Job candidates may be hired depending on the order in which they are interviewed, in the same way X Factor contestants who sing later in the show are less likely to be voted off, research indicates. In an article published by Personnel Today, we see that researchers at Cambridge University’s Judge Business School have been investigating what gives winning game-show contestants the edge. They found that the order in which contestants perform plays an important role in their success.
Wyn Llewellyn, Director at ValueFlows Ltd, observed, “An interesting piece of research and a creative and invalid extrapolation – from X factor to recruitment process. A conflicting hypothesis might be – the later you are in the process, the more tired and jaded the selection panel will be, they may have seen a ‘star’ earlier and become biased towards them, etc.
Also consider the following; professional recruiters are working to clear selection criteria and are professionally trained to do so – the public who vote in X factor have neither of these characteristics – they just vote for who they like! Wait a minute – maybe a proportion of recruiters do that too!”
This got me thinking…
So hang on a minute. How many initial applicants for a role would you expect to be able to fulfil the requirements? 1 in 10 maybe? And how often would you expect the interview process to deliver the right results? Most experts would say that a 7 out of 10 success rate of interview process delivering right result is high. Hey, it’s nearly Christmas so let’s be generous and say 8 times out of 10.
Using these assumptions the chances of selecting the right candidate are about 2 to 1 against (and if you do turn the dial to a 7 out of 10 hit rate on interview process the odds against correct selection increase to nearly 4 to 1 against).
Let’s do the maths:
1000 applicants, 900 can not do the job, 100 can.
Of the 900 who cannot do the job, an 80% correct interview process will deliver:
720 of the 900 correctly identified as not able
180 of the 900 as able, even though they are not
Of the 100 who can do the job, an 80% correct interview process will deliver:
80 identified as able
20 identified as not able, even though they are
So if we divide the 180 incorrect able candidates by the 80 able candidates, 180/80 gives us 2.25 to 1 against.
So if the formal interview seems to correlate so poorly (if at all) with future effectiveness, well maybe we’d be better off taking our chances with Simon Cowell et al after all? At the very least this leaves me wondering if we should be considering alternative ways of connecting the right people with the right job role…what do you think?
This would work on the assumption that you forget all about the previous candidates pros and cons before each interview. Add into the mix that most interviewers take notes, most interview panels are more than one person and most selection processes involve a comparison of opinions, then the odds must surely reduce to favour the suitable candidates.
Personally, I end up benchmarking candidates against the previous ones as I go along plus the inevitable tickboxes of course!!
I really like Seth Godin’s take on job interviews: http://sethgodin.typepad.com/seths_blog/2009/08/two-ways-to-hire-and-a-wrong-way.html
Thanks Kolya, a long time has passed since your visit and I apologise for the delay in acknowledging the link you posted. It’s an interesting read thanks.
Doug – I like the way you’re thinking here, however, I’m not sure I go along with your maths. To start, the 70% predictive validity is what is quoted for a fully professional assessment centre process – having a professional sifting of your 1000 candidates down to a manageable short list against strict criteria. You cannot predict the validity of that part of the process as you have not interviewed them or given them a job. Reasons for rejection are numerous – not meeting the criteria, poor CV design, poor expression, missing the deadline, no cover letter, etc.
The predictive validity of 70% refers to the short listed candidates who go through the assessment process. And what it means is that for every 10 people hired, 7 will be performing well in their role – equating to a sound hiring decision. The other 3 will not be performing well in the capabilities sought.
So, what are the chances of picking a good performer from the short list? 70% at best. But what about the high potential people you might have missed in the sift? You’ll probably never know – unless they pop up again applying for another job.