Psychometric testing has become more common in the workplace as employers turn to tests to give them definitive answers. Tests are used by employers directly or via search consultants, even and especially for the most senior positions.

Well used, tests (individual or in combination) are one of the most effective methods for helping employers with selection, promotion and development. Also, now that the era of the lone ranger emotionless manager is over, learning how to work together in and build teams is crucial. Advocates say tests are valuable management tools to achieve this.

Many tests are proprietary, and are sold exclusively by consultants or agencies. Each one says their test is the best; the British Psychological Society says tests are the best predictor of performance.

Personality tests are especially popular. But there is a big debate currently raging in occupational psychology circles: many academics and consultants who are great fans of ability tests don't think personality tests are useful. Even some of the strongest proponents of personality tests admit that they must be used with other evidence.

The latest theory is that psychometric tests should only be used in the context of wanting to understand individuals for career development and team building; and that they should not be used just for selection, nor should they be the only selection method.

Some factors to consider when choosing or using psychometric tests: 

• blinded by science

Tests are not neutral scientific tools. Many non-numerate managers are taken in by the mystique of the numbers they generate; the belief can be if "scientific" must be "objective"; not always so. Also, a lot of studies confirming test success are done by test vendors themselves.

• varying standards

Standards vary for testers who are professionally qualified by the British Psychological Society (self regulated industry) which issues certificates for ability testing and personality testing, to bucket shops and the suspicious DIY discs that come free with certain software packages.

• bias in test construction

Subtle biases can exist, eg minorities and women are less likely to guess at an answer than white men.

Many tests devised for the American market are imported here without changes. British managers without a working knowledge of "American" are at a disadvantage eg words like "pep" and "sophomore".

• discrimination issues

Legal problems may arise unless tests are validated by "directly objective" criteria. Employers may find themselves breaking UK anti discrimination laws, where the onus is on the employer to prove an attribute is essential to good performance in task, work and organisation.

Psychometric tests fell out of favour in the US in the '80s over just such radical and sexual discrimination. They are now making a comeback there, but only where employers can show a particular test is relevant and valid for diverse groups of people.

• manipulation by test taker

Test results can be distorted by the mood of a person on test day. The type of question in some tests makes it fairly easy to fake answers eg "do you prefer baths to showers?", "do you consider yourself well organised?"

As tests become more used, those takers with more experience of tests in general and the one they are taking in particular develop an unfair advantage.

• bias of interpreter

Poor or inappropriate interpretation by the user is one of the major problems with tests. There is a danger that they will be used as a substitute for rather than a basis of decision making.

Even properly interpreted, tests can typecast and pigeonhole people who may have broader and more complex personality or potential than the test is capable of showing.

• ethical issues

Many employers are using tests for more than they were originally designed for eg deciding who to get rid of.

The effect of testing on prospective and current employees is hardly ever considered. For example, what does the fact that you give this test to prospects say about your company?

• distance issues

Tests have often been devised by occupational psychologists or consultants remote from what goes on in the workplace and what it takes for successful performance in a particular organisation or situation.

Some HR departments seem to be taking ownership of all psychometric testing in their companies in order to reassert professional authority.

• interpretation issues

Interpretation should not be used as a blueprint for the perfect person generally or any particular group. Another common mistake is to use static tests as predictors.

Sex differences on tests must be interpreted carefully; any difference may not be due to an underlying trait but the way it is expressed eg "do you enjoy going to parties alone?" is biased because society still thinks it is ok for a man but not a woman to behave this way, so fewer women will answer yes even if they do.

• validation issues

Although tests are useful for career counselling and team building, there is little conclusive evidence that test results are related to future job performance.

Up to 75% of tests have no "objective" criteria to prove that high test scores are related to high job performance (eg value of orders generated; volume of output). In most kinds of work these types of measures simply do not exist, or cannot be attributed to one person alone. Tests are validated by subjective measures like bosses rating employees.
 

One alternative to written testing is "power interviewing": structured interviews in lieu of pen and paper tests. UK companies like B&Q and Joshua Tetley use them.

Interviews can include:

• behavioural questions that relate responses to past experiences

• achievement orientated questions looking for specific track record of eg increasing revenues or customer loyalty

• holistic questions that aim to measure the whole person, including work patterns, goals, ability to see big picture

• aims: reveal nature of preferred worksyle

• applications: selection, team-building, career planning, counselling

• format: choose one from each of 90 pairs of statements that best describes taker at work

• scoring: 20 different factors (eg need for change) in 7 key areas (eg temperament)