csdvsmdd.t {MDD}R Documentation

CSD vs. MDD for t-Test (Equal Variances)

Description

For a trial evaluated using a t-test with equal variances, calculates the probabilities of either having a clinically significant difference declared statistically insignificant, or of having a statistically significant difference declared clinically insignificant. Also ouputs the power of the test.

Usage

csdvsmdd.t(n1, n2, sigma, ate, csd, alpha = 0.05, alternative = "two.sided")

Arguments

n1 number of subjects in control group.
n2 number of subjects in treatment group.
sigma anticipated common standard deviation.
ate anticipated treatment effect.
csd clinically significant difference.
alpha significance level.
alternative indicates the alternative hypothesis and must be one of "two.sided", "greater", or "less". You can specify just the initial letter.

Details

The “clinically significant difference” is the smallest difference that will be meaningful either clinically or commercially. This is usually obtained from the clinical scientists and/or business partners you are working with.

The “anticipated treatment effect” is the quantity on which the study will be powered; it may, in general, be different from the clinically significant difference.

Value

A list with the following three components:

statsig.not.clinsig the probability that the trial will have a statistically significant but clinically insignificant result.
clinsig.not.statsig the probability that the trial will have a clinically significant but statistically insignificant result.
power the power of the test.

Author(s)

Don Barkauskas (barkda@wald.ucdavis.edu)

See Also

csdvsmdd.satter for unequal variances

gui.csdvsmdd for a GUI version

Examples

#Note that, unlike for csdvsmdd.chisq and csdvsmdd.fisher,
#at most one of the first two components can be zero.

csdvsmdd.t(20, 20, .35, .4, .3)
csdvsmdd.t(20, 20, .35, .4, .1)

[Package MDD version 0.5 Index]