Reeshad is the CEO of a small company. In an employee survey conducted last week, his employees exhibited relatively high levels of work stress. Reeshad therefore decided to stage an intervention with his employees. He forced his employees to come to the office on a Sunday morning, gave each of them a set of colored pencils and a coloring book that supposedly contained stress-relieving patterns, and required the employees to spend 4 hours coloring the patterns in the book without food or bathroom breaks. Reeshad then gave his employees a follow-up survey. Reeshad obviously hopes that his employees' stress levels will have improved (i.e., decreased) as a result of the intervention he staged.

Both before and after the intervention, stress was measured for the same set of 100 employees using the same set of questions and using a 1-4 scale, with higher numbers indicating more (i.e., higher) stress. In looking at the data, Reeshad sees the following:

Time 1 (i.e., pre-intervention survey) Mean Stress = 2.74 and SD = 1.44

Time 2 (i.e., post-intervention survey) Mean Stress = 3.66 and SD = 1.16

t (99) = -3.76, two-tailed p < .001

1. What is the null hypothesis in this case?
2. What is Reeshad's alternative hypothesis in this case?
3. Which type of t-test should be used to analyze these data? Why?
4. On the basis of the data, what should you conclude about whether stress levels improved (i.e., decreased) or worsened (i.e., increased) on the basis of the intervention? Was Reeshad's alternative hypothesis supported?