The head of the police force in the city of Cheadle suspects that increasing the pay of his officers might increase their efficiency. A police researcher is assigned to test whether there is a difference between the crime-solving rates of a group of detectives who have been randomly awarded pay raises and a control group of detectives who have not been awarded pay raises. In writing up his report, the researcher concludes as follows:

The results show that the observed significance level is 0.14, meaning that rejecting the null hypothesis would run a 14% risk of a Type I error. Although a 5% significance level is considered standard, in light of the potential benefits of salary increases for crime control rates, a higher 15% threshold is justified here, and the H0 may therefore be rejected.

a. What is the null hypothesis to which the researcher refers?

b. Explain why the researchers statistical reasoning is problematic.