hEstimates: Point estimates: For the heterogeneity parameter

In metagen: Inference in Meta Analysis and Meta Regression

Description

Returns a list of tau estimates based on different approximative methods. Different point estimates for the heterogeneity parameter are calculated: HD (Hedges), SL (DerSimonian-Laird), SJ (Sidik-Jonkman), MP (Mandel-Paule), ML (maximum likelihood), REML (restricted maximum-likelihood). Since any of these methods may fail to converge, there result may be 'NA' in this case.

Usage

  hEstimates(y, d, x)

Arguments

y	study responses
d	heteroscedasticity
x	design matrix

Value

A data frame containing point estimates. Variables are 'type' and 'h'.

Examples

bcg   <- bcgVaccineData()
bcg_y <- bcg$logrisk
bcg_d <- bcg$sdiv
bcg_x <- cbind(1,bcg$x)
hEstimates(y=bcg_y, d=bcg_d, x=bcg_x)

# The implementation can also handle the case in which
# a meta regression is planed with no intercept and only a
# single covariate (i.e. dim(x) = 1).  In this case,
# the design matrix can simply be provided by a vector.
# (This makes no sense in this example and shall only prove
# feasibility)
hEstimates(y=bcg_y, d=bcg_d, x=bcg$x)

# When performing a meta analysis, provide the function
# with a vector of 1s.
hEstimates(y=bcg_y, d=bcg_d, x=rep(1, length(bcg_y)))

metagen documentation built on May 2, 2019, 6:08 a.m.