Formulas, guidelines, hiring decisions

I’m struck by an article by Julian Baggini in this month’s Prospect. I'm a fan of Baggini. Here, he argues that recipes constrain us and stop us exercising judgment, which over time makes us less wise, less able to discern. That’s recipes of all sorts, not just cooking. Checklists and rules may be crucial for some things – technical things like flying a plane or manufacturing a gas appliance – but in the main, ‘we have become too prescriptive. We should stipulate only what must be stipulated and leave the rest to discretion.’

Included in his list of situations where guidelines stifle wisdom is recruitment. Basically, he argues that relying on tests or standardised criteria reduces candidates to their lowest common denominator and misses what makes them individuals.

This line is a marked contrast from Daniel Kahneman’s view, namely that in many ways – including in recruitment and selection – we are generally worse at prediction than very basic algorithms. Kahneman argues this is even the case with experts. In fact, experts can be even less good at prediction, because they can be overconfident and tend to overemphasise factors that play only a nuanced role in determining outcomes. They misjudge ‘the boundaries of their skill’.

And what we all do in predictions, he says, is underestimate the role of luck. We look at very high performers one day and assume they will continue to perform at that high level, whereas the chances are that they’ve reached a peak and next time their performance will 'regress towards the mean'. (Just as very poor performers one day will tend to improve next time, also regressing towards the mean.) He cites research that highlights our low ability to predict people’s performance and suggests that final decisions in recruitment and selection should be based on track records and more objective tests.

There is a good deal of research on selection interviewing and what techniques work best – both in terms of how well it predicts job performance (i.e. its validity) and how fairly it discriminates between candidates (its reliability). Generally, the evidence shows that that the more specific, detailed and structured a selection interview, the better. This gives a nod to our biases and limitations that Kahneman points out.

So, assuming we carry on interviewing (not a huge assumption to make) it should be more along the lines of administering a test than practising an art form. Set questions, applying a scoring system to responses on a range of criteria – all good. Casual interviewing – not. And here I picture an interviewer who doodles on a piece of paper or leans back in their chair, hands behind head; who consciously enjoys the charm of one candidate, laps up the interactions, maybe allows her/himself to pontificate a little, then radiates boredom with the next candidate. What we might be tempted to see as maverick can just be plain sloppy.

What of Baggini’s line? Is there nothing for recruitment in the idea that we should throw away the recipe book? Can't we take into account candidate’s individual zest, insightfulness or interpersonal skills? Does too standardised an approach mean you overlook genuine and relevant talents?

And shouldn't we also apply an understanding of inter-personal dynamics to the interview process itself? Most parts of most jobs don’t reflect the experience of a job interview, so do we want to bias people who naturally interview well? If the role of the interviewer is above all to gauge each candidate’s strengths, shouldn't we adjust our style so that each can evidence their strengths?

My sense is that the best recruitment processes will be ones that (1) are multifaceted, part standardised, part freer flowing, and (2) that make the different elements explicit, so they can be evaluated systematically. But then I get the niggling thought: isn’t that the danger of being overcomplicated?

Thank you for your comments. There may be a short delay in this going live on the blog page as we moderate the comments added to our blogs.