Assignment 5
Sample Size Determination for Paired and Two-Sample t-Tests
———————————————————-
Question 1: Aspirin study — paired t-test
H0: mu_B >= mu_A
Ha: mu_B < mu_A
Equivalently, defining D = A - B:
H0: mu_D <= 0 versus Ha: mu_D > 0
———————————————————-
aspirin <- power.t.test( delta = 1.5, # Expected mean paired
difference (%) sd = 3, # Standard deviation of paired differences
sig.level = 0.05, power = 0.75, type = “paired”, alternative =
“one.sided” )
aspirin
Round upward to obtain the required number of participants
aspirin_participants <- ceiling(aspirin$n)
aspirin_participants
Each participant provides two urine specimens: one for A and one for
B
aspirin_specimens <- 2 * aspirin_participants
aspirin_specimens
Conclusion: I will Enroll 23 participants and collect 46 urine
specimens.
Question 1 (Paired t-test, Aspirin):
n = 23 participants
specimens = 2 * 23 = 46 urine specimens
———————————————————-
Question 2: Infant walking study — two-sample t-test
H0: mu_active >= mu_control
Ha: mu_active < mu_control
———————————————————-
infants <- power.t.test( delta = 0.5, # Difference is half the
pooled standard deviation sd = 1, # Standardized scale sig.level = 0.10,
power = 0.85, type = “two.sample”, alternative = “one.sided” )
infants
Required number of infants in each group
infants_per_group <- ceiling(infants$n) infants_per_group
Total number of infants
total_infants <- 2 * infants_per_group total_infants
Conclusion: I will Enroll 44 infants per group, for 88 infants
total.
Question 2 (Two-sample t-test, Infant Walking):
n = 44 infants per group
total = 2 * 44 = 88 infants