Assignment
The Chief Human Resource Officer at your organization has tasked the Talent Management (TM) team with determining the value of using various pre-hire assessments in the selection process-specifically, can using pre-hire assessments help in making better hiring decisions? If so, which assessments are useful? All come with associated monetary and time costs, and so it is necessary to understand predictive ability. The head of TM has asked the People Analytics team to help in evaluating the efficacy of various assessments by leading the validation effort.
Validity is the degree to which the pre-hire assessments accurately predict job performance. A validity study is the process of assessing whether the pre-hire assessments are predictive of job performance.
There are 3 types of assessments that are being considered: (i) cognitive ability; (ii) work samples; and (iii) situational judgement. Two hundred current employees participated in the validation study; each participant completed all 3 assessments.
You have HRIS data (Performance) and pre-hire assessment data, which have been used in three separate regression analyses. You have access to the output from these analyses below. What can you conclude about the pre-hire assessments and how useful they are in predicting performance?
Some things to note:
1) Employees that were part of the validation study completed all 3 assessments in February 2022
2) Performance reviews for employees are based on calendar year performance (Jan - Dec, 2022)
3) Higher scores on the Performance variable indicate better performance
4) Higher scores on Cognitive Ability Test variable indicate higher cognitive ability
5) Higher scores on the Situational Judgement Test variable indicate better situational judgment
6) Higher scores on the Work Sample variable indicate more highly rated work samples
Using cognitive ability test as a predictor:
b0=1.99
b1=0.18, S.E.=0.02, t=8.62, p<.001, std b1=0.52
R2
=0.27
F(1,198)=74.32, p<.001
Using situational judgement test as a predictor:
b0=1.65
b1=0.20, S.E.=0.02, t=9.09, p<.001, std b1=0.54
R2
=0.29
F(1,198)=82.56, p<.001
Using work samples as a predictor:
b0=1.72
b1=0.13, S.E.=0.04, t=4.54, p=.001, std b1=0.24
R2
=0.06
F(1,198)=11.56, p=.001
Task
1) What is/are the hypothesis/hypotheses you would test? (Hint: you want one focal hypothesis pertaining to each variable relationship you wish to test.)
2) Given the results you have, and your information regarding the context, what research design is most applicable to test the hypothesis(/hypotheses) given the data collected? Explain your answer.
3) Describe the analytic techniques that are being reported here, and why they help address your hypothesis (or hypotheses).
4) Using the results provided by your instructor, was/were your hypothesis/hypotheses supported? What numbers are you using and why?
5) What are the insights you can glean from the results?
6) What next steps would you recommend to leadership/stakeholders?