Recruitment in IT is always a hot issue and getting the right person for the job is just as important now as it’s ever been.
Dr Paul Taylor of Waikato University is something of an expert on the best methods you can use to ensure you get the right person for the job whether it’s in IT, or a more general job.
He is a senior lecturer in Waikato’s psychology department, and his research interests include personnel selection and organisational training.
Speaking at a Recruitment and Consulting Services Association breakfast in Auckland last week, he said that while some studies in the 1980s cast doubt on the validity of interviews, reference checks and work samples as good predictors, more recent research suggests they are in fact useful. But he says they are more valid when they are structured.
The first step in ensuring they’re structured is to have well-defined competencies for the job. He says the second factor is to have pre-planned questions which always target those competencies. If you’re still asking things like “What are your greatest strengths? What are your greatest weaknesses?” and “If you were an animal what would you be?”, you’re probably not finding out if their skills match the core competencies, Taylor says.
Interviewers should also have a pre-planned means of evaluating the answers. Anticipate what the answers might be and what would make one answer superior to another. Use statistics, not just judgement, to rate the answers. If there are two or more interviewers then you should rate the person and come up with an average score.
The benefit of the structured interview is that if for some reason you have a gut feeling that one person is right for the job, you can analyse why. Is it, for example, because they’re articulate? If so, is articulateness a core competency? If it isn’t, it’s irrelevant. And if it is, all candidates should be judged against it.
Reference checks should also be structured. He says referees need to be appropriate. Don’t ask for them at the application. The best place to ask for referees is in the interview. You can ask the interviewee to walk you through a situation that is relevant to the job, such as a presentation.
“Then you can ask, ‘Is there somebody I can talk to who’s familiar with your work in that project?’ That way, you’re targeting your referees to the kinds of competencies you want to ask about.”
Taylor says this approach will also ensure interviewees are honest about their achievements because they know you’re going to check on them.
The questions you ask referees must also address the critical job competencies.
If you want to know about leadership abilities, ask if the referee has seen evidence of this and then ask not only about how the candidate has done, but how they compare in relation to other people in the same or similar jobs. This enables you to get away from the positive skew where everyone is “great” at everything.
Taylor says mental ability tests are quite good predictors of job performance where there cognitive ability is important. Their disadvantage is that studies in the US have shown they give non-whites poor scores. Taylor knows of no similar studies in New Zealand, but believes the issue will be aired more often, particularly from an equal opportunity perspective. “As more and more Maori and Pacific Islanders move into management positions I think it will be an increasing concern. Someone who hasn’t grown up in the culture [on which] the test is based will not score as well.”
He says personality tests are also most useful if they are targeted at a job’s critical competencies. There is always the concern that some people will try “faking good”, and give the testers what they are looking for.
“If you’re applying for a sales job and they ask you whether in your spare time you would rather read a good book or meet new people, the answer is fairly transparent.”
He says organisational psychologists have tried to make tests less transparent but they tend to become less valid — with those setting and sitting a test not understanding what it means.
Research shows people can “fake good” easily if they choose. However, even when people do “fake good” it doesn’t seem to influence validity much — they still end up being the right person for the job. Taylor says that could be because you have to be reasonably intelligent to “fake good” well.
One of the best predictors is a work sample — getting the applicant to actually do the work they do on the job. “If it’s a job that involves analysis, get them to go to the library and research and write up a report about what they’ve found.”