Paired T-Test

I am trying to see if there is a strong correlation between two datasets, but I am not confident that the paired t-test could be used. The first dataset is a score from 1 to 6 and the second set is percent grade received from a score. I am trying to determine if those with a higher score tend to have higher grades.
Do the datasets have to be of the same scale? (Such as percent grade vs. percent grade)