No, I am sorry to say, you are not doing them right.
What exactly are you trying to do?. I see a possible relationship to a Riemann sum there.
By the right endpoint method:
Now, use the identities you learned. I see you used one in your post(the sum of the squares). Then, take the limit as .
You're shooting for 21, not -1/2.
Try the same technique on this one, if that's what you need to do.