combine.est {survcomp} | R Documentation |
The function combines several estimators using meta-analytical formula to compute a meta-estimate.
combine.est(x, x.se, hetero = FALSE, na.rm = FALSE)
x |
vector of estimates |
x.se |
vector of standard errors of the corresponding estimates |
hetero |
TRUE is the heterogeneity should be taken into account (random effect model), FALSE otherwise (fixed effect model) |
na.rm |
TRUE if the missing values should be removed from the data, FALSE otherwise |
estimate |
meta-estimate |
se |
standard error of the meta-estimate |
Benjamin Haibe-Kains
Cochrane, W. G. (1954) "The combination of estimates from different experiments", Biometrics, 10, pages 101–129.
test.hetero.est
set.seed(12345) x1 <- rnorm(100, 50, 10) + rnorm(100, 0, 2) m1 <- mean(x1) se1 <- sqrt(var(x1)) x2 <- rnorm(100, 75, 15) + rnorm(100, 0, 5) m2 <- mean(x2) se2 <- sqrt(var(x2)) #fixed effect model combine.est(x=c(m1, m2), x.se=c(se1, se2), hetero=FALSE) #random effect model combine.est(x=c(m1, m2), x.se=c(se1, se2), hetero=TRUE)