In: Statistics and Probability
A company was thinking about instituting a policy that allowed for no limits to the number of sick days that employees could take rather than limiting the number of paid sick days. Each column represents a different sample of employees during the trial period.
A) Explain why an alpha level of 0.05 is appropriate for the test (compared to, say, 0.01 or 0.001).
B) What are we doing by setting an alpha level? Explain it for the layperson.
C) What is the null hypothesis and alternate hypothesis? (Do not worry about formatting the statements - I'll get it from what you write.)
D) Conduct the t-test at α = 0.05, and report the results and conclusion. Provide in your statement the necessary statistics to support your answer.
Limits | No Limits |
35 | 23 |
21 | 12 |
26 | 6 |
24 | 15 |
17 | 18 |
23 | 5 |
37 | 21 |
22 | 18 |
16 | 34 |
38 | 10 |
23 | 23 |
41 | 14 |
27 | 19 |
24 | 23 |
32 | 8 |
A)
Level of significance alpha shows the probability of rejecting the true null hypothesis. That is we incorrectly conclude that number of leaves with limit and no limit are not same while actually they are same. Since this error is not very critical so we can assume alpha = 0.05.
B)
It is maximum probability that we allow to conclude that number of leaves with limit and no limit are not same while actually they are same.
C)
Following table shows the calculations:
Limits | No Limits | d=limits-no limits | (d-mean)^2 |
35 | 23 | 12 | 2.3409 |
21 | 12 | 9 | 2.1609 |
26 | 6 | 20 | 90.8209 |
24 | 15 | 9 | 2.1609 |
17 | 18 | -1 | 131.5609 |
23 | 5 | 18 | 56.7009 |
37 | 21 | 16 | 30.5809 |
22 | 18 | 4 | 41.8609 |
16 | 34 | -18 | 810.5409 |
38 | 10 | 28 | 307.3009 |
23 | 23 | 0 | 109.6209 |
41 | 14 | 27 | 273.2409 |
27 | 19 | 8 | 6.1009 |
24 | 23 | 1 | 89.6809 |
32 | 8 | 24 | 183.0609 |
Total | 157 | 2137.7335 |