# Sample size calculation for t-test

I need help calculateing the sample size for a study. I am using a group of students in a particular gradelevel. My study will be pretest-posttest type. I only have about 45 students in that grade, but my advisor said I have to perfrom a sample size. I will like to know the effects of Polya's four-step problem solving on the mathematical understanding of word problems. Can someone please help me. Thank you in advance.

-

Since your study will be pre/post test you will have two sets of measurements that are dependent since they are repeat measurements on the same people. If your scores are normally distributed then you can use a t-test for dependent samples. There is much material on the web about sample size calculations for t-tests. Just make sure your study set up satisifies the assumptions of the t-test before spending time trying to understand how to calculate the sample size (I do not want you to waste your time if the t-test is not appropriate).

-
+1. @Fran, what you are doing is called a power analysis. You will need to specify the effect size you are anticipating as well as the alpha level (usually 0.05) as well as the power you want (i.e., the probability of rejecting the null hypothesis if it is in fact false) - a normal setting would be 0.80. Then you can feed these data into a sample size calculator (make sure to specify that your t-test will be paired) and get the minimum sample size required. – Stephan Kolassa Jan 17 '13 at 19:20
Thank you very much for your help. – Fran Jan 22 '13 at 14:30