I) (50.4-44)/3.2=2; (37.6-44)/3.2)=-2. The difference between these is 4 standard deviations which means that by far the majority of the data lies between 37.6 and 50.4 (almost 100%). So the statement is false.
II) The data lies between 1 and 2 standard deviations below the mean (Z-scores between -2 and -1):
0.1587-0.0228=0.1359 or around 13.5%, so the statement is true.
III) This sounds about right: 37.6 is 2 SD's below the mean (2.28%) and 50.5 is 2 SD's above the mean (2.28%), so the relevant combined percentage is 4.56%, close to 5%.