n.ttest: n.ttest computes sample size for paired and unpaired t-tests.

In samplesize: Sample Size Calculation for Various t-Tests and Wilcoxon-Test

Description

n.ttest computes sample size for paired and unpaired t-tests. Design may be balanced or unbalanced. Homogeneous and heterogeneous variances are allowed.

Usage

n.ttest(power = 0.8, alpha = 0.05, mean.diff = 0.8, sd1 = 0.83, sd2 = sd1,
        k = 1, design = "unpaired", fraction = "balanced", variance = "equal")

Arguments

power	Power (1 - Type-II-error)
alpha	Two-sided Type-I-error
mean.diff	Expected mean difference
sd1	Standard deviation in group 1
sd2	Standard deviation in group 2
k	Sample fraction k
design	Type of design. May be paired or unpaired
fraction	Type of fraction. May be balanced or unbalanced
variance	Type of variance. May be homo- or heterogeneous

Value

Total sample size	Sample size for both groups together
Sample size group 1	Sample size in group 1
Sample size group 2	Sample size in group 2

Author(s)

Ralph Scherer

References

Bock J., Bestimmung des Stichprobenumfangs fuer biologische Experimente und kontrollierte klinische Studien. Oldenbourg 1998

Examples

n.ttest(power = 0.8, alpha = 0.05, mean.diff = 0.80, sd1 = 0.83, k = 1,
design = "unpaired", fraction = "balanced", variance = "equal")

n.ttest(power = 0.8, alpha = 0.05, mean.diff = 0.80, sd1 = 0.83, sd2 =
2.65, k = 0.7, design = "unpaired", fraction = "unbalanced", variance =
"unequal")

Example output

$`Total sample size`
[1] 36

$`Sample size group 1`
[1] 18

$`Sample size group 2`
[1] 18

$`Total sample size`
[1] 153

$`Sample size group 1`
[1] 37

$`sample size group 2`
[1] 116

Warning message:
In n.ttest(power = 0.8, alpha = 0.05, mean.diff = 0.8, sd1 = 0.83,  :
  Arguments -fraction- and -k- are not used, when variances are unequal

samplesize documentation built on May 2, 2019, 5:56 a.m.