sampleSizeReplicationSuccess: Computes the required relative sample size to achieve...

In ReplicationSuccess: Design and Analysis of Replication Studies

Description

The relative sample size to achieve replication success is computed based on the z-value of the original study, the replication success level, the type of recalibration and either the power or the minimum relative effect size. When the approach based on power is used, the design prior also has to be specified.

Usage

sampleSizeReplicationSuccess(zo, 
                             power, 
                             d,
                             level = 0.025,
                             alternative = "one.sided",
                             type = "golden", 
                             designPrior = "conditional")

Arguments

zo	A vector of z-values from original studies.
power	The power to achieve replication success.
d	The minimum relative effect size (ratio of the effect estimate from the replication study to the effect estimate from the original study) to achieve replication success.
level	Replication success level. Default is 0.025.
alternative	either "one.sided" or "two.sided". Defaults to "one.sided". Specifies if the replication success level is one-sided or two-sided. If the replication success level is one-sided, then sample size calculations are based on a one-sided assessment of replication success in the direction of the original effect estimate.
type	Type of recalibration. Can be either "nominal" (no recalibration), "liberal", "controlled" or "golden". The default is "golden" which ensures that for an original study just significant at the specified level, replication success is only possible if the replication effect estimate is at least as large as the original one. See ?levelSceptical for details about recalibration types.
designPrior	Is only taken into account when "power" is specified. Either "conditional" or "predictive". Defaults to "conditional".

Value

The relative sample size for replication success. If larger than 1000 Inf is returned.

Author(s)

Leonhard Held, Charlotte Micheloud

References

Held, L. (2020). A new standard for the analysis and design of replication studies (with discussion). Journal of the Royal Statistical Society: Series A (Statistics in Society). 183(2):431 - 448. https://doi.org/10.1111/rssa.12493

Held, L., Micheloud, C. & Pawel, S. (2020). The assessment of replication success based on relative effect size. https://arxiv.org/abs/2009.07782

See Also

pSceptical, powerReplicationSuccess, levelSceptical

Examples

## based on power
sampleSizeReplicationSuccess(zo = p2z(0.0025), power = 0.8, level = 0.025, 
                             type = "golden")
sampleSizeReplicationSuccess(zo = p2z(0.0025), power = 0.8, level = 0.025, 
                             type = "golden", designPrior = "predictive")
           
## based on minimum relative effect size                  
sampleSizeReplicationSuccess(zo = p2z(0.0025), d = 0.9, level = 0.025, 
                             type = "nominal")
                             
sampleSizeReplicationSuccess(zo = p2z(0.0025), d = 0.9, level = 0.025, 
                             type = "golden")

ReplicationSuccess documentation built on Dec. 2, 2020, 3 p.m.