In: Statistics and Probability
Mr. Jones wanted to check if there was an observable impact of a particular worksheet on his students’ test scores for Chapter 1. His idea was to test if there was a significant variation in the average test score of the class that used the worksheet in comparison to the class that did not. In the class before lunch, he used the worksheet but not the one after lunch. He has 23 students in each class. Their test scores are below:
Before Lunch |
After Lunch |
||
86 |
79 |
23 |
67 |
59 |
45 |
74 |
46 |
98 |
67 |
38 |
90 |
81 |
80 |
14 |
79 |
32 |
72 |
38 |
63 |
35 |
84 |
51 |
38 |
31 |
60 |
18 |
56 |
57 |
20 |
44 |
70 |
68 |
43 |
27 |
33 |
24 |
16 |
16 |
14 |
31 |
58 |
24 |
30 |
82 |
17 |
||
a)
Since
1) samples are paired and 2) sample size is not large
non- pamametric sign test is prefered.
b)
The hypotheses are given below.
(null hypothesis) H0: The median difference is zero
versus
(alternative hypothesis) H1: The median difference is not zero.
c) Counting the number of positives and negatives
The test statistic is the number of negative signs.(number of positive signs, whichever is smaller.)
d)
We have a critical value of 6 from table and p-value is 0.037
Since the P-value is less than 0.05(> 0.037). Therefore, we can reject the null hypothesis. There is sufficient evidence, at the 0.05 level there is a significant median difference.
e)
There is sufficient evidence of significant variation in the average test score of the class that used the worksheet in comparison to the class that did not.(Mr. Jones idea to test if there was significant variation was right )