HP 39g Master Manual page 155

We are dealing with two independent samples
in this case and so we need to choose from
those tests which involve two samples.
Since we know the standard deviations of only
these samples, we must again use a Student-t
test. We then change to the NUM SETUP view
to import the summary statistics.
Since the two standard deviations for the
samples are similar, and the sample sizes are
the same, a more accurate estimate of the
population standard deviation will be obtained
if the data is pooled, so ensure that this option
is ed.
The NUM view shows the critical values. We
can see that the probability of obtaining a test
student-t value of 3.38 is 0.0015 and this is
well below the permitted test level of 1%.
The PLOT view also shows that the vertical
line representing the value of or is
− x ∆
x x
1 2
well into the region of rejection indicated by
the R .
From the evidence we should reject the null hypothesis and accept instead
that the average weight of the group receiving supplement A is significantly
higher than those receiving supplement C with only a 1% chance of being
incorrect.
155