Some scenarios adapted from Gravetter, F.J. & Wallnau, L.B. (1996). Statistics for the behavioral sciences (4th ed.). St. Paul, MN: West Publishing.
6. Some young entrepreneurs designed a computer program meant to increase students’ SAT score. The national average on the SAT is 1497 and the standard deviation is 110. To test their program, they give it to some students, then give those students the SAT. Below are their results.
subject SAT score (after using the
program) A 1590 B 1460 C 1610 D 1670 E 1500 F 1580 G 1620
Use the 0.01 level of significance (do a directional test). The obtained statistic in this case was 5.76. Questions 1. Would the critical statistic be larger, smaller, or the same if a larger sample was used? What did you base your answer on?
Scenario #6 A. A z test should be used to analyze the data because there is one group of
subjects, and both the population mean and the population standard ����������ȋρ�����ɐȌ����������Ǥ�
B. It was found that the computer program significantly increases SAT scores, z = +5.76, p < 0.01, one tailed.
a. This conclusion was based on the fact that z-obtained (5.76) was larger than z-critical (2.33), and a one-tailed test was performed with a 0.01 alpha level.
C.
