A BBC article caught my eye. It was by the BBC’s business reporter Suzanne Bearne covering an interview she did with the leader of family high-street firm Timpson Ltd.
Timpson is in every town and shopping centre and employs around 3,000 ‘colleagues’ in 1,325 branches. And the reason for our interest is what’s claimed to be the firm’s “rather novel approach to recruitment”.
Suzanne Bearne reports that Timpson’s interviewers have one criterion in selection - they seek to determine which Mr Men character the candidate most closely resembles. If they’re Mr Grumpy, they’re shown the door. Only Mr Happy and his friends get hired.
So what’s the sense in this?
Interviewee John Timpson suggests that “you can train someone to do a job (but) you cannot train their personality”. That’s true. That has scientific support given that personality is generally considered to be substantially fixed in early adulthood. Personality determines how we will behave in the various scenarios we experience day to day. And John Timpson is not making the common mistake of confusing personality with attitude. Attitude is a function of environment - and that’s created by past experiences and by the current employer and its working practices.
Mr Men (and Little Miss) and Personality
If indeed the Timpson method of determining a candidate’s Mr Men character can be done in any objective fashion, and if it truly reflects personality, science tells us that it might yield a predictive validity of about 0.4. Predictive validity is a measure of the success of a selection method in picking winners.
At 0.4, selecting Mr Happy is a hell of a lot better than a traditional, unstructured interview asking trite, canned questions that tell, in truth, little about the candidate. But it’s not as good as could be. To investigate Timpson’s method further, we see that all down-selected candidates - presumably Mr Happys only - are given a trial for a morning in a shop. Ah, so selection is not just on the basis of personality!
But Competence Too
By asking a candidate to work in the shop, the interviewer gets a chance to assess their people skills (competency), their ability to use basic machines (competency) and their knowledge of business routines at point of sale (competency). So they are also assessing competencies. And adding personality and competencies through work sample tests elevates the predictive validity to around 0.55 - good but not quite the best available.
So it seems that John Timpson’s approach to selection has sound scientific support - though he arrived at it by accident and presumably trial and error.
So where’s the rub?
Prone to Error
Firstly, many jobs can’t be trained for in anything like reasonable time: pilot, surgeon, scientist, engineer, software developer, school teacher, butcher, electrician - the list is seemingly endless. And even those that can be learned on-the-job need intelligence to facilitate that learning. So for a host of jobs, recruiters must assess intelligence and task-related skills and knowledge. Adding these to the mix secures the best predictive validity available - around 0.6.
Secondly, lay assessment of personality is prone to error. Telling a Mr Happy from a Mr Grumpy might seem simple enough, but happiness can be feigned. It’s likely to come down to liking. If the interviewer likes the candidate and the candidate seems happy, that may be enough. The predictive validity in such a case is likely to erode from the really-very-good 0.4 to a paltry 0.2. It takes controlled testing to assess personality to achieve anything like a valid outcome.
Determining Personality Using Mr Men
So is the Timpson method novel?
No. It follows established science.
And will it succeed in picking winners?
Possibly. It all depends on how it’s done. Untrained managers picking Mr Happys is unlikely to be of high validity.
Looking online for interview questions used by Timpson, there seems to be a mix used of trite and very sensible questions that might indeed tease out personality if used by a trained manager. Many of this latter category are typical of questions used by organisational psychologists, so it shows the sort of training that might be needed to interpret candidate responses.
So how do other managers exploit what John Timpson has used?
Helping Managers Select Using Valid Methods
A set of four assessments are needed to predict winners: general mental ability (or intelligence in lay terms), personality, task related skills and knowledge and task related behaviours. The emphasis on each may vary according to the needs of the job being recruited for. Then it’s down to how this set is built into assessment instruments with suitably high integrity.
General mental ability and personality can be assessed against what’s needed to excel using psychometric tests. And skills and knowledge are assessed using carefully constructed job-related work samples. Behaviours are best assessed using scenario-based interviewing - though using job-specific questions. We’ve written extensively in our blogs and papers on how this might be done.
If you want to talk recruitment and recruitment methods, call us.
 Predictive validity is the measure of a selection instrument’s ability to predict candidates that will excel when in the job. A predictive validity of near-zero would be useless and little better than guessing. A predictive validity of 0.6 is about the best possible and is arrived at by adding several good instruments together. More instruments add incrementally towards, but never achieving, the theoretical value of 1.
MR. MEN™ LITTLE MISS™ Copyright © 2017 THOIP