Description Usage Arguments Value Examples
View source: R/eefAnalyticPerm_modified_05_02_2016.R
crtFREQ
is a frequentist method that can be used to calculate effect size from cluster randomised trials based on residual variance or total variance. .
1 |
formula |
specifies the model to be analysed. It is of the form form y ~ x1+x2 +..., where y is the outcome variable and X's are the predictors. |
random |
a string variable specifying the "clustering" variable as contained in the data. This must be put between quotes. For example, "school". |
intervention |
specifies the name of the intervention variable as appeared in formula. This must be put between quotes. For example "intervention" or "treatment" or "group".. |
nPerm |
number of permutations required to generate permutation p-value. Default is NULL. |
nBoot |
number of bootstrap required to generate bootstrap confidence interval. Default is NULL. |
data |
the data frame to be analysed. |
S3 object; a list consisting of
Beta
. Estimates and confidence intervals for the predictors specified in the model.
It will be the slope for a continuous predictor and the mean difference for a dummy variable or a categorical predictor.
ES
. Hedges' effect size for the intervention effect. If nBoot is not specified,
the confidence intervals are 95
they are non-parametric bootstrapped confidence intervals.
covParm
. A vector of variance decomposition into between-variance (Schools), within-variance (Pupils) and total variance.
It also contains the intra-cluster correlation (ICC).
SchEffects
. Individual school effects at baseline.
Perm
. A matrix of the distribution of ES under the null hypothesis.
The two columns in the matrix represents ES based on within variance and total variance.
Produced only if nPerm is specified.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 | data(iwq)
########################################################
## MLM analysis of cluster randomised trials + 1.96SE ##
########################################################
output1 <- crtFREQ(Posttest~ Intervention+Prettest,random="School",
intervention="Intervention",data=iwq)
### Fixed effects
beta <- output1$Beta
beta
### Effect size
ES1 <- output1$ES$Intervention1
ES1
## Covariance matrix
covParm <- output1$covParm
covParm
### random effects for schools
randOut <- output1$"SchEffects"
randOut <- randOut[order(randOut$Estimate),]
barplot(randOut$Estimate,ylab="Deviations from Overall Average",
names.arg=randOut$Schools,las=2)
###############################################
## MLM analysis of cluster randomised trials ##
## with bootstrap confidence intervals ##
###############################################
output2 <- crtFREQ(Posttest~ Intervention+Prettest,random="School",
intervention="Intervention",nBoot=1000,data=iwq)
### Effect size
ES2 <- output2$ES
ES2
#######################################################################
## MLM analysis of cluster randomised trials with permutation p-value##
#######################################################################
output3 <- crtFREQ(Posttest~ Intervention+Prettest,random="School",
intervention="Intervention",nPerm=1000,data=iwq)
#### Distribution under the null
perm <- output3$Perm
#### Permutation P-value using total variance
obsg <- output3$ES$Intervention1[2,1]
p_value <- ifelse(mean(perm$"InterventionTotal" > obsg)==0,"<0.001",
mean(perm$"InterventionTotal" > obsg) )
p_value
hist(perm$"InterventionTotal", breaks=40, col="white", border="blueviolet",
xlab="Distribution Under Null Hypothesis",
main=paste("P(X|NULL)= ",p_value,sep=""),
xlim=range(c(perm$"InterventionTotal",obsg), na.rm=TRUE))
abline(v=obsg,lwd=2,col=4)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.