In: Statistics and Probability
A company applies an aptitude test to all those who apply for jobs as programmers. The results of these tests are analyzed according to their precedence (general advertising, ads in specialized magazines, employment agencies, personal recommendations or people who come spontaneously). With a significance level of 0.025, determine if there is a difference between the results of these tests according to their precedence. Your work should have detailed procedure, in this exercise you can use the tool offered by Excel for the required test, but you have to have the appropriate analysis.
General Pub | 36 | 47 | 38 | 51 | 62 | 78 | 60 | 47 | 49 | 53 | 26 | 38 | 61 | 39 | 43 |
Rev Esp. | 58 | 64 | 62 | 47 | 71 | 90 | 65 | 82 | 61 | 59 | |||||
Ag Emp. | 47 | 59 | 48 | 81 | 66 | 50 | 42 | 53 | |||||||
Rec Per. | 67 | 61 | 82 | 97 | 65 | 72 | 54 | 59 | 58 | ||||||
Spontaneous | 38 | 47 | 80 | 41 | 38 | 66 | 50 |
The hypotheses will be
H0: all the precedence is equal .i.e. there is no significant difference between the precedence
H1: At least two of the precedence is not equal i.e. there is a significant difference between the precedence
The appropriate analysis that can be used is ANOVA for a single factor. We can conduct this test using MS Excel.
Enter data → Data → Data Analysis → Anova: Single Factor
Provide the necessary details and make α=0.025 them the following appears.
Anova: Single Factor |
||||
SUMMARY |
||||
Groups |
Count |
Sum |
Average |
Variance |
General Pub. |
15 |
728 |
48.53333 |
169.6952 |
Rev Esp. |
10 |
659 |
65.9 |
152.9889 |
Ag Emp. |
8 |
446 |
55.75 |
159.9286 |
Rec Per. |
9 |
615 |
68.33333 |
186 |
Spontaneous |
7 |
360 |
51.42857 |
253.2857 |
ANOVA |
||||||
Source of Variation |
SS |
df |
MS |
F |
P-value |
F crit |
Between Groups |
3248.561 |
4 |
812.1401 |
4.53488 |
0.003736 |
3.093288 |
Within Groups |
7879.848 |
44 |
179.0874 |
|||
Total |
11128.41 |
48 |
The critical value of F for (4, 44) d.f and α=0.025 level of significance is 3.093288.
Since the calculated value is greater than critical value and the p-value is less than 0.025, we reject the null hypothesis and conclude that with 97.5% confidence that the precedence differ significantly. So thers is a difference between the results of these tests according to their precedence.
I request you to provide a feedback regarding the answer.