Independent Samples t-test
Validation
0.1.0
Here we compare the results of PAMLj with other software that performs power analysis. In particular, we will compare our results with R pwr package and G*Power.
Example 1 : Power
Setup
- Aim = Power
- Expected \(\delta\) = .5
- N per group = 30
- Alpha = .05
- Tails: Two
PAMLj
The expected power is .4779.
R
##
## Two-sample t test power calculation
##
## n = 30
## d = 0.5
## sig.level = 0.05
## power = 0.4778965
## alternative = two.sided
##
## NOTE: n is number in *each* group
G*Power
Results are the same
Example 2: Sample size
Setup
- Aim = N
- Expected \(\delta\) = .4
- Power = .90
- Alpha = .05
- Tails = Two
PAMLj
The expected N is 266, 113 per group. ## R
##
## Two-sample t test power calculation
##
## n = 132.3105
## d = 0.4
## sig.level = 0.05
## power = 0.9
## alternative = two.sided
##
## NOTE: n is number in *each* group
G*Power
Results are the same.
Example 3: Effect size
Setup
- Aim = minimal effect size
- power = .95
- N per group = 52
- Alpha = .01
PAMLj
R
##
## Two-sample t test power calculation
##
## n = 52
## d = 0.8415181
## sig.level = 0.01
## power = 0.95
## alternative = two.sided
##
## NOTE: n is number in *each* group
G*Power
Again, rounding a the third decimal place, results are the same.
Example 4: One-tail
Setup
- Aim = N
- Expected \(\delta\) = .4
- Power = .90
- Alpha = .05
- Tails = One
PAMLj
R
##
## Two-sample t test power calculation
##
## n = 107.7313
## d = 0.4
## sig.level = 0.05
## power = 0.9
## alternative = greater
##
## NOTE: n is number in *each* group
Comments?
Got comments, issues or spotted a bug? Please open an issue on PAMLj at github or send me an email