The Office of Personnel Management has been asked to evaluate the clerical test
ID: 2923862 • Letter: T
Question
The Office of Personnel Management has been asked to evaluate the clerical test that is used for the State of Washington. All applicants are given the test, but the test is not used for hiring or placement. After a six month probationary period, applicants that are hired can be made permanent or terminated. One hundred persons are hired for clerical positions. This dependent variable is coded 1 if the person is hired and 0 if the person is terminated. The independent variable is the person’s score on the clerical test. Scores can range from 0 to 100. A regression was run and yielded the following results:
Y = .23 + .0064X
R2 = .55 Sy|x = .09 sb = .0012 mean X = 75
Interpret the slope, intercept, R2. Is the slope statistically significant? What can you say about the relationship between the score and employment?
Explanation / Answer
Here slope value is 0.0064 that means if there is increase in score in 10 marks then there we will be more chance that the person is hired and it increase it by 0.06%.and if any person has more than 40 marks, it has Y score above 40 and that produce high chances of being selected..
If any person who is scoring 0 marks in the test. her will get a dependent variable score 0.23 that means it has high chances of being terminated.
Here R2 = 0.55 that shows that there 55% of the variability or changes in hiring or terminating for clerical positions are determined by marks scored in clerical test 6 month back. Rest 45% is deduced by enviromental factors.
Is slope significant?
the 95% confidence interval = b0 +- Z95% sb1 = 0.0064 +- 1.96 * 0.0012 = (0.004, 0.0088)
so these interval doesn't involve value of 0 so yes slope is statistically significant.
There is strong linear relationship between being hired and high marks on the clerical test.
Related Questions
drjack9650@gmail.com
Navigate
Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.