Question 1136460
There needs to be clarification of
1. What is the sd of the 90 hires?
2. This should be a 2-sample t-test given that s is being used to estimate sigma.  Where does the "population standard deviation" come from?  Is s of the prior year 17.2?

half-interval is t*(sp*sqrt((1/n1)+(1/n2)))

If I can assume the sd of the prior year for 17.2, and that was a sample, 
then sp^2=(n1-1)s1^2+(n2-2)s2^2/(n1+n2-2)
sp^2=64*15.6^2+89*17.2^2/(153)
sp^2=273.888; sp=16.550
t df=153,0.975=1.976
difference in means is -4.40
denominator of (sqrt ((1/n1)+(1/n2)))=0.02650; sp*sqrt((1/n1)+(1/n2))=2.694.
Multiply that by t of 1.976 to get 5.32
interval is (-9.72, 0.82)

(0 is in the interval so it would not be significant)