First I enter the data for the 10 people.
A <- c(15, 26, 13, 28, 17, 20, 7, 36, 12, 18)
B <- c(13, 20, 10, 21, 17, 22, 5, 30, 7, 11)
The same person took both types of aspirin, so the data is paired.
H0: the mean difference between A and B is 0 (mu_d = 0)
H1: the mean difference between A and B is not 0 (mu_d != 0)
t.test(A, B, paired = TRUE)
##
## Paired t-test
##
## data: A and B
## t = 3.6742, df = 9, p-value = 0.005121
## alternative hypothesis: true mean difference is not equal to 0
## 95 percent confidence interval:
## 1.383548 5.816452
## sample estimates:
## mean difference
## 3.6
The p-value is 0.0051. This is less than 0.05, so we reject H0. There is a difference in the mean concentration of the two aspirins. Aspirin A has a higher concentration, about 3.6 mg% more on average.
If we use a two-sample t-test instead:
t.test(A, B)
##
## Welch Two Sample t-test
##
## data: A and B
## t = 0.9802, df = 17.811, p-value = 0.3401
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -4.12199 11.32199
## sample estimates:
## mean of x mean of y
## 19.2 15.6
The p-value would be 0.34. With this test we would not reject H0. This happens because the two-sample test does not use the fact that the measurements come from the same person, so the differences between people make the variation bigger.
active <- c(9.50, 10.00, 9.75, 9.75, 9.00, 13.00)
no_exercise <- c(11.50, 12.00, 13.25, 11.50, 13.00, 9.00)
boxplot(active, no_exercise, names = c("Active", "No Exercise"),
ylab = "Time (months)")
H0: the mean time to walk is the same for both groups (mu_active = mu_no)
H1: the mean time to walk is less for the active exercise group (mu_active < mu_no)
We might want to use a non-parametric method because the samples are very small (only 6 babies in each group), so it is hard to know if the data is normal. Also, the active group has a value (13.0) that looks like an outlier compared to the others. I checked normality:
shapiro.test(active)
##
## Shapiro-Wilk normality test
##
## data: active
## W = 0.72061, p-value = 0.01012
The p-value is 0.01, so the active group does not look normal. The Mann-Whitney test does not need the data to be normal.
wilcox.test(active, no_exercise, alternative = "less", exact = FALSE)
##
## Wilcoxon rank sum test with continuity correction
##
## data: active and no_exercise
## W = 9, p-value = 0.08523
## alternative hypothesis: true location shift is less than 0
The p-value is 0.085. This is greater than 0.05, so we fail to reject H0. There is not enough evidence to say that active exercise makes babies learn to walk faster.