MIE 437 HR Analytics – Project #2 Instructions
The course projects will require you to conduct analyses, answer questions, create tables/figures, and provide a written report. You will use the real-world dataset to complete each project.
Project #2: Reliability, Validity, and t-tests (due Thursday [3/31] through Moodle – no exceptions!)
Background: Your HR manager at Wolfpack Widgets Inc. has asked you to conduct a deeper examination of the results of the most recent engagement survey. You will need to conduct new analyses to help your manager give a report to the chief executives about emerging issues in the workforce. Your project must include an executive summary (1-page single-spaced maximum) AND an appendix (after the summary) with the required tables and figures of the results. You must carefully review each of the three steps that is outlined below, and ensure that your Executive Summary and Appendix include every requested component.
Step 1: Evaluate the quality of job satisfaction and affective commitment scores (reliability)
1. Calculate coefficient alpha for affective commitment, engagement, social integration, and burnout (at t1). In your Executive Summary: (a) report these values, (b) describe what type of reliability is estimated by coefficient alpha, (c) describe what these values tell us about reliability, and (d) evaluate each coefficient in terms of whether it provides strong or weak reliability evidence.
2. In your executive summary, answer the following questions: (a) What type of reliability estimate do you believe is missing? That is, although you know the estimate of coefficient alpha, what else do you believe is important to understand reliability and why? (b) How could you address this in next year’s engagement survey? State this as a recommendation to your HR manager in your executive summary.
Step 2: Report on any significant differences between employees in their affective commitment and engagement scores (t-test)
3. Conduct the appropriate t-tests (two-tailed, α = .05) to investigate whether scores significantly differ on the basis of (a) sex, and on the basis of (b) newcomers v tenured employees. Note: you will need to conduct and report the results of four t-tests!
4. In your executive summary, summarize the results and conclusions of all t-tests. In your Appendix, include tables for your t-test statistical results, as well as figures with bar charts that show each group’s mean on the variable analyzed in the particular t-test.
Step 3: Evaluate burnout, an important expected outcome of affective commitment and engagement (Use burnout.t1).
5. Calculate the correlation between affective commitment (affcom.t1) and burnout (=correl(n1:n131,q2:q131). Is this strong or weak evidence of criterion-related validity? (Report and interpret in Executive Summary)
6. Calculate the correlation between engagement (engage.t1) and burnout (=correl(o1:o131,q2:q131). Is this strong or weak evidence of criterion-related validity? (Report and interpret in Executive Summary)
7. Conduct the appropriate t-tests (two-tailed, α = .05) to investigate whether the burnout scores significantly differ on the basis of (a) sex, and (b) newcomers v tenured employees.
a. Summarize your results and conclusions in the executive summary.
b. In your Appendix, include tables for your t-test statistical results, as well as figures with bar charts that show each group’s mean on the variable analyzed in the particular t-test.