tests/testthat/test_analysis_example_jackson2014.r

In metafor: Meta-Analysis Package for R

### library(metafor); library(testthat); Sys.setenv(NOT_CRAN="true")

context("Checking analysis example: jackson2014")

source("settings.r")

test_that("confint() gives correct results for example 1 in Jackson et al. (2014).", {

   skip_on_cran()

   ### example 1 ###

   yi <- c(0.0267, 0.8242, 0.3930, 2.4405, 2.1401, 1.2528, 2.4849, 0.3087,
           1.4246, 0.1823, 1.1378, 1.2321, 2.0695, 4.0237, 1.4383, 1.6021)
   vi <- c(0.1285, 0.0315, 0.0931, 2.0967, 1.0539, 0.1602, 1.0235, 0.0218,
           0.5277, 0.0556, 0.3304, 0.1721, 0.4901, 2.0200, 0.3399, 0.1830)
   xi <- c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1)

   ### random/mixed-effects meta-regression model (REML estimation by default)
   res <- rma(yi, vi, mods = ~ xi, digits=3)

   ### approximate 95% CI for tau^2 based on REML estimate and its SE
   ci <- exp(log(res$tau2) + c(-1.96,1.96)*(1/res$tau2 * res$se.tau2))
   expect_equivalent(ci[1], 0.0110, tolerance=.tol[["var"]])
   expect_equivalent(ci[2], 0.6330, tolerance=.tol[["var"]])

   ### generalised Cochran heterogeneity estimate and CI (inverse variance weights)
   res <- rma(yi, vi, mods = ~ xi, method="GENQ", weights=1/vi, digits=3)
   ci <- confint(res)
   expect_equivalent(ci$random[1,2], 0.0029, tolerance=.tol[["var"]])
   expect_equivalent(ci$random[1,3], 0.6907, tolerance=.tol[["var"]])

   ### generalised Cochran heterogeneity estimate and CI (inverse SE weights)
   res <- rma(yi, vi, mods = ~ xi, method="GENQ", weights=1/sqrt(vi), digits=3)
   ci <- confint(res)
   expect_equivalent(ci$random[1,2], 0.0000, tolerance=.tol[["var"]])
   expect_equivalent(ci$random[1,3], 1.1245, tolerance=.tol[["var"]])

   ### Paule-Mandel estimate and CI
   res <- rma(yi, vi, mods = ~ xi, method="PM", digits=3)
   ci <- confint(res)
   expect_equivalent(ci$random[1,2], 0.0023, tolerance=.tol[["var"]])
   expect_equivalent(ci$random[1,3], 1.4871, tolerance=.tol[["var"]])

})

test_that("confint() gives correct results for example 2 in Jackson et al. (2014).", {

   skip_on_cran()

   ### example 2 ###

   yi <- c(0.54, 0.4, 0.64, 0.365, 0.835, 0.02, 0.12, 0.085, 1.18, 0.08, 0.18,
           0.325, 0.06, 0.715, 0.065, 0.245, 0.24, 0.06, 0.19)
   vi <- c(0.0176, 0.019, 0.0906, 0.0861, 0.0063, 0.0126, 0.0126, 0.0041, 0.0759,
           0.0126, 0.0104, 0.0242, 0.0026, 0.2629, 0.0169, 0.0156, 0.0481, 0.0084,
           0.0044)
   xi <- c(1986, 1987, 1988, 1988, 1998, 1999, 2000, 2000, 2000, 2001, 2001, 2001,
           2002, 2002, 2002, 2002, 2003, 2003, 2003)

   ### random/mixed-effects meta-regression model (REML estimation by default)
   res <- rma(yi, vi, mods = ~ xi, digits=3)

   ### approximate 95% CI for tau^2 based on REML estimate and its SE
   ci <- exp(log(res$tau2) + c(-1.96,1.96)*(1/res$tau2 * res$se.tau2))
   expect_equivalent(ci[1], 0.0163, tolerance=.tol[["var"]])
   expect_equivalent(ci[2], 0.1108, tolerance=.tol[["var"]])

   ### generalised Cochran heterogeneity estimate and CI (inverse variance weights)
   res <- rma(yi, vi, mods = ~ xi, method="GENQ", weights=1/vi, digits=3)
   ci <- confint(res)
   expect_equivalent(ci$random[1,2], 0.0170, tolerance=.tol[["var"]])
   expect_equivalent(ci$random[1,3], 0.1393, tolerance=.tol[["var"]])

   ### generalised Cochran heterogeneity estimate and CI (inverse SE weights)
   res <- rma(yi, vi, mods = ~ xi, method="GENQ", weights=1/sqrt(vi), digits=3)
   ci <- confint(res)
   expect_equivalent(ci$random[1,2], 0.0180, tolerance=.tol[["var"]])
   expect_equivalent(ci$random[1,3], 0.1375, tolerance=.tol[["var"]])

   ### Paule-Mandel estimate and CI
   res <- rma(yi, vi, mods = ~ xi, method="PM", digits=3)
   ci <- confint(res)
   expect_equivalent(ci$random[1,2], 0.0178, tolerance=.tol[["var"]])
   expect_equivalent(ci$random[1,3], 0.1564, tolerance=.tol[["var"]])

})

rm(list=ls())