Whereas the Wald confidence interval to estimate a population proportion uses the sample proportion in computing its standard error, hypothesis tests about a population proportion should use the hypothesized value, p0, in its standard error. As a result, using the Wald confidence interval to test hypotheses about a population proportion could lead to an increase in either a Type I or a Type II error.

The use of real contexts and history in the teaching of statistical principles holds much attraction in the classroom. This is especially so when the examples represent unusual situations but can lead to further investigation by the students. Previous instances of this have been utilized to great effect with sporting data and even courtroom decisions. This paper tackles a very practical problem in forensic science using simple linear regression, with some spectacular and extremely useful results.