Yeah it's a good idea to do that in order to properly answer the question. Don't use it for variance though because that still has the squared values. Just an approximation. Try typing into cas 'normcdf(-1,1,0,1)'. It gives the result ~0.682689 (not 0.68 exactly) which obivously represents the probability within one standard deviation either side of the mean.
Hope that's all clear 
Thanks, yep, very clear- aaand I got another one

Mr Lim, a physics teacher, sets a particularly hard test for his students. I finds the average mark is 54 and the standard deviation is 8. He decides to award the top 10% an A, and fail the bottom 10%. Find, correct to the nearest whole number:
the lowest mark required to achieve an A.
So I just did the usual find the z score that correlates to the bottom 90%, then found the actual mark required using the rule, which was 64.25 to two decimal places. However this is saying that the lowest mark required to get an A is 64.25, which means if we have to round to the nearest whole number, would be round up or down? My problem is that if we round down (which is closer), we get a value of 64, but if a student gets a mark of 64, then technically they haven't scored in the top 10% and shouldn't get an A. So I rounded up instead to 65, but BOB says 64..
So the question is, what would you do in these sorts of situations- I kinda saw it like one of those "break-even" questions in Further, but would be good to confirm which answer is correct!