The teacher from Problem 21 also tried a different approach to determining whether changing answers helps or hurts exam grades. In another class, students were told to review their final exams and change any answers they wanted to before turning them in. However, the students had to indicate both the original answer and the changed answer for each question. The teacher graded each exam twice, once using the set of original answers and once with the changes. In the class of n = 22 students, the exam scores improved by an average of MD5 2.5 points with the changed answers. The standard deviation for the difference scores was σ = 3.1. Are the data sufficient to conclude that rethinking and changing answers can significantly improve scores? Use a onetailed test at the .01 level of significance.