R/statpsych1.R
sim.ci.median.ps.Rd
Performs a computer simulation of confidence interval performance for a median difference in a paired-samples design. Sample data within each level of the within-subjects factor can be generated from bivariate population distributions with five different marginal distributions. All distributions are scaled to have standard deviations of 1.0 at level 1. Bivariate random data with specified marginal skewness and kurtosis are generated using the unonr function in the mnonr package.
sim.ci.median.ps(alpha, n, sd.ratio, cor, dist1, dist2, rep)
alpha level for 1-alpha confidence
sample size
ratio of population standard deviations
population correlation of paired observations
type of distribution at level 1 (1, 2, 3, 4, or 5)
type of distribution at level 2 (1, 2, 3, 4, or 5)
1 = Gaussian (skewness = 0 and excess kurtosis = 0)
2 = platykurtic (skewness = 0 and excess kurtosis = -1.2)
3 = leptokurtic (skewness = 0 and excess kurtsois = 6)
4 = moderate skew (skewness = 1 and excess kurtosis = 1.5)
5 = large skew (skewness = 2 and excess kurtosis = 6)
number of Monte Carlo samples
Returns a 1-row matrix. The columns are:
Coverage - probability of confidence interval including population median difference
Lower Error - probability of lower limit greater than population median difference
Upper Error - probability of upper limit less than population median difference
Ave CI Width - average confidence interval width
# Only 1000 reps because CRAN runs all examples and requires short execution time.
sim.ci.median.ps(.05, 30, 1.5, .7, 4, 3, 1000)
#> Coverage Lower Error Upper Error Ave CI Width
#> [1,] 0.96 0.032 0.008 0.9459518
# Should return (within sampling error):
# Coverage Lower Error Upper Error Ave CI Width
# [1,] 0.961 0.026 0.013 0.9435462