qIf
sample sizes are small (< 30), we use t-distribution
for the test:
(remember: )
qNull
Hypothesis H0: the
two samples come from the same population
q
qWe can
then use t-distribution to determine if these two sample means are
significantly different.