There is a lot of controversy over the effectiveness of interviews as a final selection technique, and a lot of evidence to suggest that their effectiveness is poor, which is worrying and perhaps surprising, given how widely they are used and the extent to which companies rely on them.
A great deal depends on the structure of the interview. In particular, the research suggests that it is only structured interviews that are effective as predictors of future job performance. Indeed, they rank amongst the most effective selection techniques with a validity of 0.62. Unstructured interviews, by contrast, have a much lower validity rating of 0.31 (Anderson and Shackle ton, 1993).
Why are structured interviews so much more effective than unstructured interviews? A great deal comes down to the fact that it is easier to objectively compare the responses from a range of candidates when interviews are structured – they are all asked to respond to the same set of questions and their responses can all be rated on a standard scale. Where unstructured interviews are concerned, candidates are not necessarily asked the same set of questions. Comparing the quality of responses from one candidate to the next therefore becomes extremely difficult.
Structured interviews are seen as particularly effective when they are conducted as a behavioral interview (asking, for example, how in the past, the candidate displayed leadership skills, showed initiative or persuasiveness, for example) rather than as a situational interview (how the candidate says they would respond in a certain hypothetical situation) (Barclay, 2001).
The argument here is that it is difficult for a candidate to fake a response in relation to something they actually did in the past. In the extreme, the interviewer could verify candidates’ responses with former employers.
By contrast, when interviews are unstructured, it is difficult to assess the responses given by candidates in any systematic manner. The process becomes highly subjective, and it is this subjectivity that reduces the validity of the process. Researchers have found that the subjectivity within the unstructured interview process takes a number of forms:
• Expectancy effect: interviewers develop an expectancy of the candidate based on prior information, for example from their application form. This expectancy, which can either be positive or negative, can cloud the interviewer’s judgment of the candidate during the interview.
This in turn can introduce ‘confirmatory information-seeking bias’, where the interviewer deliberately sets out to use the interview to confirm their prejudgment of the candidate.
• Primacy effect: interviewers tend to be much more strongly influenced by what the candidate says at the start of the interview than later on, and they will make decisions on candidates within the first few minutes of the interview. One study conducted over a 10-year period at McGill University suggested that interviewers make up their minds on candidates, on average, within the first four minutes of the interview (Webster, 1982). Interviewers then use the rest of the interview to confirm the snap judgments made early on.
• Contrast effect: where the previous candidate was exceptional, this can lead to the following candidate being rated poorly. Conversely, where the previous candidate was exceptionally poor, this can lead to the following candidate being given a high rating.
• Quota effect: in some instances, interviewers have to fill a quota of successful candidates. If this quota is filled early on, candidates interviewed later on in the process are less likely to be successful, irrespective of their performance relative to earlier candidates.
• Similar-to-me effect: this refers to the phenomenon of interviewers preferring candidates who have similar biographical background and attitudes to themselves. Such a situation leads to the potential for race, age and gender bias. It can also lead to a situation in which there are too many like-minded people in an organization, which in itself can have negative performance consequences (see ‘Managing diversity at Marks and Spencer, and British Telecom’ on p.42.)
• Personal liking bias: where an interviewer develops a personal liking for a candidate on the basis of non-relevant common ground (sporting interests, for example) irrespective of the candidate’s suitability for the job.
• Physical cues: for example, wearing glasses is often equated with intelligence.
• Ability to recall information: in an unstructured interview situation, interviewers are in the position of having to think up the next question, while simultaneously attempting to commit to memory the answer the candidate is giving to their previous question. As a result, it becomes extremely difficult for interviewers to recall information once the interview has been completed, particularly in instances where the interview is conducted on a one-to-one basis. The result of this information recall problem is that the interviewer will end up making decisions on candidates on only a fraction of the information imparted.
In sum, unstructured interviewing is seen by recruitment and selection experts as the hallmark of an incompetent interviewer, and many experts argue that they should not be used to make final selection decisions.
However, the majority of companies continue to use unstructured oneto- one interviews as a basis for assessment, despite the fact that they are so poor a predictor of future job performance. While the vast majority of companies use interviews of one sort or another, estimates from the US suggest that only about 35 per cent of companies use structured interviews (Cascio, 1991) – the implication being that the remaining 65 per cent of companies rely on unstructured interviewing. It would seem that recruitment and selection in many companies continues to be carried out by recruiters who remain unaware of the problems and complexities that exist within the interview process. This may be because small companies do not often have HR experts and in any case many line managers are responsible for selecting their own staff. The continuing use of interviews may also have something to do with the expectations of candidates. Would you take a job if an interview were not part of the selection methods?
Interviews enable two-way communication, and candidates can ask questions as well as answer them.
Follow