caceCRTBoot: CACE Analysis of Cluster Randomised Trials using MLM.

In eefMLM: Analytical Package for Analysing Education Trials

Description

caceCRTBoot performs CACE analysis of cluster randomised trials.

Usage

caceCRTBoot(formula, random, intervention, compliance, nBoot, data)

Arguments

formula	model specification of the form posttest ~ pretests+Intervention+....the model to be analysed. It is of the 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	the name of the intervention variable as appeared in formula. This must be put in quotes. For example "intervention" or "treatment" or "group".
compliance	percentages of sessions attended by pupils.
nBoot	number of bootstrap required to generate bootstrap confidence interval. This must be specified.
data	data frame containing the data to be analysed.

Value

S3 object; a list consisting of

• CACE. Estimated CACE effect size based on percentages of sessions attended by pupils. The percentage data is converted into the following grids (0, 10, 20, 30, 40, 50, 60, 70, 80, 90) and CACE effect size is calculated for each grid.

• Compliers. A summary table of the percentage of pupils in the intervention and control groups that attended more than a pre-specified percentage of sessions. The values for the control group should be zeros if there is no dilution in which a pupil or school in the control group receives intervention.

Examples

data(iwq)
############# CRT
caceOutput<- caceCRTBoot(Posttest~ Prettest+ Intervention,
			random="School",intervention="Intervention",
			compliance = "Percentage_Attendance",
			nBoot=1000,data=iwq)

### visualising CACE ES

require(metafor)
forest(x=caceOutput$CACE$ES, ci.lb=caceOutput$CACE$LB, 
		ci.ub=caceOutput$CACE$UB, xlab="CACE ES",
		slab=as.character(caceOutput$CACE$Compliance), 
		xlim=c(-1,2), alim=c(-1,3.75), cex=1,lwd=1.5)

eefMLM documentation built on May 2, 2019, 5:46 p.m.