limits_tau: Confidence limits of tau

In jadahlke/psychmeta: Psychometric Meta-Analysis Toolkit

Confidence limits of tau

Description

Note that this interval does not incorporate uncertainty in artifact estimates, so the interval will be somewhat conservative when applied to individual-correction or artifact-distribution meta-analyses.

Usage

limits_tau(
  var_es,
  var_pre,
  k,
  method = c("profile_var_es", "profile_Q", "normal_logQ"),
  conf_level = 0.95,
  var_unbiased = TRUE
)

Arguments

var_es	The observed variance of effect sizes.
var_pre	The predicted variance of effect sizes due to artifacts.
k	The number of studies in a meta-analysis.
method	Which method to use to estimate the limits. Options are profile_var_es for a profile-likelihood interval assuming \sigma^{2}_es ~ \chi^{2}(k-1), profile_Q for a profile-likelihood interval assuming Q ~ \chi^{2}(k-1, \lambda), \lambda = \sum_{i=1}{k}{w_i(\theta - \bar{\theta})^{2}}, and normal_logQ for a delta method assuming log(Q) follows a standard normal distribution.
conf_level	Confidence level.
var_unbiased	Are variances computed using the unbiased (TRUE) or maximum likelihood (FALSE) estimator?

Value

The confidence limits of tau

Examples

limits_tau(var_es = 0.008372902, var_pre = 0.004778935, k = 20)

jadahlke/psychmeta documentation built on Feb. 11, 2024, 9:15 p.m.