First and foremost, interviewers need to determine if a candidate will be a high-quality employee. A significant body of research suggests that the way an interview is constructed has a significant bearing on whether it can foretell later job performance. The following characteristics are the most important:
Stay Tuned to the Job
Research teaches that interviews and interview questions need to be job-related. This is the best way to ensure that applicants’ answers will help the interviewer make an informed decision. One of the most important functions of the interview is to help interviewers assess the degree to which a candidate has the necessary skills and competencies to be successful on the job, both behavioral (customer orientation or interpersonal skills) and technical. When questions are not job-relevant, and the interviewer begins to wander or ask inappropriate questions, the likelihood of coming away with the best information dramatically decreases.
Interviews built on job-related questions also are less likely to result in claims of unfair discrimination and, if they are questioned in court, are usually more defensible. Interviewers should be trained to not only ask job-related questions, but also avoid asking illegal questions. A recent DDI survey shows that the latter remains a serious problem. Only 68 percent of interviewers were aware that asking a candidate’s age was illegal, 60 percent knew that asking whether a candidate was married was illegal, and only 56 percent knew that it was illegal to ask if a candidate planned to have children.
Structure Your Questions
One way to keep focused on job requirements is to structure the interview. All candidates for the same position should be interviewed using the same set of rules, developed in advance, in order to gather the information needed to evaluate their potential for success. Researchers have demonstrated that an organization would need to conduct three to four unstructured interviews in order to get the same benefit or predictive power as they would from one structured interview (Schmidt & Zimmerman, 2004).
Research proves that outcomes or scores of structured interviews are indicative of how the candidate will actually perform on the job (Jelf, 1999; Moscoso, 2000). Additionally, structured interviews tend to result in very low levels of adverse impact; that is, they are less likely to select significantly smaller percentages of protected group candidates compared to the majority group (Hough, Oswald, & Ployhart, 2001). In particular, there are smaller race differences in interview results than in many other selection methods (Huffcutt & Roth, 1998; Williamson, Campion, Malos, Roehling, & Campion, 1997).
Ask When, Not If
Two types of highly structured interviews have become increasingly popular as ways of eliciting examples of behavior related to a job or job competency. The difference between the two is the types of questions asked:
Situational interview questions ask an applicant to describe what he or she would do in a hypothetical situation. For example, “Imagine you were in a situation where your team was under significant time pressure to complete a project. How would you handle that situation?”
Behavior-based interview questions ask applicants what they did in a specific situation. For example, “Think of a situation where you had to lead a team in a time crunch. Describe the situation, your actions, and the results of your actions.”
There has been much debate regarding the merits of these two approaches. Research results have been mixed, with both types shown to be related to later job performance. Key elements of this debate include the following:
Relevance to audience.
Some researchers indicate that situational interview questions are more appropriate for a broader audience because they do not require that individuals have extensive work experience (Pulakos & Schmitt, 1995). Other research, however, tested this assumption and found this not to be the case. Instead, these researchers argued that the nature of behavior-based questions allows for applicants to apply all types of life experiences to the questions (Gibb & Taylor, 2003). For example, the behavior-based question example that asked about leading a team could be answered by describing a work team, a sports team, or a school group. Research also suggests that individuals explaining past behavior are able to give considerably more information and detail, helping to provide the interviewer with additional insight.
Job complexity. Some researchers have argued that situational interviews are less able than behavior-based interviews to address complex and high-level jobs. Situational questions must accurately represent potential situations that job applicants might face. The higher one is in an organization, or the more complex the job, the more difficult it is to develop situational questions that fully encompass all parts of that job. This means that while the interviewer may be able to differentiate between those applicants who might be poor performers and those who would be good performers, it will be significantly more difficult for the interviewer to differentiate between those who will be good and those who will be great. Behavior-based interview questions, on the other hand, allow applicants to choose their own examples, elaborate on what might be very complex considerations and actions, and then explain in detail the different ways their actions lead to results (Huffcutt et al., 2004).
Typical or best performance. Behavior-based interview questions assess typical performance—what happens day to day—and show an applicant’s motivation to carry out the behaviors he or she thinks appropriate. Situational questions assess maximal performance, or what an applicant would do or consider the best option; they do not measure motivation to apply those skills or whether they are likely to actually perform the behaviors (Taylor & Small, 2002). A number of studies have compared the two types of questioning with later job performance. In each instance, the ability of behavior-based questions to predict performance on the job was significantly higher than for situational questions.
Figuring out the answer. Behavior-based question responses tend to be less related to cognitive ability of the applicant than situational questions. In other words, it is easier for people with high intellect to guess at and develop better answers for hypothetical situations, but they cannot change what has already happened. Because selection tools measuring cognitive ability tend to result in significant adverse impact, situational questions also tend to result in higher levels of adverse impact than do past behavior-based questions (Moscoso, 2000).
The preponderance of the evidence favors interview questions that ask about what interviewees actually did (“when” questions) and not what they would do (“if” questions). Although situational questions are not without merit, they might be more appropriately used as a test than in an interview.
At DDI, we have conducted a number of research studies demonstrating that our version of behavior-based interviewing, Targeted Selection® (TS®), produces a variety of positive organizational outcomes. For example, hiring with TS® has resulted in better hiring decisions, faster speed to productivity, and better job performance. Moreover, it has had a significant long-term impact, including greater customer satisfaction, reduced employee turnover, and enhanced organization performance. A number of these client results are detailed on our Web site, along with more information on our selection solutions.
About the Author
How Interviewers Identify The Best Candidates
