col_double(), Ethical.Risk = col_double(),
Finacial.Risk = col_double(), `Health_Safety Risk` = col_double(),
Hostility = col_double(), Physical.Aggression = col_double(),
`Recreational Risk` = col_double(),
Social = col_double(), Total.Aggression = col_double(),
Total.Risk = col_double(), Verbal.Aggression = col_double()))## Warning: Duplicated column names deduplicated: 'Sex' => 'Sex_1' [7]## Warning: The following named parsers don't match the column names:
## Health_Safety Risk, Recreational RiskT_DF <- as.data.frame(T_DF)
###Change data types
T_DF$ID <- as.factor(T_DF$ID)
T_DF$Pre <- as.factor(T_DF$Pre)
T_DF$Sex_1 <- as.factor(T_DF$Sex_1)
T_DF$RinseC <- as.factor(T_DF$RinseC)
T_DF$Session <- as.factor(T_DF$Session)
T_DF$InterventionC <- as.factor(T_DF$InterventionC)
T_DF.RAW <- T_DF
###GrandMeanCenter all columns
T_DF.GM.pred <- T_DF[,11:22]
for (i in 1:ncol(T_DF.GM.pred)) {
T_DF.GM.pred[,i] <- GM.center(T_DF.GM.pred[,i])
name <- colnames(T_DF.GM.pred)[i]
colnames(T_DF.GM.pred)[i] <- paste(name,".","GM",sep="")
}
T_DF <- cbind.data.frame(T_DF,T_DF.GM.pred)
T_DF$GM.time <- GM.center(T_DF$Time) #GM.time
T_DF$Test.l <- r.t(T_DF$Test,type="log")
T_DF$Test.l.GM <- GM.center(T_DF$Test.l) #GMC Testosterone
# Remove Missing Testosterone Data
MissingDV <- which(is.na(T_DF$Test)) #Remove Mising values due to Not having data T data
T_DF.NA <- T_DF[-MissingDV,]#Time Trends
Model1 <- lme(fixed=Test.l~GM.time*Sex+I(GM.time^2)*Sex,random=list(ID=pdDiag(~GM.time+I(GM.time^2)),Session=pdDiag(~GM.time+I(GM.time^2))),correlation = corAR1(),data=T_DF.NA)
summary(Model1)## Linear mixed-effects model fit by REML
## Data: T_DF.NA
## AIC BIC logLik
## 487.8053 535.5057 -229.9027
##
## Random effects:
## Formula: ~GM.time + I(GM.time^2) | ID
## Structure: Diagonal
## (Intercept) GM.time I(GM.time^2)
## StdDev: 0.2420114 4.139588e-11 3.807967e-11
##
## Formula: ~GM.time + I(GM.time^2) | Session %in% ID
## Structure: Diagonal
## (Intercept) GM.time I(GM.time^2) Residual
## StdDev: 0.0001421114 1.576449e-11 2.098194e-12 0.7372307
##
## Correlation Structure: AR(1)
## Formula: ~1 | ID/Session
## Parameter estimate(s):
## Phi
## 0.5770667
## Fixed effects: Test.l ~ GM.time * Sex + I(GM.time^2) * Sex
## Value Std.Error DF t-value p-value
## (Intercept) 5.585607 0.16707264 179 33.43220 0.0000
## GM.time 0.081279 0.04996890 179 1.62658 0.1056
## Sex 0.218064 0.24032604 10 0.90737 0.3856
## I(GM.time^2) -0.235528 0.03141161 179 -7.49811 0.0000
## GM.time:Sex 0.045294 0.07255579 179 0.62427 0.5332
## Sex:I(GM.time^2) 0.071660 0.04561273 179 1.57106 0.1179
## Correlation:
## (Intr) GM.tim Sex I(GM.^ GM.t:S
## GM.time 0.000
## Sex -0.695 0.000
## I(GM.time^2) -0.509 0.000 0.354
## GM.time:Sex 0.000 -0.689 0.001 0.000
## Sex:I(GM.time^2) 0.350 0.000 -0.510 -0.689 -0.008
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.34632690 -0.76212607 -0.08414732 0.89875816 2.24221901
##
## Number of Observations: 229
## Number of Groups:
## ID Session %in% ID
## 12 46#Remove First time point
T_DF.NA.rm1 <- subset(T_DF.NA,Time==2|Time==3|Time==4|Time==5) #only for the last 4 time points
#T_DF.NA.rm1 <- subset(T_DF.NA,Time==3|Time==4|Time==5) #only for the last 3 time points
T_DF.NA.rm1$GM.time2 <- GM.center(T_DF.NA.rm1$Time)
options(contrasts=c("contr.treatment","contr.poly"))
Model1 <- lme(fixed=Test.l~GM.time2*Sex_1,random=list(ID=pdDiag(~GM.time2),Session=pdDiag(~GM.time2)),correlation = corAR1(),data=T_DF.NA.rm1) # Linear Model
summary(Model1)## Linear mixed-effects model fit by REML
## Data: T_DF.NA.rm1
## AIC BIC logLik
## 243.6604 275.59 -111.8302
##
## Random effects:
## Formula: ~GM.time2 | ID
## Structure: Diagonal
## (Intercept) GM.time2
## StdDev: 2.689795e-12 3.576292e-16
##
## Formula: ~GM.time2 | Session %in% ID
## Structure: Diagonal
## (Intercept) GM.time2 Residual
## StdDev: 0.0004542804 0.1515614 0.825445
##
## Correlation Structure: AR(1)
## Formula: ~1 | ID/Session
## Parameter estimate(s):
## Phi
## 0.9209313
## Fixed effects: Test.l ~ GM.time2 * Sex_1
## Value Std.Error DF t-value p-value
## (Intercept) 5.288249 0.15896407 136 33.26695 0.0000
## GM.time2 -0.290712 0.04835715 136 -6.01177 0.0000
## Sex_1Male 0.332197 0.22986166 10 1.44520 0.1790
## GM.time2:Sex_1Male 0.161458 0.06992432 136 2.30903 0.0224
## Correlation:
## (Intr) GM.tm2 Sx_1Ml
## GM.time2 0.000
## Sex_1Male -0.692 0.000
## GM.time2:Sex_1Male 0.000 -0.692 0.000
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.11089632 -0.85068311 -0.02536655 0.91844062 1.70287034
##
## Number of Observations: 184
## Number of Groups:
## ID Session %in% ID
## 12 46 #Fitting a quadraditc model does not fit as well as the linear model
# Model1 <- lme(fixed=Test.l~GM.time2+I(GM.time2^2)*Sex_1,random=list(ID=pdDiag(~GM.time2+I(GM.time2^2)),
# Session=pdDiag(~GM.time2+I(GM.time2^2))),correlation = corAR1(),data=T_DF.NA.rm1)
# summary(Model1)#Effect of the Rinse Condition
Model2 <- lme(fixed=Test.l~GM.time*RinseC*Sex,random=list(ID=pdDiag(~GM.time),Session=pdDiag(~GM.time)),correlation = corAR1(),data=T_DF.NA)
summary(Model2)## Linear mixed-effects model fit by REML
## Data: T_DF.NA
## AIC BIC logLik
## 550.647 598.2213 -261.3235
##
## Random effects:
## Formula: ~GM.time | ID
## Structure: Diagonal
## (Intercept) GM.time
## StdDev: 0.2198911 1.756494e-09
##
## Formula: ~GM.time | Session %in% ID
## Structure: Diagonal
## (Intercept) GM.time Residual
## StdDev: 0.0001328074 1.604299e-09 0.8297201
##
## Correlation Structure: AR(1)
## Formula: ~1 | ID/Session
## Parameter estimate(s):
## Phi
## 0.5217231
## Fixed effects: Test.l ~ GM.time * RinseC * Sex
## Value Std.Error DF t-value p-value
## (Intercept) 5.090170 0.1829617 179 27.820963 0.0000
## GM.time 0.132222 0.0809577 179 1.633225 0.1042
## RinseCRinse -0.237126 0.2254608 32 -1.051737 0.3008
## Sex 0.320958 0.2636437 10 1.217394 0.2514
## GM.time:RinseCRinse -0.112276 0.1144915 179 -0.980647 0.3281
## GM.time:Sex 0.007129 0.1170646 179 0.060902 0.9515
## RinseCRinse:Sex 0.177963 0.3265849 32 0.544922 0.5896
## GM.time:RinseCRinse:Sex 0.071118 0.1662653 179 0.427739 0.6694
## Correlation:
## (Intr) GM.tim RnsCRn Sex GM.t:RCR GM.t:S RnCR:S
## GM.time 0.000
## RinseCRinse -0.616 0.000
## Sex -0.694 0.000 0.428
## GM.time:RinseCRinse 0.000 -0.707 0.000 0.000
## GM.time:Sex 0.000 -0.692 0.000 0.000 0.489
## RinseCRinse:Sex 0.425 0.000 -0.690 -0.617 0.000 0.000
## GM.time:RinseCRinse:Sex 0.000 0.487 0.000 0.000 -0.689 -0.704 -0.005
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -1.5715334 -0.5610296 -0.2266947 0.9530433 2.4671850
##
## Number of Observations: 229
## Number of Groups:
## ID Session %in% ID
## 12 46 anova(Model2)## numDF denDF F-value p-value
## (Intercept) 1 179 2488.6042 <.0001
## GM.time 1 179 5.4315 0.0209
## RinseC 1 32 0.8808 0.3550
## Sex 1 10 3.9011 0.0765
## GM.time:RinseC 1 179 0.8978 0.3446
## GM.time:Sex 1 179 0.2630 0.6087
## RinseC:Sex 1 32 0.2995 0.5880
## GM.time:RinseC:Sex 1 179 0.1830 0.6694 Model2 <- lme(fixed=Test.l~GM.time*RinseC*Sex,random=list(ID=pdDiag(~GM.time),Session=pdDiag(~GM.time)),correlation = corAR1(),data=subset(T_DF.NA,InterventionC=="Testosterone"))
summary(Model2)## Linear mixed-effects model fit by REML
## Data: subset(T_DF.NA, InterventionC == "Testosterone")
## AIC BIC logLik
## 337.5185 375.5775 -154.7592
##
## Random effects:
## Formula: ~GM.time | ID
## Structure: Diagonal
## (Intercept) GM.time
## StdDev: 0.2161682 1.383131e-05
##
## Formula: ~GM.time | Session %in% ID
## Structure: Diagonal
## (Intercept) GM.time Residual
## StdDev: 1.269412e-05 8.314083e-06 0.8247038
##
## Correlation Structure: AR(1)
## Formula: ~1 | ID/Session
## Parameter estimate(s):
## Phi
## 0.01262178
## Fixed effects: Test.l ~ GM.time * RinseC * Sex
## Value Std.Error DF t-value p-value
## (Intercept) 5.741646 0.16742537 92 34.29376 0.0000
## GM.time 0.125586 0.09906292 92 1.26774 0.2081
## RinseCRinse -0.232546 0.20826830 10 -1.11657 0.2903
## Sex 0.276768 0.25444787 10 1.08772 0.3022
## GM.time:RinseCRinse -0.148169 0.14581668 92 -1.01613 0.3122
## GM.time:Sex 0.070192 0.15346762 92 0.45738 0.6485
## RinseCRinse:Sex 0.162862 0.30892165 10 0.52719 0.6096
## GM.time:RinseCRinse:Sex 0.102852 0.21552353 92 0.47722 0.6343
## Correlation:
## (Intr) GM.tim RnsCRn Sex GM.t:RCR GM.t:S RnCR:S
## GM.time 0.000
## RinseCRinse -0.581 0.000
## Sex -0.658 0.000 0.382
## GM.time:RinseCRinse 0.000 -0.679 0.000 0.000
## GM.time:Sex 0.000 -0.645 0.000 0.000 0.439
## RinseCRinse:Sex 0.391 0.000 -0.674 -0.626 0.000 0.000
## GM.time:RinseCRinse:Sex 0.000 0.460 0.000 0.000 -0.677 -0.712 0.000
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -1.99646476 -0.89344179 0.08229893 0.85675813 1.61804519
##
## Number of Observations: 120
## Number of Groups:
## ID Session %in% ID
## 12 24 anova(Model2)## numDF denDF F-value p-value
## (Intercept) 1 92 3436.734 <.0001
## GM.time 1 92 4.180 0.0438
## RinseC 1 10 0.824 0.3855
## Sex 1 10 3.301 0.0993
## GM.time:RinseC 1 92 0.722 0.3978
## GM.time:Sex 1 92 1.289 0.2591
## RinseC:Sex 1 10 0.278 0.6096
## GM.time:RinseC:Sex 1 92 0.228 0.6343#Effect of the Intervention
Model3 <- lme(fixed=Test.l~GM.time+InterventionC*Sex,random=list(ID=pdDiag(~GM.time),Session=pdDiag(~GM.time)),correlation = corAR1(),data=T_DF.NA)
summary(Model3)## Linear mixed-effects model fit by REML
## Data: T_DF.NA
## AIC BIC logLik
## 479.1987 516.7268 -228.5994
##
## Random effects:
## Formula: ~GM.time | ID
## Structure: Diagonal
## (Intercept) GM.time
## StdDev: 0.2750274 1.771653e-05
##
## Formula: ~GM.time | Session %in% ID
## Structure: Diagonal
## (Intercept) GM.time Residual
## StdDev: 6.186147e-06 4.088338e-06 0.6172766
##
## Correlation Structure: AR(1)
## Formula: ~1 | ID/Session
## Parameter estimate(s):
## Phi
## 0.008256705
## Fixed effects: Test.l ~ GM.time + InterventionC * Sex
## Value Std.Error DF t-value p-value
## (Intercept) 4.507605 0.14048287 182 32.08651 0.0000
## GM.time 0.054954 0.02900793 182 1.89445 0.0598
## InterventionCTestosterone 1.119666 0.11545986 32 9.69744 0.0000
## Sex 0.473418 0.19913718 10 2.37735 0.0388
## InterventionCTestosterone:Sex -0.118666 0.16584871 32 -0.71551 0.4795
## Correlation:
## (Intr) GM.tim IntrCT Sex
## GM.time 0.000
## InterventionCTestosterone -0.445 0.000
## Sex -0.705 -0.005 0.314
## InterventionCTestosterone:Sex 0.310 0.007 -0.696 -0.435
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.3753415 -0.4171013 0.0568346 0.7002506 2.4142695
##
## Number of Observations: 229
## Number of Groups:
## ID Session %in% ID
## 12 46#Models for correlations with predictors
MissingDV2 <- which(is.na(T_DF.NA$Total.Aggression)) #this removes time points 2,3, 4 from the data frame
T_DF.NA2 <- T_DF.NA[-MissingDV2,]
##Calculate Change Scores
uni <- unique(T_DF.NA2$ID)
T_DF.chng <- NULL
for (i in 1:length(uni)) { #this subtracts the post time from the previous time
df.temp <- subset(T_DF.NA2,ID==uni[i])
uni2 <- unique(df.temp$Session)
for (ii in 1:length(uni2)) {
df.temp2 <- subset(df.temp,Session==uni2[ii])
chng <- df.temp2[2,10:24]-df.temp2[1,10:24]
chng2 <- rbind.data.frame(chng,chng)
df.out <- cbind.data.frame(df.temp2,chng2)
T_DF.chng <- rbind.data.frame(T_DF.chng,df.out)
}
}## Warning in data.frame(..., check.names = FALSE): row names were found from
## a short variable and have been discardedT_DF.chng <- na.exclude(T_DF.chng)
T_DF.chng <- subset(T_DF.chng,Pre=="Before" & InterventionC=="Testosterone")
colMeans(T_DF.chng[,26:37]) #AVerage Change scores observed## Anger.GM Hostility.GM Total.Risk.GM
## 0.5326087 0.1068841 -3.0996377
## Finacial.Risk.GM Ethical.Risk.GM Health_Safety.Risk.GM
## 0.4275362 -0.8496377 -0.6467391
## Recreational.Risk.GM Social.GM Empathy.GM
## -1.0760870 -0.9547101 -0.1884058
## GM.time Test.l Test.l.GM
## -2.0000000 4.5932844 -0.6974727 mean(T_DF.chng$Test.1)## [1] 207.362 #Analysis of change scores for Total Aggression
Model4.total.a <- lme(fixed=Test.1~Total.Aggression.1,random=~1 |ID/Session,data=T_DF.chng)
summary(Model4.total.a)## Linear mixed-effects model fit by REML
## Data: T_DF.chng
## AIC BIC logLik
## 297.7943 303.2496 -143.8972
##
## Random effects:
## Formula: ~1 | ID
## (Intercept)
## StdDev: 165.7434
##
## Formula: ~1 | Session %in% ID
## (Intercept) Residual
## StdDev: 64.90407 37.07113
##
## Fixed effects: Test.1 ~ Total.Aggression.1
## Value Std.Error DF t-value p-value
## (Intercept) 210.72255 50.51118 11 4.171800 0.0016
## Total.Aggression.1 0.46801 3.06407 11 0.152741 0.8814
## Correlation:
## (Intr)
## Total.Aggression.1 -0.085
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -0.89491748 -0.28316357 0.05252704 0.24032316 0.60103553
##
## Number of Observations: 24
## Number of Groups:
## ID Session %in% ID
## 12 24 p.c <- ggplot(T_DF.chng,aes(y=Test.1,x=Total.Aggression.1)) + geom_point(size=2) + geom_smooth(method="lm")
p.c
T_DF.max <- NULL
uni <- unique(T_DF.NA$ID)
for (i in 1:length(uni)) { #this subtracts the post time from the previous time
df.temp <- subset(T_DF.NA,ID==uni[i])
uni2 <- unique(df.temp$Session)
for (ii in 1:length(uni2)) {
df.temp2 <- subset(df.temp,Session==uni2[ii])
df.temp2$Test.l.GM[df.temp2$Time==5] <- df.temp2$Test.l.GM[df.temp2$Time==5]
T_DF.max <- rbind.data.frame(T_DF.max,df.temp2)
}
}
MissingDV3 <- which(is.na(T_DF.max$Total.Aggression)) #this removes time points 2,3, 4 from the data frame
T_DF.NA3.max <- T_DF.max[-MissingDV3,]
Model.t_max.1 <- lme(fixed=Total.Aggression~Test.l.GM*Sex_1,random=~1|ID/Session,data=subset(T_DF.max,Time==5),method="ML")
summary(Model.t_max.1)## Linear mixed-effects model fit by maximum likelihood
## Data: subset(T_DF.max, Time == 5)
## AIC BIC logLik
## 322.2298 335.0303 -154.1149
##
## Random effects:
## Formula: ~1 | ID
## (Intercept)
## StdDev: 7.443528
##
## Formula: ~1 | Session %in% ID
## (Intercept) Residual
## StdDev: 4.675683 2.271383
##
## Fixed effects: Total.Aggression ~ Test.l.GM * Sex_1
## Value Std.Error DF t-value p-value
## (Intercept) 29.853292 3.510297 32 8.504493 0.0000
## Test.l.GM 3.658112 2.253224 32 1.623501 0.1143
## Sex_1Male 0.984791 4.894156 10 0.201218 0.8446
## Test.l.GM:Sex_1Male -3.772855 3.097254 32 -1.218129 0.2321
## Correlation:
## (Intr) Ts..GM Sx_1Ml
## Test.l.GM 0.281
## Sex_1Male -0.717 -0.202
## Test.l.GM:Sex_1Male -0.205 -0.727 0.104
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -0.92072312 -0.23831532 0.00476136 0.24137541 0.79640484
##
## Number of Observations: 46
## Number of Groups:
## ID Session %in% ID
## 12 46Model.t_max.2 <- lme(fixed=Total.Aggression~Test.l.GM*Sex_1,random=~1|ID/Session,data=T_DF.NA3.max,method="ML")
summary(Model.t_max.2)## Linear mixed-effects model fit by maximum likelihood
## Data: T_DF.NA3.max
## AIC BIC logLik
## 609.8407 627.4167 -297.9203
##
## Random effects:
## Formula: ~1 | ID
## (Intercept)
## StdDev: 7.410028
##
## Formula: ~1 | Session %in% ID
## (Intercept) Residual
## StdDev: 1.132564 5.234442
##
## Fixed effects: Total.Aggression ~ Test.l.GM * Sex_1
## Value Std.Error DF t-value p-value
## (Intercept) 28.171744 3.401748 43 8.281549 0.0000
## Test.l.GM 2.895538 1.787298 43 1.620064 0.1125
## Sex_1Male 3.139065 4.685347 10 0.669975 0.5180
## Test.l.GM:Sex_1Male -3.651545 2.419893 43 -1.508969 0.1386
## Correlation:
## (Intr) Ts..GM Sx_1Ml
## Test.l.GM 0.341
## Sex_1Male -0.726 -0.248
## Test.l.GM:Sex_1Male -0.252 -0.739 0.211
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.39348694 -0.48410852 -0.03356185 0.49387347 2.60556556
##
## Number of Observations: 91
## Number of Groups:
## ID Session %in% ID
## 12 46Total Aggression
# Sex 1 = Male , 0 = Female
# Model4.total.a <- lme(fixed=Total.Aggression~Test.l.GM,random=~1 |ID/Session,data=subset(T_DF.NA2,InterventionC=="Testosterone"), method="ML")
# summary(Model4.total.a)
#
# Model4.total.a <- lme(fixed=Total.Aggression~Test.l.GM*Pre*Sex_1,random=~1 |ID/Session,data=subset(T_DF.NA2,InterventionC=="Testosterone"), method="ML")
# summary(Model4.total.a)
#
# lsmeans(Model4.total.a, pairwise~Pre|Sex_1)
Model4 <- lme(fixed=Total.Aggression~Pre*Sex_1*InterventionC,random=~1 |ID/Session,data=T_DF.NA2, method="ML")
summary(Model4)## Linear mixed-effects model fit by maximum likelihood
## Data: T_DF.NA2
## AIC BIC logLik
## 605.5606 633.18 -291.7803
##
## Random effects:
## Formula: ~1 | ID
## (Intercept)
## StdDev: 7.621099
##
## Formula: ~1 | Session %in% ID
## (Intercept) Residual
## StdDev: 1.338002 4.768877
##
## Fixed effects: Total.Aggression ~ Pre * Sex_1 * InterventionC
## Value Std.Error DF
## (Intercept) 25.359792 3.617005 41
## PreBefore -1.818182 2.129200 41
## Sex_1Male 5.231249 5.116883 10
## InterventionCTestosterone 5.335768 2.144077 32
## PreBefore:Sex_1Male 2.579198 3.055619 41
## PreBefore:InterventionCTestosterone -3.874126 2.893013 41
## Sex_1Male:InterventionCTestosterone -5.244859 3.080167 32
## PreBefore:Sex_1Male:InterventionCTestosterone 5.385837 4.207889 41
## t-value p-value
## (Intercept) 7.011269 0.0000
## PreBefore -0.853927 0.3981
## Sex_1Male 1.022351 0.3307
## InterventionCTestosterone 2.488609 0.0182
## PreBefore:Sex_1Male 0.844084 0.4035
## PreBefore:InterventionCTestosterone -1.339132 0.1879
## Sex_1Male:InterventionCTestosterone -1.702784 0.0983
## PreBefore:Sex_1Male:InterventionCTestosterone 1.279938 0.2078
## Correlation:
## (Intr) PreBfr Sx_1Ml IntrCT
## PreBefore -0.294
## Sex_1Male -0.707 0.208
## InterventionCTestosterone -0.321 0.497 0.227
## PreBefore:Sex_1Male 0.205 -0.697 -0.290 -0.346
## PreBefore:InterventionCTestosterone 0.217 -0.736 -0.153 -0.675
## Sex_1Male:InterventionCTestosterone 0.224 -0.346 -0.313 -0.696
## PreBefore:Sex_1Male:InterventionCTestosterone -0.149 0.506 0.211 0.464
## PrB:S_1M PB:ICT S_1M:I
## PreBefore
## Sex_1Male
## InterventionCTestosterone
## PreBefore:Sex_1Male
## PreBefore:InterventionCTestosterone 0.513
## Sex_1Male:InterventionCTestosterone 0.482 0.470
## PreBefore:Sex_1Male:InterventionCTestosterone -0.726 -0.688 -0.673
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.4851087 -0.5981793 0.0109560 0.5627690 2.1926572
##
## Number of Observations: 91
## Number of Groups:
## ID Session %in% ID
## 12 46lsmeans(Model4, pairwise~Pre|Sex_1*InterventionC)## $lsmeans
## Sex_1 = Female, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 25.35979 3.617005 11 17.39882 33.32077
## Before 23.54161 3.617005 11 15.58064 31.50258
##
## Sex_1 = Male, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 30.59104 3.619360 10 22.52661 38.65548
## Before 31.35206 3.655052 10 23.20810 39.49602
##
## Sex_1 = Female, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 30.69556 3.563655 11 22.85201 38.53911
## Before 25.00325 3.563655 11 17.15970 32.84680
##
## Sex_1 = Male, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 30.68195 3.619360 10 22.61751 38.74639
## Before 32.95468 3.619360 10 24.89024 41.01911
##
## Confidence level used: 0.95
##
## $contrasts
## Sex_1 = Female, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before 1.818182 2.129200 41 0.854 0.3981
##
## Sex_1 = Male, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before -0.761016 2.191647 41 -0.347 0.7302
##
## Sex_1 = Female, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before 5.692308 1.958579 41 2.906 0.0059
##
## Sex_1 = Male, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before -2.272727 2.129200 41 -1.067 0.2920lsmip(Model4,InterventionC~Pre|Sex_1)
Physical Aggression
Model4 <- lme(fixed=Physical.Aggression~Pre*Sex_1*InterventionC,random=~1 |ID/Session,data=T_DF.NA2, method="ML")
summary(Model4)## Linear mixed-effects model fit by maximum likelihood
## Data: T_DF.NA2
## AIC BIC logLik
## 397.0568 424.6762 -187.5284
##
## Random effects:
## Formula: ~1 | ID
## (Intercept)
## StdDev: 2.490678
##
## Formula: ~1 | Session %in% ID
## (Intercept) Residual
## StdDev: 0.56272 1.475714
##
## Fixed effects: Physical.Aggression ~ Pre * Sex_1 * InterventionC
## Value Std.Error DF
## (Intercept) 4.524563 1.1767679 41
## PreBefore -0.272727 0.6588743 41
## Sex_1Male 4.214662 1.6647477 10
## InterventionCTestosterone 1.493114 0.6839883 32
## PreBefore:Sex_1Male 0.545731 0.9461480 41
## PreBefore:InterventionCTestosterone -1.342657 0.8952339 41
## Sex_1Male:InterventionCTestosterone -2.584024 0.9823838 32
## PreBefore:Sex_1Male:InterventionCTestosterone 1.342381 1.3025512 41
## t-value p-value
## (Intercept) 3.844907 0.0004
## PreBefore -0.413929 0.6811
## Sex_1Male 2.531712 0.0298
## InterventionCTestosterone 2.182953 0.0365
## PreBefore:Sex_1Male 0.576793 0.5672
## PreBefore:InterventionCTestosterone -1.499784 0.1413
## Sex_1Male:InterventionCTestosterone -2.630360 0.0130
## PreBefore:Sex_1Male:InterventionCTestosterone 1.030578 0.3088
## Correlation:
## (Intr) PreBfr Sx_1Ml IntrCT
## PreBefore -0.280
## Sex_1Male -0.707 0.198
## InterventionCTestosterone -0.315 0.482 0.223
## PreBefore:Sex_1Male 0.195 -0.696 -0.276 -0.335
## PreBefore:InterventionCTestosterone 0.206 -0.736 -0.146 -0.654
## Sex_1Male:InterventionCTestosterone 0.219 -0.335 -0.307 -0.696
## PreBefore:Sex_1Male:InterventionCTestosterone -0.142 0.506 0.200 0.450
## PrB:S_1M PB:ICT S_1M:I
## PreBefore
## Sex_1Male
## InterventionCTestosterone
## PreBefore:Sex_1Male
## PreBefore:InterventionCTestosterone 0.513
## Sex_1Male:InterventionCTestosterone 0.467 0.456
## PreBefore:Sex_1Male:InterventionCTestosterone -0.726 -0.687 -0.652
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.31287462 -0.62168809 0.04883832 0.60603591 2.23693095
##
## Number of Observations: 91
## Number of Groups:
## ID Session %in% ID
## 12 46lsmeans(Model4, pairwise~Pre|Sex_1*InterventionC)## $lsmeans
## Sex_1 = Female, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 4.524563 1.176768 11 1.934514 7.114612
## Before 4.251836 1.176768 11 1.661787 6.841884
##
## Sex_1 = Male, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 8.739225 1.177541 10 6.115501 11.362949
## Before 9.012229 1.188507 10 6.364070 11.660388
##
## Sex_1 = Female, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 6.017677 1.160085 11 3.464349 8.571006
## Before 4.402293 1.160085 11 1.848964 6.955622
##
## Sex_1 = Male, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 7.648316 1.177541 10 5.024592 10.272040
## Before 7.921043 1.177541 10 5.297319 10.544767
##
## Confidence level used: 0.95
##
## $contrasts
## Sex_1 = Female, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before 0.2727273 0.6588743 41 0.414 0.6811
##
## Sex_1 = Male, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before -0.2730041 0.6790292 41 -0.402 0.6897
##
## Sex_1 = Female, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before 1.6153846 0.6060762 41 2.665 0.0110
##
## Sex_1 = Male, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before -0.2727273 0.6588743 41 -0.414 0.6811lsmip(Model4,InterventionC~Pre|Sex_1)
Verbal Aggression
Model4 <- lme(fixed=Verbal.Aggression~Pre*Sex_1*InterventionC,random=~1 |ID/Session,data=T_DF.NA2, method="ML")
summary(Model4)## Linear mixed-effects model fit by maximum likelihood
## Data: T_DF.NA2
## AIC BIC logLik
## 401.3351 428.9545 -189.6675
##
## Random effects:
## Formula: ~1 | ID
## (Intercept)
## StdDev: 2.181625
##
## Formula: ~1 | Session %in% ID
## (Intercept) Residual
## StdDev: 0.7481456 1.494487
##
## Fixed effects: Verbal.Aggression ~ Pre * Sex_1 * InterventionC
## Value Std.Error DF
## (Intercept) 7.892168 1.0729000 41
## PreBefore -0.545455 0.6672558 41
## Sex_1Male 0.393946 1.5179686 10
## InterventionCTestosterone 1.275997 0.7240290 32
## PreBefore:Sex_1Male 1.230903 0.9589514 41
## PreBefore:InterventionCTestosterone -1.762238 0.9066222 41
## Sex_1Male:InterventionCTestosterone -0.639633 1.0397234 32
## PreBefore:Sex_1Male:InterventionCTestosterone 1.985881 1.3196786 41
## t-value p-value
## (Intercept) 7.355922 0.0000
## PreBefore -0.817459 0.4184
## Sex_1Male 0.259522 0.8005
## InterventionCTestosterone 1.762356 0.0876
## PreBefore:Sex_1Male 1.283592 0.2065
## PreBefore:InterventionCTestosterone -1.943740 0.0588
## Sex_1Male:InterventionCTestosterone -0.615196 0.5428
## PreBefore:Sex_1Male:InterventionCTestosterone 1.504821 0.1400
## Correlation:
## (Intr) PreBfr Sx_1Ml IntrCT
## PreBefore -0.311
## Sex_1Male -0.707 0.220
## InterventionCTestosterone -0.366 0.461 0.258
## PreBefore:Sex_1Male 0.216 -0.696 -0.306 -0.321
## PreBefore:InterventionCTestosterone 0.229 -0.736 -0.162 -0.626
## Sex_1Male:InterventionCTestosterone 0.255 -0.321 -0.356 -0.696
## PreBefore:Sex_1Male:InterventionCTestosterone -0.157 0.506 0.223 0.430
## PrB:S_1M PB:ICT S_1M:I
## PreBefore
## Sex_1Male
## InterventionCTestosterone
## PreBefore:Sex_1Male
## PreBefore:InterventionCTestosterone 0.512
## Sex_1Male:InterventionCTestosterone 0.447 0.436
## PreBefore:Sex_1Male:InterventionCTestosterone -0.727 -0.687 -0.624
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.43919401 -0.38927298 0.09737046 0.58049346 2.07676605
##
## Number of Observations: 91
## Number of Groups:
## ID Session %in% ID
## 12 46lsmeans(Model4, pairwise~Pre|Sex_1*InterventionC)## $lsmeans
## Sex_1 = Female, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 7.892168 1.072900 11 5.530731 10.253605
## Before 7.346714 1.072900 11 4.985277 9.708151
##
## Sex_1 = Male, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 8.286114 1.073832 10 5.893468 10.678760
## Before 8.971562 1.086829 10 6.549957 11.393167
##
## Sex_1 = Female, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 9.168165 1.052345 11 6.851970 11.484361
## Before 6.860473 1.052345 11 4.544278 9.176668
##
## Sex_1 = Male, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 8.922478 1.073832 10 6.529832 11.315123
## Before 9.831569 1.073832 10 7.438923 12.224214
##
## Confidence level used: 0.95
##
## $contrasts
## Sex_1 = Female, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before 0.5454545 0.6672558 41 0.817 0.4184
##
## Sex_1 = Male, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before -0.6854481 0.6887362 41 -0.995 0.3255
##
## Sex_1 = Female, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before 2.3076923 0.6137861 41 3.760 0.0005
##
## Sex_1 = Male, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before -0.9090909 0.6672558 41 -1.362 0.1805lsmip(Model4,InterventionC~Pre|Sex_1)
Anger
options(contrasts=c("contr.sum","contr.poly"))
Model4 <- lme(fixed=Anger~Pre*Sex_1*InterventionC,random=~1 |ID/Session,data=T_DF.NA2, method="ML")
summary(Model4)## Linear mixed-effects model fit by maximum likelihood
## Data: T_DF.NA2
## AIC BIC logLik
## 418.0813 445.7008 -198.0407
##
## Random effects:
## Formula: ~1 | ID
## (Intercept)
## StdDev: 2.319354
##
## Formula: ~1 | Session %in% ID
## (Intercept) Residual
## StdDev: 0.6210482 1.707768
##
## Fixed effects: Anger ~ Pre * Sex_1 * InterventionC
## Value Std.Error DF t-value p-value
## (Intercept) 6.701245 0.7331684 41 9.140117 0.0000
## Pre1 0.060202 0.1884060 41 0.319535 0.7509
## Sex_11 -0.255319 0.7331684 10 -0.348241 0.7349
## InterventionC1 -0.414366 0.2131753 32 -1.943779 0.0608
## Pre1:Sex_11 0.175812 0.1884060 41 0.933153 0.3562
## Pre1:InterventionC1 0.210552 0.1884060 41 1.117544 0.2703
## Sex_11:InterventionC1 0.015480 0.2131753 32 0.072615 0.9426
## Pre1:Sex_11:InterventionC1 -0.128384 0.1884060 41 -0.681423 0.4994
## Correlation:
## (Intr) Pre1 Sex_11 IntrC1 Pr1:S_11 P1:IC1
## Pre1 -0.004
## Sex_11 -0.007 0.004
## InterventionC1 0.015 -0.014 0.008
## Pre1:Sex_11 0.004 -0.055 -0.004 0.014
## Pre1:InterventionC1 -0.004 0.055 0.004 -0.014 0.024
## Sex_11:InterventionC1 0.008 0.014 0.015 -0.036 -0.014 0.014
## Pre1:Sex_11:InterventionC1 0.004 0.024 -0.004 0.014 0.055 -0.055
## S_11:I
## Pre1
## Sex_11
## InterventionC1
## Pre1:Sex_11
## Pre1:InterventionC1
## Sex_11:InterventionC1
## Pre1:Sex_11:InterventionC1 -0.014
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.28620697 -0.45611486 -0.09021087 0.54395482 2.06876022
##
## Number of Observations: 91
## Number of Groups:
## ID Session %in% ID
## 12 46anova(Model4)## numDF denDF F-value p-value
## (Intercept) 1 41 84.13264 <.0001
## Pre 1 41 0.09429 0.7603
## Sex_1 1 10 0.11202 0.7448
## InterventionC 1 32 3.73411 0.0622
## Pre:Sex_1 1 41 0.89173 0.3505
## Pre:InterventionC 1 41 1.16835 0.2861
## Sex_1:InterventionC 1 32 0.00399 0.9500
## Pre:Sex_1:InterventionC 1 41 0.46434 0.4994lsmeans(Model4, pairwise~Pre|Sex_1*InterventionC)## $lsmeans
## Sex_1 = Female, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 6.365222 1.146921 10 3.809722 8.920722
## Before 5.728858 1.146921 10 3.173359 8.284358
##
## Sex_1 = Male, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 6.750046 1.147886 10 4.192396 9.307695
## Before 6.303392 1.162792 10 3.712530 8.894254
##
## Sex_1 = Female, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 6.998658 1.124212 10 4.493759 9.503558
## Before 6.690966 1.124212 10 4.186066 9.195866
##
## Sex_1 = Male, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 6.931864 1.147886 10 4.374215 9.489513
## Before 7.840955 1.147886 10 5.283305 10.398604
##
## Confidence level used: 0.95
##
## $contrasts
## Sex_1 = Female, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before 0.6363636 0.7624811 41 0.835 0.4088
##
## Sex_1 = Male, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before 0.4466539 0.7855616 41 0.569 0.5727
##
## Sex_1 = Female, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before 0.3076923 0.7013806 41 0.439 0.6632
##
## Sex_1 = Male, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before -0.9090909 0.7624811 41 -1.192 0.2400lsmip(Model4,InterventionC~Pre|Sex_1)
Hostility
Model4 <- lme(fixed=Hostility~Pre*Sex_1*InterventionC,random=~1 |ID/Session,data=T_DF.NA2, method="ML")
summary(Model4)## Linear mixed-effects model fit by maximum likelihood
## Data: T_DF.NA2
## AIC BIC logLik
## 383.4122 411.0317 -180.7061
##
## Random effects:
## Formula: ~1 | ID
## (Intercept)
## StdDev: 2.149448
##
## Formula: ~1 | Session %in% ID
## (Intercept) Residual
## StdDev: 1.046028 1.195227
##
## Fixed effects: Hostility ~ Pre * Sex_1 * InterventionC
## Value Std.Error DF t-value p-value
## (Intercept) 7.100458 0.6834277 41 10.389479 0.0000
## Pre1 0.174749 0.1321078 41 1.322779 0.1932
## Sex_11 -0.012620 0.6834277 10 -0.018465 0.9856
## InterventionC1 -0.430935 0.2106004 32 -2.046222 0.0490
## Pre1:Sex_11 0.281544 0.1321078 41 2.131172 0.0391
## Pre1:InterventionC1 -0.145181 0.1321078 41 -1.098956 0.2782
## Sex_11:InterventionC1 -0.265003 0.2106004 32 -1.258320 0.2174
## Pre1:Sex_11:InterventionC1 -0.129295 0.1321078 41 -0.978707 0.3335
## Correlation:
## (Intr) Pre1 Sex_11 IntrC1 Pr1:S_11 P1:IC1
## Pre1 -0.003
## Sex_11 -0.007 0.003
## InterventionC1 0.015 -0.012 0.010
## Pre1:Sex_11 0.003 -0.059 -0.003 0.012
## Pre1:InterventionC1 -0.003 0.059 0.003 -0.012 0.020
## Sex_11:InterventionC1 0.010 0.012 0.015 -0.031 -0.012 0.012
## Pre1:Sex_11:InterventionC1 0.003 0.020 -0.003 0.012 0.059 -0.059
## S_11:I
## Pre1
## Sex_11
## InterventionC1
## Pre1:Sex_11
## Pre1:InterventionC1
## Sex_11:InterventionC1
## Pre1:Sex_11:InterventionC1 -0.012
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.51342419 -0.43631181 0.01467211 0.43841887 2.12039068
##
## Number of Observations: 91
## Number of Groups:
## ID Session %in% ID
## 12 46anova(Model4)## numDF denDF F-value p-value
## (Intercept) 1 41 108.84685 <.0001
## Pre 1 41 2.38619 0.1301
## Sex_1 1 10 0.00091 0.9765
## InterventionC 1 32 4.46537 0.0425
## Pre:Sex_1 1 41 4.86353 0.0331
## Pre:InterventionC 1 41 1.30833 0.2593
## Sex_1:InterventionC 1 32 1.61373 0.2131
## Pre:Sex_1:InterventionC 1 41 0.95787 0.3335lsmeans(Model4, pairwise~Pre|Sex_1*InterventionC)## $lsmeans
## Sex_1 = Female, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 6.573719 1.048283 10 4.237999 8.909439
## Before 6.210083 1.048283 10 3.874363 8.545803
##
## Sex_1 = Male, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 6.824465 1.049285 10 4.486512 9.162417
## Before 7.069826 1.059199 10 4.709783 9.429869
##
## Sex_1 = Female, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 8.514545 1.029223 10 6.221293 10.807798
## Before 7.053007 1.029223 10 4.759754 9.346260
##
## Sex_1 = Male, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 7.188101 1.049285 10 4.850149 9.526053
## Before 7.369919 1.049285 10 5.031967 9.707871
##
## Confidence level used: 0.95
##
## $contrasts
## Sex_1 = Female, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before 0.3636364 0.5336429 41 0.681 0.4994
##
## Sex_1 = Male, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before -0.2453615 0.5535745 41 -0.443 0.6599
##
## Sex_1 = Female, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before 1.4615385 0.4908801 41 2.977 0.0049
##
## Sex_1 = Male, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before -0.1818182 0.5336429 41 -0.341 0.7351lsmip(Model4,InterventionC~Pre|Sex_1)
#Total Risk
Model4 <- lme(fixed=Total.Risk~Pre*Sex_1*InterventionC,random=~1 |ID/Session,data=T_DF.NA2, method="ML")
summary(Model4)## Linear mixed-effects model fit by maximum likelihood
## Data: T_DF.NA2
## AIC BIC logLik
## 735.8626 763.4821 -356.9313
##
## Random effects:
## Formula: ~1 | ID
## (Intercept)
## StdDev: 16.18329
##
## Formula: ~1 | Session %in% ID
## (Intercept) Residual
## StdDev: 0.8562673 9.984673
##
## Fixed effects: Total.Risk ~ Pre * Sex_1 * InterventionC
## Value Std.Error DF t-value p-value
## (Intercept) 122.88101 5.019526 41 24.480602 0.0000
## Pre1 1.19823 1.100821 41 1.088486 0.2827
## Sex_11 -8.37550 5.019526 10 -1.668585 0.1262
## InterventionC1 -0.40468 1.117786 32 -0.362041 0.7197
## Pre1:Sex_11 3.32275 1.100821 41 3.018430 0.0044
## Pre1:InterventionC1 -1.50457 1.100821 41 -1.366770 0.1791
## Sex_11:InterventionC1 -1.02921 1.117786 32 -0.920755 0.3641
## Pre1:Sex_11:InterventionC1 -0.74368 1.100821 41 -0.675571 0.5031
## Correlation:
## (Intr) Pre1 Sex_11 IntrC1 Pr1:S_11 P1:IC1
## Pre1 -0.003
## Sex_11 -0.004 0.003
## InterventionC1 0.012 -0.014 0.006
## Pre1:Sex_11 0.003 -0.054 -0.003 0.014
## Pre1:InterventionC1 -0.003 0.054 0.003 -0.014 0.025
## Sex_11:InterventionC1 0.006 0.014 0.012 -0.037 -0.014 0.014
## Pre1:Sex_11:InterventionC1 0.003 0.025 -0.003 0.014 0.054 -0.054
## S_11:I
## Pre1
## Sex_11
## InterventionC1
## Pre1:Sex_11
## Pre1:InterventionC1
## Sex_11:InterventionC1
## Pre1:Sex_11:InterventionC1 -0.014
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.58275535 -0.54790763 0.06764513 0.42338187 3.94295963
##
## Number of Observations: 91
## Number of Groups:
## ID Session %in% ID
## 12 46anova(Model4)## numDF denDF F-value p-value
## (Intercept) 1 41 599.2693 <.0001
## Pre 1 41 1.8888 0.1768
## Sex_1 1 10 2.7057 0.1310
## InterventionC 1 32 0.1991 0.6585
## Pre:Sex_1 1 41 9.5322 0.0036
## Pre:InterventionC 1 41 1.9399 0.1712
## Sex_1:InterventionC 1 32 0.8658 0.3591
## Pre:Sex_1:InterventionC 1 41 0.4564 0.5031lsmeans(Model4, pairwise~Pre|Sex_1*InterventionC)## $lsmeans
## Sex_1 = Female, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 115.3443 7.613207 10 98.38106 132.3076
## Before 110.7989 7.613207 10 93.83561 127.7622
##
## Sex_1 = Male, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 128.9956 7.617606 10 112.02255 145.9687
## Before 134.7664 7.687990 10 117.63654 151.8964
##
## Sex_1 = Female, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 122.7086 7.509607 10 105.97618 139.4411
## Before 109.1702 7.509607 10 92.43772 125.9026
##
## Sex_1 = Male, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 129.2684 7.617606 10 112.29527 146.2414
## Before 131.9956 7.617606 10 115.02255 148.9687
##
## Confidence level used: 0.95
##
## $contrasts
## Sex_1 = Female, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before 4.545455 4.457939 41 1.020 0.3139
##
## Sex_1 = Male, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before -5.770820 4.581853 41 -1.259 0.2150
##
## Sex_1 = Female, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before 13.538462 4.100708 41 3.301 0.0020
##
## Sex_1 = Male, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before -2.727273 4.457939 41 -0.612 0.5441lsmip(Model4,InterventionC~Pre|Sex_1)
Financial Risk
Model4 <- lme(fixed=Finacial.Risk~Pre*Sex_1*InterventionC,random=~1 |ID/Session,data=T_DF.NA2, method="ML")
summary(Model4)## Linear mixed-effects model fit by maximum likelihood
## Data: T_DF.NA2
## AIC BIC logLik
## 503.375 530.9945 -240.6875
##
## Random effects:
## Formula: ~1 | ID
## (Intercept)
## StdDev: 3.597261
##
## Formula: ~1 | Session %in% ID
## (Intercept) Residual
## StdDev: 1.220844 2.672917
##
## Fixed effects: Finacial.Risk ~ Pre * Sex_1 * InterventionC
## Value Std.Error DF t-value p-value
## (Intercept) 17.847788 1.1439204 41 15.602299 0.0000
## Pre1 -0.117701 0.2949821 41 -0.399009 0.6920
## Sex_11 -2.883152 1.1439204 10 -2.520413 0.0304
## InterventionC1 -0.466859 0.3530446 32 -1.322380 0.1954
## Pre1:Sex_11 0.463854 0.2949821 41 1.572483 0.1235
## Pre1:InterventionC1 -0.259309 0.2949821 41 -0.879067 0.3845
## Sex_11:InterventionC1 -0.457505 0.3530446 32 -1.295884 0.2043
## Pre1:Sex_11:InterventionC1 -0.086845 0.2949821 41 -0.294408 0.7699
## Correlation:
## (Intr) Pre1 Sex_11 IntrC1 Pr1:S_11 P1:IC1
## Pre1 -0.004
## Sex_11 -0.007 0.004
## InterventionC1 0.016 -0.014 0.009
## Pre1:Sex_11 0.004 -0.056 -0.004 0.014
## Pre1:InterventionC1 -0.004 0.056 0.004 -0.014 0.023
## Sex_11:InterventionC1 0.009 0.014 0.016 -0.035 -0.014 0.014
## Pre1:Sex_11:InterventionC1 0.004 0.023 -0.004 0.014 0.056 -0.056
## S_11:I
## Pre1
## Sex_11
## InterventionC1
## Pre1:Sex_11
## Pre1:InterventionC1
## Sex_11:InterventionC1
## Pre1:Sex_11:InterventionC1 -0.014
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.7551738 -0.3660375 -0.1168382 0.3643547 3.9662413
##
## Number of Observations: 91
## Number of Groups:
## ID Session %in% ID
## 12 46anova(Model4)## numDF denDF F-value p-value
## (Intercept) 1 41 243.71657 <.0001
## Pre 1 41 0.05741 0.8118
## Sex_1 1 10 6.14615 0.0326
## InterventionC 1 32 1.94968 0.1722
## Pre:Sex_1 1 41 2.55483 0.1176
## Pre:InterventionC 1 41 0.77459 0.3839
## Sex_1:InterventionC 1 32 1.69009 0.2029
## Pre:Sex_1:InterventionC 1 41 0.08668 0.7699lsmeans(Model4, pairwise~Pre|Sex_1*InterventionC)## $lsmeans
## Sex_1 = Female, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 14.04027 1.798399 10 10.03319 18.04736
## Before 14.04027 1.798399 10 10.03319 18.04736
##
## Sex_1 = Male, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 19.96757 1.800037 10 15.95683 23.97830
## Before 21.47560 1.824318 10 17.41077 25.54044
##
## Sex_1 = Female, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 16.58131 1.760476 10 12.65872 20.50389
## Before 15.19669 1.760476 10 11.27411 19.11928
##
## Sex_1 = Male, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 20.33120 1.800037 10 16.32047 24.34194
## Before 21.14938 1.800037 10 17.13865 25.16012
##
## Confidence level used: 0.95
##
## $contrasts
## Sex_1 = Female, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before -1.332268e-15 1.193399 41 0.000 1.0000
##
## Sex_1 = Male, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before -1.508038e+00 1.231027 41 -1.225 0.2276
##
## Sex_1 = Female, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before 1.384615e+00 1.097768 41 1.261 0.2143
##
## Sex_1 = Male, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before -8.181818e-01 1.193399 41 -0.686 0.4968lsmip(Model4,InterventionC~Pre|Sex_1)
Ethical Risk
Model4 <- lme(fixed=Ethical.Risk~Pre*Sex_1*InterventionC,random=~1 |ID/Session,data=T_DF.NA2, method="ML")
summary(Model4)## Linear mixed-effects model fit by maximum likelihood
## Data: T_DF.NA2
## AIC BIC logLik
## 517.5502 545.1696 -247.7751
##
## Random effects:
## Formula: ~1 | ID
## (Intercept)
## StdDev: 3.955385
##
## Formula: ~1 | Session %in% ID
## (Intercept) Residual
## StdDev: 0.0003743887 3.108673
##
## Fixed effects: Ethical.Risk ~ Pre * Sex_1 * InterventionC
## Value Std.Error DF t-value p-value
## (Intercept) 19.427060 1.2451138 41 15.602637 0.0000
## Pre1 0.479160 0.3427107 41 1.398149 0.1696
## Sex_11 -1.460429 1.2451138 10 -1.172928 0.2680
## InterventionC1 -0.263292 0.3454198 32 -0.762237 0.4515
## Pre1:Sex_11 0.648462 0.3427107 41 1.892156 0.0655
## Pre1:InterventionC1 -0.603007 0.3427107 41 -1.759523 0.0859
## Sex_11:InterventionC1 -0.387139 0.3454198 32 -1.120778 0.2707
## Pre1:Sex_11:InterventionC1 -0.115524 0.3427107 41 -0.337089 0.7378
## Correlation:
## (Intr) Pre1 Sex_11 IntrC1 Pr1:S_11 P1:IC1
## Pre1 -0.004
## Sex_11 -0.006 0.004
## InterventionC1 0.015 -0.014 0.008
## Pre1:Sex_11 0.004 -0.054 -0.004 0.014
## Pre1:InterventionC1 -0.004 0.054 0.004 -0.014 0.025
## Sex_11:InterventionC1 0.008 0.014 0.015 -0.037 -0.014 0.014
## Pre1:Sex_11:InterventionC1 0.004 0.025 -0.004 0.014 0.054 -0.054
## S_11:I
## Pre1
## Sex_11
## InterventionC1
## Pre1:Sex_11
## Pre1:InterventionC1
## Sex_11:InterventionC1
## Pre1:Sex_11:InterventionC1 -0.014
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.5973768 -0.4497526 -0.1082360 0.4969251 3.0938972
##
## Number of Observations: 91
## Number of Groups:
## ID Session %in% ID
## 12 46anova(Model4)## numDF denDF F-value p-value
## (Intercept) 1 41 243.74789 <.0001
## Pre 1 41 2.65428 0.1109
## Sex_1 1 10 1.29627 0.2814
## InterventionC 1 32 0.72052 0.4023
## Pre:Sex_1 1 41 3.79764 0.0582
## Pre:InterventionC 1 41 3.11602 0.0850
## Sex_1:InterventionC 1 32 1.26715 0.2687
## Pre:Sex_1:InterventionC 1 41 0.11363 0.7378lsmeans(Model4, pairwise~Pre|Sex_1*InterventionC)## $lsmeans
## Sex_1 = Female, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 17.72529 1.957243 10 13.36428 22.08630
## Before 16.90711 1.957243 10 12.54610 21.26812
##
## Sex_1 = Male, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 20.35455 1.958758 10 15.99017 24.71894
## Before 21.66812 1.985030 10 17.24520 26.09104
##
## Sex_1 = Female, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 20.46322 1.918366 10 16.18883 24.73760
## Before 16.77091 1.918366 10 12.49652 21.04529
##
## Sex_1 = Male, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 21.08182 1.958758 10 16.71744 25.44621
## Before 20.44546 1.958758 10 16.08108 24.80984
##
## Confidence level used: 0.95
##
## $contrasts
## Sex_1 = Female, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before 0.8181818 1.387955 41 0.589 0.5588
##
## Sex_1 = Male, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before -1.3135695 1.426168 41 -0.921 0.3624
##
## Sex_1 = Female, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before 3.6923077 1.276733 41 2.892 0.0061
##
## Sex_1 = Male, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before 0.6363636 1.387955 41 0.458 0.6490lsmip(Model4,InterventionC~Pre|Sex_1)
Health_Safety risk
Model4 <- lme(fixed=Health_Safety.Risk~Pre*Sex_1*InterventionC,random=~1 |ID/Session,data=T_DF.NA2, method="ML")
summary(Model4)## Linear mixed-effects model fit by maximum likelihood
## Data: T_DF.NA2
## AIC BIC logLik
## 486.8408 514.4603 -232.4204
##
## Random effects:
## Formula: ~1 | ID
## (Intercept)
## StdDev: 4.324106
##
## Formula: ~1 | Session %in% ID
## (Intercept) Residual
## StdDev: 0.8125637 2.42408
##
## Fixed effects: Health_Safety.Risk ~ Pre * Sex_1 * InterventionC
## Value Std.Error DF t-value p-value
## (Intercept) 22.366692 1.3410361 41 16.678665 0.0000
## Pre1 0.337761 0.2674136 41 1.263066 0.2137
## Sex_11 -0.392460 1.3410361 10 -0.292654 0.7758
## InterventionC1 0.046093 0.2979223 32 0.154716 0.8780
## Pre1:Sex_11 0.489162 0.2674136 41 1.829235 0.0746
## Pre1:InterventionC1 -0.125526 0.2674136 41 -0.469407 0.6413
## Sex_11:InterventionC1 -0.105308 0.2979223 32 -0.353475 0.7261
## Pre1:Sex_11:InterventionC1 -0.201397 0.2674136 41 -0.753130 0.4557
## Correlation:
## (Intr) Pre1 Sex_11 IntrC1 Pr1:S_11 P1:IC1
## Pre1 -0.003
## Sex_11 -0.004 0.003
## InterventionC1 0.012 -0.014 0.006
## Pre1:Sex_11 0.003 -0.055 -0.003 0.014
## Pre1:InterventionC1 -0.003 0.055 0.003 -0.014 0.024
## Sex_11:InterventionC1 0.006 0.014 0.012 -0.035 -0.014 0.014
## Pre1:Sex_11:InterventionC1 0.003 0.024 -0.003 0.014 0.055 -0.055
## S_11:I
## Pre1
## Sex_11
## InterventionC1
## Pre1:Sex_11
## Pre1:InterventionC1
## Sex_11:InterventionC1
## Pre1:Sex_11:InterventionC1 -0.014
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.69178992 -0.42644215 -0.01037085 0.48637162 3.67093205
##
## Number of Observations: 91
## Number of Groups:
## ID Session %in% ID
## 12 46anova(Model4)## numDF denDF F-value p-value
## (Intercept) 1 41 278.19743 <.0001
## Pre 1 41 2.04751 0.1600
## Sex_1 1 10 0.08136 0.7813
## InterventionC 1 32 0.01473 0.9042
## Pre:Sex_1 1 41 3.54650 0.0668
## Pre:InterventionC 1 41 0.25685 0.6150
## Sex_1:InterventionC 1 32 0.13253 0.7182
## Pre:Sex_1:InterventionC 1 41 0.56721 0.4557lsmeans(Model4, pairwise~Pre|Sex_1*InterventionC)## $lsmeans
## Sex_1 = Female, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 22.41502 2.018634 10 17.91722 26.91281
## Before 21.41502 2.018634 10 16.91722 25.91281
##
## Sex_1 = Male, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 22.83502 2.019814 10 18.33460 27.33545
## Before 22.98608 2.036733 10 18.44796 27.52421
##
## Sex_1 = Female, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 23.18729 1.993176 10 18.74622 27.62836
## Before 20.87960 1.993176 10 16.43853 25.32067
##
## Sex_1 = Male, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 22.38048 2.019814 10 17.88005 26.88090
## Before 22.83502 2.019814 10 18.33460 27.33545
##
## Confidence level used: 0.95
##
## $contrasts
## Sex_1 = Female, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before 1.0000000 1.0822992 41 0.924 0.3609
##
## Sex_1 = Male, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before -0.1510600 1.1147811 41 -0.136 0.8929
##
## Sex_1 = Female, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before 2.3076923 0.9955705 41 2.318 0.0255
##
## Sex_1 = Male, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before -0.4545455 1.0822992 41 -0.420 0.6767lsmip(Model4,InterventionC~Pre|Sex_1)
Model4 <- lme(fixed=Recreational.Risk~Pre*Sex_1*InterventionC,random=~1 |ID/Session,data=T_DF.NA2, method="ML")
summary(Model4)## Linear mixed-effects model fit by maximum likelihood
## Data: T_DF.NA2
## AIC BIC logLik
## 513.192 540.8115 -245.596
##
## Random effects:
## Formula: ~1 | ID
## (Intercept)
## StdDev: 6.473518
##
## Formula: ~1 | Session %in% ID
## (Intercept) Residual
## StdDev: 1.682729 2.442329
##
## Fixed effects: Recreational.Risk ~ Pre * Sex_1 * InterventionC
## Value Std.Error DF t-value p-value
## (Intercept) 32.88364 1.9934830 41 16.495572 0.0000
## Pre1 0.16713 0.2697859 41 0.619484 0.5390
## Sex_11 -1.75055 1.9934830 10 -0.878137 0.4005
## InterventionC1 0.19961 0.3781087 32 0.527904 0.6012
## Pre1:Sex_11 0.87833 0.2697859 41 3.255643 0.0023
## Pre1:InterventionC1 -0.21924 0.2697859 41 -0.812628 0.4211
## Sex_11:InterventionC1 -0.10250 0.3781087 32 -0.271093 0.7881
## Pre1:Sex_11:InterventionC1 -0.23531 0.2697859 41 -0.872209 0.3882
## Correlation:
## (Intr) Pre1 Sex_11 IntrC1 Pr1:S_11 P1:IC1
## Pre1 -0.002
## Sex_11 -0.003 0.002
## InterventionC1 0.009 -0.013 0.006
## Pre1:Sex_11 0.002 -0.057 -0.002 0.013
## Pre1:InterventionC1 -0.002 0.057 0.002 -0.013 0.021
## Sex_11:InterventionC1 0.006 0.013 0.009 -0.032 -0.013 0.013
## Pre1:Sex_11:InterventionC1 0.002 0.021 -0.002 0.013 0.057 -0.057
## S_11:I
## Pre1
## Sex_11
## InterventionC1
## Pre1:Sex_11
## Pre1:InterventionC1
## Sex_11:InterventionC1
## Pre1:Sex_11:InterventionC1 -0.013
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.365962669 -0.552012559 0.006438086 0.498206336 2.643651873
##
## Number of Observations: 91
## Number of Groups:
## ID Session %in% ID
## 12 46anova(Model4)## numDF denDF F-value p-value
## (Intercept) 1 41 271.78081 <.0001
## Pre 1 41 0.81537 0.3718
## Sex_1 1 10 0.76032 0.4037
## InterventionC 1 32 0.23026 0.6346
## Pre:Sex_1 1 41 11.09037 0.0018
## Pre:InterventionC 1 41 0.74097 0.3944
## Sex_1:InterventionC 1 32 0.07978 0.7794
## Pre:Sex_1:InterventionC 1 41 0.76075 0.3882lsmeans(Model4, pairwise~Pre|Sex_1*InterventionC)## $lsmeans
## Sex_1 = Female, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 31.82110 2.923242 10 25.30771 38.33449
## Before 30.63929 2.923242 10 24.12590 37.15267
##
## Sex_1 = Male, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 34.24118 2.924605 10 27.72475 40.75760
## Before 35.63143 2.938509 10 29.08402 42.17883
##
## Sex_1 = Female, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 32.53599 2.899541 10 26.07541 38.99657
## Before 29.53599 2.899541 10 23.07541 35.99657
##
## Sex_1 = Male, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 33.60481 2.924605 10 27.08839 40.12124
## Before 35.05936 2.924605 10 28.54293 41.57579
##
## Confidence level used: 0.95
##
## $contrasts
## Sex_1 = Female, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before 1.181818 1.090447 41 1.084 0.2848
##
## Sex_1 = Male, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before -1.390249 1.128679 41 -1.232 0.2251
##
## Sex_1 = Female, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before 3.000000 1.003065 41 2.991 0.0047
##
## Sex_1 = Male, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before -1.454545 1.090447 41 -1.334 0.1896lsmip(Model4,InterventionC~Pre|Sex_1)
Model4 <- lme(fixed=Social~Pre*Sex_1*InterventionC,random=~1 |ID/Session,data=T_DF.NA2, method="ML") #Social Risk
summary(Model4)## Linear mixed-effects model fit by maximum likelihood
## Data: T_DF.NA2
## AIC BIC logLik
## 468.0657 495.6851 -223.0328
##
## Random effects:
## Formula: ~1 | ID
## (Intercept)
## StdDev: 3.935943
##
## Formula: ~1 | Session %in% ID
## (Intercept) Residual
## StdDev: 1.574873 1.890705
##
## Fixed effects: Social ~ Pre * Sex_1 * InterventionC
## Value Std.Error DF t-value p-value
## (Intercept) 30.365879 1.2335552 41 24.616555 0.0000
## Pre1 0.320071 0.2089532 41 1.531782 0.1333
## Sex_11 -1.902173 1.2335552 10 -1.542025 0.1541
## InterventionC1 0.072304 0.3236451 32 0.223404 0.8246
## Pre1:Sex_11 0.854754 0.2089532 41 4.090649 0.0002
## Pre1:InterventionC1 -0.309300 0.2089532 41 -1.480235 0.1465
## Sex_11:InterventionC1 -0.007834 0.3236451 32 -0.024207 0.9808
## Pre1:Sex_11:InterventionC1 -0.092798 0.2089532 41 -0.444110 0.6593
## Correlation:
## (Intr) Pre1 Sex_11 IntrC1 Pr1:S_11 P1:IC1
## Pre1 -0.003
## Sex_11 -0.005 0.003
## InterventionC1 0.013 -0.012 0.009
## Pre1:Sex_11 0.003 -0.058 -0.003 0.012
## Pre1:InterventionC1 -0.003 0.058 0.003 -0.012 0.020
## Sex_11:InterventionC1 0.009 0.012 0.013 -0.031 -0.012 0.012
## Pre1:Sex_11:InterventionC1 0.003 0.020 -0.003 0.012 0.058 -0.058
## S_11:I
## Pre1
## Sex_11
## InterventionC1
## Pre1:Sex_11
## Pre1:InterventionC1
## Sex_11:InterventionC1
## Pre1:Sex_11:InterventionC1 -0.012
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -1.43182407 -0.51978560 -0.02934255 0.57896508 1.94272833
##
## Number of Observations: 91
## Number of Groups:
## ID Session %in% ID
## 12 46anova(Model4)## numDF denDF F-value p-value
## (Intercept) 1 41 605.1353 <.0001
## Pre 1 41 3.6020 0.0648
## Sex_1 1 10 2.3383 0.1572
## InterventionC 1 32 0.0265 0.8717
## Pre:Sex_1 1 41 17.3096 0.0002
## Pre:InterventionC 1 41 2.2756 0.1391
## Sex_1:InterventionC 1 32 0.0009 0.9765
## Pre:Sex_1:InterventionC 1 41 0.1972 0.6593lsmeans(Model4, pairwise~Pre|Sex_1*InterventionC)## $lsmeans
## Sex_1 = Female, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 29.30090 1.855291 10 25.16706 33.43475
## Before 27.75545 1.855291 10 23.62160 31.88929
##
## Sex_1 = Male, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 31.59701 1.856729 10 27.45995 35.73406
## Before 33.09938 1.870565 10 28.93150 37.26725
##
## Sex_1 = Female, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 29.97616 1.829493 10 25.89980 34.05252
## Before 26.82231 1.829493 10 22.74595 30.89868
##
## Sex_1 = Male, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 31.86973 1.856729 10 27.73268 36.00678
## Before 32.50610 1.856729 10 28.36905 36.64315
##
## Confidence level used: 0.95
##
## $contrasts
## Sex_1 = Female, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before 1.5454545 0.8441587 41 1.831 0.0744
##
## Sex_1 = Male, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before -1.5023703 0.8753013 41 -1.716 0.0936
##
## Sex_1 = Female, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before 3.1538462 0.7765130 41 4.062 0.0002
##
## Sex_1 = Male, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before -0.6363636 0.8441587 41 -0.754 0.4553lsmip(Model4,InterventionC~Pre|Sex_1)
Model4 <- lme(fixed=Empathy~Pre*Sex_1*InterventionC,random=~1 |ID/Session,data=T_DF.NA2, method="ML")
summary(Model4)## Linear mixed-effects model fit by maximum likelihood
## Data: T_DF.NA2
## AIC BIC logLik
## 569.3296 596.9491 -273.6648
##
## Random effects:
## Formula: ~1 | ID
## (Intercept)
## StdDev: 12.22933
##
## Formula: ~1 | Session %in% ID
## (Intercept) Residual
## StdDev: 2.996947 2.844551
##
## Fixed effects: Empathy ~ Pre * Sex_1 * InterventionC
## Value Std.Error DF t-value p-value
## (Intercept) 37.22673 3.740190 41 9.953165 0.0000
## Pre1 -0.92165 0.314583 41 -2.929748 0.0055
## Sex_11 4.21454 3.740190 10 1.126825 0.2861
## InterventionC1 -0.14401 0.565215 32 -0.254789 0.8005
## Pre1:Sex_11 -0.14129 0.314583 41 -0.449127 0.6557
## Pre1:InterventionC1 -0.42689 0.314583 41 -1.357015 0.1822
## Sex_11:InterventionC1 -0.56074 0.565215 32 -0.992075 0.3286
## Pre1:Sex_11:InterventionC1 -0.32835 0.314583 41 -1.043765 0.3027
## Correlation:
## (Intr) Pre1 Sex_11 IntrC1 Pr1:S_11 P1:IC1
## Pre1 -0.002
## Sex_11 -0.002 0.002
## InterventionC1 0.007 -0.011 0.005
## Pre1:Sex_11 0.002 -0.060 -0.002 0.011
## Pre1:InterventionC1 -0.002 0.060 0.002 -0.011 0.019
## Sex_11:InterventionC1 0.005 0.011 0.007 -0.029 -0.011 0.011
## Pre1:Sex_11:InterventionC1 0.002 0.019 -0.002 0.011 0.060 -0.060
## S_11:I
## Pre1
## Sex_11
## InterventionC1
## Pre1:Sex_11
## Pre1:InterventionC1
## Sex_11:InterventionC1
## Pre1:Sex_11:InterventionC1 -0.011
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.0199431541 -0.5239276791 0.0003454672 0.4477106074 2.3362976533
##
## Number of Observations: 91
## Number of Groups:
## ID Session %in% ID
## 12 46anova(Model4)## numDF denDF F-value p-value
## (Intercept) 1 41 99.18197 <.0001
## Pre 1 41 8.13204 0.0068
## Sex_1 1 10 1.28879 0.2827
## InterventionC 1 32 0.07990 0.7793
## Pre:Sex_1 1 41 0.13452 0.7157
## Pre:InterventionC 1 41 1.99129 0.1657
## Sex_1:InterventionC 1 32 1.00812 0.3229
## Pre:Sex_1:InterventionC 1 41 1.08944 0.3027lsmeans(Model4, pairwise~Pre|Sex_1*InterventionC)## $lsmeans
## Sex_1 = Female, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 38.91834 5.390290 10 26.90803 50.92866
## Before 42.55471 5.390290 10 30.54439 54.56502
##
## Sex_1 = Male, InterventionC = Placebo:
## Pre lsmean SE df lower.CL upper.CL
## After 32.55001 5.392005 10 20.53587 44.56414
## Before 34.30782 5.403551 10 22.26796 46.34768
##
## Sex_1 = Female, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 41.83832 5.365291 10 29.88371 53.79294
## Before 42.45371 5.365291 10 30.49910 54.40832
##
## Sex_1 = Male, InterventionC = Testosterone:
## Pre lsmean SE df lower.CL upper.CL
## After 31.91364 5.392005 10 19.89951 43.92778
## Before 33.27728 5.392005 10 21.26315 45.29142
##
## Confidence level used: 0.95
##
## $contrasts
## Sex_1 = Female, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before -3.6363636 1.270030 41 -2.863 0.0066
##
## Sex_1 = Male, InterventionC = Placebo:
## contrast estimate SE df t.ratio p.value
## After - Before -1.7578093 1.320161 41 -1.332 0.1904
##
## Sex_1 = Female, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before -0.6153846 1.168258 41 -0.527 0.6012
##
## Sex_1 = Male, InterventionC = Testosterone:
## contrast estimate SE df t.ratio p.value
## After - Before -1.3636364 1.270030 41 -1.074 0.2892lsmip(Model4,InterventionC~Pre|Sex_1)
##Plots of Time
#Both conditions
p1 <- ggplot(T_DF.NA,aes(y=Test,x=Time,color=InterventionC)) + geom_point() + geom_smooth()
p1
#Testosterone Only -- seperated by Sex
p1.b <- ggplot(subset(T_DF,InterventionC=="Testosterone"),aes(y=Test,x=Time,color=Sex_1)) + geom_point() + geom_smooth()
p1.b## Warning: Removed 5 rows containing non-finite values (stat_smooth).## Warning: Removed 5 rows containing missing values (geom_point).
#Log Transformed
p2 <- ggplot(T_DF.NA,aes(y=Test.l,x=Time,color=InterventionC)) + geom_point() + geom_smooth()
p2
#Total Aggression
#All data
p3 <- ggplot(subset(T_DF.NA2,InterventionC=="Testosterone"),aes(y=Test.l,x=Total.Aggression)) + geom_point(size=2) + geom_smooth(method="lm",se=TRUE)
p3
#Seperated by ID
p3 <- ggplot(subset(T_DF.NA2,InterventionC=="Testosterone"),aes(y=Test.l,x=Total.Aggression,color=ID)) + geom_point(size=2) + geom_smooth(method="lm",se=FALSE)
p3
p3 <- ggplot(subset(T_DF.NA2,Sex_1=="Male" & InterventionC=="Testosterone"),aes(y=Test.l,x=Total.Aggression,color=ID)) + geom_point(size=2) + geom_smooth(method="lm",se=FALSE)
p3
#Seperate Plots for each individual
uni <- unique(T_DF.NA2$ID)
for (i in 1:length(uni)) {
p3 <- ggplot(subset(T_DF.NA2,InterventionC=="Testosterone" & ID==uni[i]),aes(y=Test.l,x=Total.Aggression,color=ID,linetype=Session)) + geom_point(size=2) + geom_smooth(method="lm",se=FALSE); print(p3)
}
