If a population is approximately "normal" (in the statistical sense) then you can convert the raw data of a sample into "z-scores" -- their analogous value on the standard normal scale where the mean=median=mode=0 and the standard deviation is 1.
The z-score helps to determine the percentage of the population that has a given characteristic (the z-score) or lower. (or higher depending on the application.)
If you have the scores on a test, and we assume that the scores will be approximately normally distributed, then you can convert each raw score (the actual test score) to its corresponding z-score. Then you can determine what percentage of the population will have a test score above or below the given test score from the value of the z-score in the standard normal table.