gilberto-sassi / power

Power and sample size for Hypothesis testiing.

power

This function computes the power and sample size for basic testing hypothesis listed below.

One population and variable

• Z test for mean (bilateral and unilateral): pwr_z_test_1pop
• t-test for mean (bilateral and unilateral): pwr_t_test_1pop
• Chi-squared for variance (bilateral and unilateral): pwr_sigma_1pop
• Proportion for test: pwr_prop_1pop

Two populations and variables

• Z test for difference of means (bilateral and unilateral) with known variance: pwr_z_test_2pop
• F test to compare variance of two normal population: pwr_sigma_2pop
• t test for difference of means (bilateral and unilateral) with unknown variance and differences: pwr_t_test_2pop_hetero
• t test for difference of means (bilateral and unilateral) with unknown and equals: pwr_t_test_2pop_homo
• Paired t test: pwr_paired_t_test
• Test for proportions (bilateral and unilateral) in samples with more than 40 observations: pwr_prop_2pop

Checking association

• Chi-sqaured to check association between two qualitative variables: pwr_chisq_test_association (implementation of power of test and sample size)
• Test for Pearson's correlation using Fisher's Z tranformation: z_fisher_test (implementation of test) and pwr_z_fisher_test (implementation of power of test and sample size)

Anova

• Unbalanced ANOVA: pwr_anova_balanced (power of test and sample size)
• Balanced ANOVA: pwr_anova_unbalanced (only power of test)

References

• MONTGOMERY, Douglas C.; RUNGER, George C. Applied statistics and probability for engineers. John Wiley & Sons, 2010.