findn (version 0.1.0)

print.findn: Printing a findn Object

Description

Displays details about a sample size estimation from a findn object.

Usage

# S3 method for findn
print(
  x,
  details = c("low", "high"),
  max_n = NULL,
  digits = 3,
  invisible = FALSE,
  ...
)

Value

findn returns an object of class findn which contains the following elements:

sample_size

the sample size estimate

fit

the model coefficients and covariance matrix from the last Bayesian probit regression model

all_evals

all evaluated sample sizes

targ

the target power

level

the significance level for the confidence intervals used for the stopping criteria

exit.mes

a message about wheter the stopping criterion was reached with the number of simulations given by max_evals

By default, a list containing the point estimate for the sample size, the minimum sufficient sample size (i.e. the smallest sample size for which the lower limit of the confidence interval for the estimated power is larger than the target power) and a message whether the stopping criterion was reached is printed.

Arguments

x

Object of class findn.

details

Either "low" (default) or "high". See also 'Details'.

max_n

If details = "high" the predicted power values and confidence intervals are shown for all sample sizes from 1 to max_n if max_n is non-NULL. See also 'Details'.

digits

Number of decimal places to be shown.

invisible

Whether the results should be printed or only assigned.

...

Further arguments.

Details

When details = "low", only the point estimate (i.e., the smallest sample for which the predicted power exceeds the target power), the "minimum sufficient sample size" (i.e., the smallest sample size for which the lower limit of the level interval for the predicted power exceeds the target power) and an exit message. The exit message shows whether the chosen stopping rule was satisfied. If details = "high" then the default behaviour (i.e. when max_n = NULL) is to display all sample sizes, their predicted power values and the alpha whether their power exceeds the target power, and the three largest sample sizes that are smaller than the smallest sample size that is rated uncertain and the three smallest sample sizes which are greater than the smallest sample size that is rated uncertain. If details = "high" and max_n is non-NULL, then the sample sizes, their predicted power values and the confidence intervals for the predicted power values from 1 to max_n are displayed.

Examples

# Function that simulates the outcomes of a two-sample t-test
ttest <- function(mu1 = 0, mu2 = 1, sd, n, k) {
  sample1 <- matrix(rnorm(n = ceiling(n) * k, mean = mu1, sd = sd),
    ncol = k)
  mean1 <- apply(sample1, 2, mean)
  sd1_hat <- apply(sample1, 2, sd)
  sample2 <- matrix(rnorm(n = ceiling(n) * k, mean = mu2, sd = sd),
    ncol = k)
  mean2 <- apply(sample2, 2, mean)
  sd2_hat <- apply(sample2, 2, sd)
  sd_hat <- sqrt((sd1_hat^2 + sd2_hat^2) / 2)
  teststatistic <- (mean1 - mean2) / (sd_hat * sqrt(2 / n))
  crit <- qt(1 - 0.025, 2*n - 2)
  return(mean(teststatistic < -crit))
}

# Create a findn object
res_ttest <- findn(fun = ttest, targ = 0.8, k = 25, start = 100, 
  init_evals = 100, r = 4, stop = "evals", max_evals = 2000, 
  level = 0.05, var_a = 0.05, var_b = 1, alpha = 0.025, 
  alternative = "one.sided", sd = 2, verbose = FALSE)

# print with default settings
print(res_ttest, details = "low", digits = 3)