Supplementary Exercise 6.85 of IPS7e
------------------------------------

Test scores for eight-grade students in Boston: the average score was
262 but not statistically significant at significance level 0.05 from
the country-level average (in large cities) of 287.

The settings of their test must have been:
  H0: mu=287
  Ha: mu<>287
  Xmean=262
  P>0.05

Some wordings to explain this:
- the sample average of 262 could have happened by chance alone if the
true population mean for Boston students was 287,

- the sample average to be 262 was not a particularly surprising event
if the true population mean for Boston students was 287,

- the sample average of 262 did not constitute sufficient evidence to
rule out that the true population mean for Boston students was 287,

- the probability for the Boston sample average to differ from the
true population mean by at least 25 points was more than 5%,
if the true population mean for Boston students was 287.

Most likely the last statement will be true also without assuming
the true population mean to be 287, but this will depend on the method
used to compute the test. For z-tests or t-tests the statement will
be true generally, because the only way in which the population mean
enters the calculations is through the difference between sample average
and population in the numerator of the z- and t-statistics.