1. Reliability is about measuring a characteristic without error. Here are some possible questions to discuss: We know that various factors add to measurement error. When you go to take a test in class, what kinds of things affect how well you do on the test? Which ones tend to harm (lower) your score vs. which ones help (increase) your score? Longer tests or assessments, up to a point, lead to better reliability. Why? (As you think about this issue, consider a “dating” analogy to selection—would you want to marry someone after just one date? So why would we hire someone after just one interview?)
2. Answer 2 of the questions below and explain:
Should employers use social networking websites or “googling” to screen out applicants? Why or why not?
Should we even bother asking for references or reference letters if we are going to get little useful information from them?
Why do some interviewers like stress interviews, or like to ask off-the-wall questions? Are they getting any job-relevant information from these?
Should we use cognitive ability tests, even though they tend to have adverse impact?
An applicant asks for an ADA accommodation of a reader on a cognitive ability test. Will this change what we are measuring?