Two Sample Tests (Matched or Independent)

This is my course notes for Mathematical Biostatistics Boot Camp 2, course materials are free on github.

1. Matched data

Example

R Code

diff <- test2 - test1
n <- sum(!is.na(diff)) #49
mean(diff) #2.88
sd(diff) #7.61
testStat <- sqrt(n) * mean(diff) / sd(diff) #2.65
# below works out to be 0.01
2 * pt(abs(testStat), n -1, lower.tail = FALSE)
##uses the R function
t.test(diff)

Discussion of matched data

Regression to mediocrity

Comments

2. Two independent groups

Example

• 1 note this is not obtained by averaging the two standard deviations, it’s obtained by averaging the variances then square rooting

Comments