Publication: Personnel Psychology (SUMMER 2010)
Article: Are highly structured job interviews resistant to demographic similarity effects?
Authors: J.M. McCarthy, D.H. Van Iddekinge, and M.A. Campion
Reviewed By: Benjamin Granger
Interviews are by far one of the most commonly used personnel selection tools and for good reason: They work (at least when they’re structured)!
One potential problem with interviews is that irrelevant personal characteristics of interviewees (i.e., gender, race) may affect interview ratings; interviewees who are similar (race, gender) to interviewers will receive higher ratings in an interview than those who are dissimilar to the interviewers. This can ultimately lead to illegal practices and failing to hire the best applicants. This potential problem is known as demographic similarity. The underlying reason this may occur is that people view others who are similar to themselves more favorably than those who are different (e.g., She is just like me so she must also be awesome!).
McCarthy, Van Iddekinge and Campion (2010) explored the existence of demographic similarity effects in highly structured job interviews, which are known to be more valid than unstructured interviews. McCarthy et al. found that the demographic similarity effect was completely non-existent in highly structured interviews. In fact, their results suggest that there is no such thing as a demographic similarity effect when highly structured interviews are used.
Impressively, McCarthy et al.’s study was conducted on a total of 207 interviewers and nearly 20,000 entry level employees applying for professional level positions with the U.S. government. Moreover, McCarthy et al.’s findings were consistent across three common types of structured interviews: (1) past-behavioral based interviews, (2) interviews focusing on a person’s past experiences and education, and (3) situation-based interview (e.g., In hypothetical situation “X,” what would you do?).
So not only are highly structured interviews legally defensible and able to effectively target high future performers and weed out poor performers (i.e., criterion validity), we now know that they are highly resistant to demographic similarity effects! So here are a few quick tips for ratcheting up the structure of a job interview: (1) keep the questions of all applicants consistent and avoid shooting from the hip, (2) base interview questions off of a formal job analysis, (3) include ratings that have anchors with behavioral examples, (4) have interviewers take thorough notes during an interview, (5) avoid discussing applicants between interviews, and (6) include a panel of interviewers/raters as opposed to a single interviewer. For more information on how to increase the structure of a job interview, see the comprehensive list presented in McCarthy et al.’s article.