PPpSceptical: Compute project power of the sceptical p-value

Description Usage Arguments Details Value Author(s) References See Also Examples

Description

The project power of the sceptical p-value is computed for a specified level of replication success, the relative variance, significance level and power for a standard significance test of the original study, and the alternative hypothesis.

Usage

1
2
3
4
5
6
7
8
PPpSceptical(
  level,
  c,
  alpha,
  power,
  alternative = c("one.sided", "two.sided", "greater", "less"),
  type = c("golden", "nominal", "liberal", "controlled")
)

Arguments

level

Numeric vector of levels of replication success.

c

Numeric vector of variance ratios of the original and replication effect estimates. This is usually the ratio of the sample size of the replication study to the sample size of the original study.

alpha

Significance level for a standard significance test in the original study.

power

Power to detect the assumed effect with a standard significance test in the original study.

alternative

Either "one.sided" (default), "two.sided", "greater", or "less". If "one.sided", the type-I error rate is computed based on a one-sided assessment of replication success in the direction of the original effect estimate. If "two.sided", the type-I error rate is computed based on a two-sided assessment of replication success regardless of the direction of the original and replication effect estimate. If "greater" or "less", the type-I error rate is computed based on a one-sided assessment of replication success in the pre-specified direction of the original and replication effect estimate.

type

Type of recalibration. Can be either "golden" (default), "nominal" (no recalibration), "liberal", or "controlled". "golden" 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.

Details

PPpSceptical is the vectorized version of .PPpSceptical_. Vectorize is used to vectorize the function.

Value

The project power.

Author(s)

Samuel Pawel, Leonhard Held

References

Held, L. (2020). The harmonic mean chi-squared test to substantiate scientific findings. Journal of the Royal Statistical Society: Series C (Applied Statistics), 69, 697-708. doi: 10.1111/rssc.12410

Held, L., Micheloud, C., Pawel, S. (2021). The assessment of replication success based on relative effect size. https://arxiv.org/abs/2009.07782

Maca, J., Gallo, P., Branson, M., and Maurer, W. (2002). Reconsidering some aspects of the two-trials paradigm. Journal of Biopharmaceutical Statistics, 12, 107-119. doi: 10.1081/bip-120006450

See Also

pSceptical, levelSceptical, T1EpSceptical

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
## compare project power for different levels of replication success
levels <- c("nominal" = levelSceptical(level = 0.025, type = "nominal"),
            "liberal" = levelSceptical(level = 0.025, type = "liberal"),
            "controlled" = levelSceptical(level = 0.025, type = "controlled"),
            "golden" = levelSceptical(level = 0.025, type = "golden"))
c <- seq(0.4, 5, by = 0.01)
alpha <- 0.025
power <- 0.9
pp <- sapply(X = levels, FUN = function(l) {
  PPpSceptical(level = l, c = c, alpha, power, alternative = "one.sided",
               type = "nominal")
})

## compute project power of 2 trials rule
za <- qnorm(p = 1 - alpha)
mu <- za + qnorm(p = power)
pp2TR <- power*pnorm(q = za, mean = sqrt(c)*mu, lower.tail = FALSE)

matplot(x = c, y = pp*100, type = "l", lty = 1, lwd = 2, las = 1, log = "x",
        xlab = bquote(italic(c)), ylab = "Project power (%)", xlim = c(0.4, 5),
        ylim = c(0, 100))
lines(x = c, y = pp2TR*100, col = length(levels) + 1, lwd = 2)
abline(v = 1, lty = 2)
abline(h = 90, lty = 2, col = "lightgrey")
legend("bottomright", legend = c(names(levels), "2TR"), lty = 1, lwd = 2, 
       col = seq(1, length(levels) + 1))

ReplicationSuccess documentation built on July 16, 2021, 9:08 a.m.