How to Conduct the Perfect Interview: Insights from the Science

Despite the fact that interviews are used in almost 100% of recruitment processes, few interviewers are aware of the empirical evidence behind interviews. HR practitioners and hiring managers are, therefore, making incredibly high-stakes decisions that impact the lives of real people via interviews, without even knowing if they work. Interviews are so ingrained into the traditions of employee selection that no one even questions whether or not to conduct them, or what the key variables are which influence the efficacy of interviews as assessment tools.

Fortunately for HR professionals and hiring managers, the research does suggest that, when interviews are done well, they are exceptionally powerful tools for employee selection. However, when done badly or designed inappropriately, their utility in selection plummets. In this article, I will outline what the empirical evidence says about interviews, with a particular focus on the main drivers of effectiveness when predicting real-world performance via interview.

Interview Structure

The main determiner of interview efficacy from the research is interview structure. Structured interviews significantly outperform unstructured interviews when it comes to predictive validity, making them far stronger predictors of performance. Structured interviews are rigorously designed, carefully scored, and employ a fixed set of questions which all candidates are asked. Unstructured interviews, also known as conversational interviews, do not follow a standard process and give entirely different questions to different candidates.

The reason why structured interviews outperform is that they are more reliable, statistically speaking. Giving everyone the same questions ensures that certain candidates aren’t given an unfairly hard time, and that others aren’t shoe-horned in with easy questions. Indeed, research shows that you need to conduct 3-4 unstructured interviews to reach the same level of validity as a structured interview.

Naturally, conversational interviews feel more natural, allowing you to put the candidate at ease and connect with the candidate on a deeper level. However, the goal of the interview is to identify high-potential candidates, not to make friends, and so interviewers are well advised to stick with structured interviews instead.

Interview Format

The next most important variable is the format of the interview, i.e., how the questions themselves are designed. From the research, we see three main kinds of interviews: 1) Personality interviews, designed to measure specific personality traits. 2) Past-oriented interviews, which ask the candidate questions based on their past experiences. 3) Situational interviews, which present hypothetical job-related situations to candidates.

From the research, personality interviews tend to be the weakest. Trying to infer personality traits, i.e., conscientiousness, industriousness, resilience, integrity, etc., from interviews is rarely effective, as interviews simply aren’t the correct way to measure these constructs. Inevitably, charismatic candidates just convince their interviewers that they have these characteristics, irrespective of their actual personality, weakening the interview.

Past-oriented interviews tend to outperform personality interviews but are still not the ideal way to interview. These interviews follow the typical “tell me about a time when…” interview format, seeing how well candidates have dealt with issues historically. The drawback of this approach is that candidates will simply make up historic examples of competence, and charismatic candidates once again can talk their way through.

Lastly, we have situational interviews, which are the most effective for predicting future performance. Situational interviews involve asking candidates hypothetical scenarios and seeing how well they would resolve those issues. The advantage of this approach is that charisma is less of an advantage; you genuinely need to know how to solve the problem presented. This increases the validity of the situational interview, and we strongly recommend this approach in recruitment.

Panel vs. Individual

The final consideration is whether to conduct one-to-one interviews or whether a panel of interviewers should be used. The theory behind panel interviews is that you have multiple raters, and thus that should increase the reliability of the assessment. In practice, however, research shows that panel interviews are less effective than individual interviews, which may come as a surprise.

The reason for this difference isn’t fully understood, but there are a few potential causes. Firstly, panel interviews are simply harder to structure than individual interviews, as different people are likely to interject and go off-script. Panel interviews are also more likely to make candidates nervous, hampering their performance. Lastly, interviewers are likely to use wash-up sessions after the interview to determine the candidate's performance, which introduces significant unreliability into the process.

Given that panel interviews are particularly expensive, time-consuming, and hard to arrange logistically, organizations are well advised to drop them as an assessment method.

Conclusion and Recommendations

Unbeknownst to the vast majority of hiring managers and HR professionals, who may argue often about their perceived ideal interview format, the research is actually quite clear on this topic. The ideal interview is a structured interview, relying on a fixed set of questions that everyone completes. It uses hypothetical situations as questions, evaluating what the candidate would do in this given situation. And lastly, it's a simple one-to-one interview, rather than a panel interview involving multiple interviewers. These interview configurations will maximize the quality of hire, helping organizations to find the best candidates from their applicant pools.