Statistics can be used to determine the affect on literary scores of the number of books read by a student.
The study would involve choosing a large number of students that makes up a meaningful sample size, it should be at least 30 or more. As the size of the sample increase it increases the level of confidence with which the results obtained can be extrapolated over the entire population.
Literary scores of students would then be required with the corresponding number of books read by each of them during the year. These can be considered as two variables and regression analysis used to determine if a change in the number of books read actually leads to a meaningful change in the test scores.
Based on the margin of error required and the confidence interval that is sought, a proper sample size would have to be determined.