surv.covariance.variance: Construct covariance-variance matrix for study-level model of
MLMA
Description
Constructs the "known" covariance-variance matrix for the transformed
estimates of a survival meta-analysis. Responses are a set of survival
proportions (points on KM plot) grouped in some way (study; treatment-within-study). The GLM framework has the link function g(x) = log(-log(x)).
Usage
surv.covariance.variance(surv,sigma2,group)
Arguments
surv
vector of survival estimates ordered by group
sigma2
vector of variances for surv estimates (on survival scale)
group
vector of grouping factor which should be treatment and study interaction