A 7th grade science teacher would like to know if spending more time doing class experiments will help his students perform better on tests. In a day, he teaches 4 periods of science, so for his 1st and 3th period classes, he taught as he normally does with mostly lectures and occasional class experiments. For his 2nd and 4th period classes, he did a class experiment every day with minimal lectures. There were n = 25 different students in each of his classes. He then wanted to compare

the exam averages between his classes to see if student performance was significantly different. Which of statistical tests should he use?