The sorcerer's apprentices

Stephen Overell profiles the use of automated psychometric testing by telephone at B&Q, and asks whether this recruitment is becoming a tool for breeding social stability in the workplace.

Anyone who applies for a job in one of B&Q's DIY stores is in for a shock. Not for B&Q the terse little letter, asking promising candidates to come in for interview, or informing the unsuccessful of their doom. Instead, candidates will receive a document with a graph on it plotting their personality against population norms on such factors as conscientiousness, cleanliness and integrity, and a commentary explaining why they may or may not be cut out for a life in customer service.

Thanks to an 'Automated Telephone Screening Interview' - a telephone-based psychometric questionnaire - the store claims to be able to tell which of the 200,000 applicants each year might be suitable for one of 15,000 jobs. By pressing numbers on a phone, B&Q tests personalities to see if applicants fit with the kind of 'culture' it wants in its shops. "I prefer to have my closest relationships outside work rather than with a colleague," the system asks, or "I believe most people will steal if they can get away with it." You press five for 'very like me' or numbers down to one for 'nothing like me at all'.

Brave new techniques

Following the 'interview', the system works out a score and generates the document. The successful go on to a database so managers at one of the 320 stores can pick from a shortlist. Only then will candidates be asked about drill bits, paint finishes and pyracanthas.

There is always a section of the HR community who become incontinent with excitement about go-ahead psychological sorcery like this. Imagine, they say to themselves, a whole shop of people sharing personality traits - everybody yes-siring and can-doing and going the extra mile - but not swiping stock or impregnating their colleagues. Fantastic!

But isn't it all a bit too Brave New World? Recruitment is becoming a tool for breeding social stability in the workplace, a kind of pseudo-scientific caste system. The next step might be hatching employees in incubators. It is with a certain cheek, that the company has 'respect for people' as one of its five 'values' adorning wall plaques in stores across the land.

Prospective scoffers should not be too quick, though. B&Q has been outspokenly progressive on HR. It is in the vanguard of employers who have taken up the cause of older workers (they know more about DIY), has very generous profit-share policies and has pioneered flexible working and e learning. Moreover, its automated recruitment scheme has the advantage of consistency and does not discriminate on race or gender. Store managers whittling down a pile of applications by the time-honoured method of caprice and prejudice is not exactly ideal.

The psychometric commentary is part of an effort to provide feedback to applicants (part of psychometric best practice). Yet the most powerful argument in its favour is a simple one: since the psychometric system was adopted in 1999, staff turnover has fallen, from 35 per cent a year to 29 per cent.

Psychometric tests, of course, remain controversial. To some they are wicked because they are darkly accurate, boiling down personalities to their alchemical essence. To others they are wicked because they are inaccurate, with as much predictive veracity for employment as sorting by birth weight.

But assuming the technical bona fides of B&Q's test, it seems to me there are two good grounds for questioning if this sifting mechanism is, well, quite up to the job.

Dubious wisdom

First, it is very intrusive. Prying into the quirks and ticks of human individuality for the sake of an entry-level job does not seem proportionate, let alone wise. How would you like it if you went for a job selling paint, were quizzed about your relationships, and received an analysis of your personality?

Second, it is dubious how this system fits with the guidelines of the British Psychological Society (BPS), the body that supposedly promotes responsible test use.

The BPS code of good practice says test users should "use tests only in conjunction with other assessment methods and only when their use can be supported by the available technical information".

In B&Q's system, more than 150,000 applicants are being rejected on the basis of a psychometric instrument alone.

Fortunately for B&Q, the BPS is not clear what its own guideline means. Does it mean psychometric tests should not be the only tool used to accept or reject someone? Or does it mean a company should use interviews and references in addition to psychometrics in its overall recruitment armoury?

David Bartram, chairman of the BPS steering committee on test standards and research director of SHL, says the latter: it's a question of overall recruitment and B&Q is safe. Yet Colin Selby, a member of the division of occupational psychology at the BPS and a consultant with Penna, says the former: no-one should be rejected solely because of a psychometric test score.

To have such a confused message is a fudge of real psychological genius on the part of the BPS, leaving organisations to invent their own rules - ably assisted by suppliers with an interest in marketing psychometric applications. The truth is neither efficient nor modern: there are significant ethical downsides to relying on psychometrics as an initial filter that never existed with old-fashioned manual short-listing.