Williams Coke Distributor labels its coke bottles 32 ounces. The Weights and Measures department randomly samples 50 bottles and finds a sample mean of 31.80 ounces with a standard deviation of .75. With a significance level of .01 test this claim of cheating by Weights and Measures.

I don't understand whether to use a t score or z score. how to get that score and then how the sign level ties in (to find the critical value??) and ultimately to reject or fail to reject. Any guidance would be much appreciated