Question 1132052
Assuming normality, random sample and s being an unbiased estimator of sigma, all that
test statistic is a t 0.975, df=48
critical value |t|>2.01
calculation (x bar-mean)s/sqrt(48)
that is (-365)/351/sqrt(48)
which is -7.20 and strongly reject Ho that the two are equal with p<<0.0001 

The second is a two sample proportion
use z, critical value |z|>1.96
(p hat-p)/sqrt(p*(1-p)/n)), where n=(1/n1)+(1/n2)
the two proportions are 0.683 and 0.785
their difference is -0.102
divide by SE, using combined proportion of 280/380, or 0.737, so SE is sqrt (0.737*0.263)((1/180)+(1/200))
=0.045. This approach pools the two proportions, which is reasonable if the null hypothesis says they are equal, therefore part of the same population.
the z value is 2.27 so reject Ho and conclude the two proportions are different.