Confidence intervals for Rosenthal's fail-safe number assuming a half normal distribution with a fixed number of studies {fsn}

Confidence intervals for Rosenthal's fail-safe number assuming a half normal distribution with a fixed number of studies

Description

Confidence intervals for Rosenthal's fail-safe number assuming a half normal distribution with a fixed number of studies.

Usage

halfnorm.fixednr.ci(stat, se, alpha = 0.05, type = "dist", B = 1000)

Arguments

stat

A vector with the statistics.

se

A vector with the standard errors of the stat.

alpha

The significance level, set to 0.05 by default.

type

The type of confidence intervals to construct. Based on distributional assumptions ("dist"), based on the method of moments ("mom"), using non-parametric bootstrap ("boot") or all of these three ("all").

B

Number of bootstrap samples to generate.

Details

The function computes confidence intervals assuming a half normal distribution assuming that the number of studies is fixed and estimating the variance either via MLE, moments or bootstrap as described in Fragkos, Tsagris & Frangos (2014).

Value

A list including:

Nr

Rosenthal's fail safe number.

variance

The variance of Rosenthal's fail safe number.

ci

The (1-alpha)% confidence interval for the true Rosenthal's fail safe number.

Author(s)

Michail Tsagris and Constantinos Frangos

R implementation and documentation: Michail Tsagris [e-mail] and Constantinos Frangos [e-mail].

References

Konstantinos C. Fragkos, Michail Tsagris and Christos C. Frangos (2014). Publication Bias in Meta-Analysis: Confidence Intervals for Rosenthal's Fail-Safe Number. International Scholarly Research Notices, Volume 2014.

See Also

halfnorm.randomnr.ci, den.plot, rosenthal, convergence.rate

Examples

stat <- rnorm(30, 3, 0.2)
se <- rchisq(30, 1)
halfnorm.fixednr.ci(stat, se)