Richard Donkin .com
Donkin on Work
Donkin on Fishing
Donkin on Travel
Donkin on Sailing

Donkin Life
The Future of Work
Tight Lines - Fishing Blog
Cardinal Points - Sailing Blog
About me
Contact me
Public Speaking
Media Clinic
Blood, Sweat & Tears
Children's Book
Future of Work

Connect with Richard Donkin at Linked in

Donkin on Work - Psychometrics

May 1994 - Psychometrics and recruiting

Why do recruiters place so much store on psychometric testing? It is a question puzzling many would-be barristers who, in spite of sometimes dazzling academic records, have failed to win a place at the Bar law school this year.

One of the students who failed the selection system, which included the use of a critical reasoning examination - a type of ability test - was on Tuesday given leave to have a judicial review of the decision. This will examine the revised entry procedure, used this year for the first time in awarding the 800 places on the one-year vocational course that is the main requirement for entry to the Bar.

The irony is that the new system, devised partly to sift out perceived prejudices, particularly against women and ethnic minority candidates, appears to have given minorities no better chance of entry and perhaps less than before.

There may be question marks, it seems, over the design of the selection system which was intended to broaden the basis of recruitment to the Bar. The motive was laudable. But some believe the law school should no longer retain its monopoly as the only route to the Bar.

The Council of Legal Education, which runs the course, is now looking at the possibilities of making changes to the way the tests are implemented. It may not have been unreasonable, however, to introduce additional selection criteria, looking beyond the possession of university degrees.

Part of any new test might involve knowledge of the real world with multiple choice questions such as 'Is Gazza (a) a city in Italy, (b) a place where Palestinians live, or (c) a footballer?' It is only two years ago, you may recall, that the High Court judge, Sir Jeremiah Harman, asked 'Who is Gazza?' during a court hearing.

Sometimes the way that tests are applied can make little sense. British Rail was last year found to have discriminated against Asian rail guards who wanted to be train drivers. They were given verbal tests which they failed. When it was pointed out that verbal skills were ranked higher in the requirements for the jobs they were already doing than the jobs they had applied for, the industrial tribunal had no hesitation over ruling in their favour.

The problem with psychometric tests, as in many selection procedures, is probably one of balance in the way they are used. Finding that balance, given the continuing debate about the effectiveness of testing, is proving difficult and causing confusion among managers interested in exploring such methods.

Psychometric testing is divided into two areas - ability tests and personality tests or questionnaires. While ability tests in themselves appear to have widespread support among psychologists, there is dissenting opinion about the use of personality testing to measure performance (personality criteria were not used for the bar exam).

Steve Blinkhorn and Charles Johnson, occupational psychologist at Psychometric Research and Development, a St Alban's consultancy, earned the opprobrium of many of their fellow practitioners three years ago for casting doubt on the use of personality testing to predict job performance. They accused those who applied some tests of using pseudo science which, they said, 'bamboozles an unsophisticated public'.

It was strong stuff, particularly since it criticised some of the leading tests on the market. In the April edition of The Psychologist magazine they buttressed their arguments with further claims that 'proponents of the use of personality tests for occupational selection continue to play fast and loose with statistical methods, and to make claims which do not stand up to close inspection.'

Their latest critique in The Psychologist concluded: 'There is no body of public knowledge relating scores on personality tests taken as part of a selection procedure to objective criteria of later performance sufficient to form a basis of routine use of the tests, despite 40 or more years of research. Yet their routine use is widespread and growing.'

Their argument is contradicted by leading test publishers such as Saville and Holdsworth. Roy Davis, the company's head of marketing, said the value of personality questionnaires as a predictor of success in a job was 'statistically proven' but, he added, 'they must be used with other evidence'.

No wonder some recruiters have reservations. Many, however, do not. Employers appear to be placing increasing trust on the recruitment skills of their personnel departments who in turn are often eager to embrace the latest testing methods. 'Going on tests and training courses is like getting a badge in the boy scouts for personnel officers,' said Blinkhorn.

He suggests that psychometric testing is being commandeered by personnel officers in order to better define their professional expertise. If this is true there may well be a danger of too much weight given to testing in selection. It should not become the altar on which budding careers are sacrificed.

Whatever the outcome of the judicial review in the Bar selection case, the argument for a more considered evaluation of tests seems convincing.

© 1994 Financial Times Ltd. All rights reserved

Download as a pdf file

©2006 Richard Donkin - all rights reserved