Psychometrics is a statistics heavy discipline and has been researched for decades. The topics surrounding Psychometrics are endless and so we have carefully selected some of the most relevant areas that may be of interest to graduates and provided some information around these below.

Why use psychometrics?

There is a vast amount of research demonstrating the positive relationship between ability, as assessed using psychometric tests, and job performance. Psychometric tests are used by thousands of organisations across the world and chances are you are likely to come across these at least once in your lifetime!

Psychometric tests also help organisations with a large volume of applicants to virtually halve the number of candidates that they progress through to the time consuming and expensive assessment centre stages. Most organisations use a single Psychometric test to get rid of the bottom 30% of candidates who performed the worst. By using 2 tests, they can remove more than 50% of people before the assessment centre stage.

Consider this example:

100,000 candidates apply for a graduate role and take a Numerical Reasoning test. The bottom scoring 30% of candidates are rejected leaving 70,000 candidates. The remaining 70,000 candidates then take a Verbal Reasoning test and another 30% of poorest performing individuals are removed leaving 49,000. The Psychometric tests have removed over 50% of applicants already.

This year, it is estimated that 300,000 applicants are applying for 30,000 jobs at leading organisations (AGR, 2009). Psychometric tests will probably remove more than 150,000 of these at the online stage.

How reliable are psychometric tests?

As with all assessments, there is not one that is 100% infallible. As such, psychometric tests can also sometimes incorrectly accept low ability individuals, and other times incorrectly reject individuals who may prove to be an asset for the organisation. However, more often than not, candidates are accepted and rejected correctly.

Consider this example:

Candidate 1 - A candidate takes an online numerical test at the onset of a recruitment process for a graduate job. The candidate achieves the 95th percentile which is exceptional. The candidate continues through to the assessment centre stage and is asked to retake a similar test but this time achieves only the 20th percentile. Some may say the candidate cheated – although another reason may that the candidate’s performance had been impacted, at the assessment centre stage, by something significant.

Candidate 2 -Another candidate takes the online numerical test at the onset of a recruitment process for a graduate job. The candidate achieves the 31st percentile which is acceptable (most organisations tend to place a cut-off at the 30th percentile level – below this you are automatically rejected). The candidate continues through to the assessment centre stage and is asked to retake a similar test but this time obtains an impressive 70th percentile. How would you explain this?

Assessment Specialists would say that Candidate 2 has sufficient ability and potential but the lack of familiarity with psychometric tests may have had an impact on their demonstration of their ability in the first test.

Why is practising so important?

When faced with a test for the first time, a candidate will spend some of their mental resources familiarising themselves with the format of the test, rather than focussing on demonstrating their true ability. They are therefore unable to demonstrate their true ability as accurately as they would have had they practised the test.

Imagine this then – what if Candidate 2 (from the example above) had achieved the 28th percentile in the online test (remember: most organisations reject those who achieve less than the 30th percentile)? Yes, they would automatically have been rejected, even though their true ability is of a high level. This is why practising psychometric tests is vital!

‘Classical Test’ theory

Classical Test theory suggests that any assessment will only reveal an individual’s observed score, and that this is not always reflective of the ‘true’ score, as there is always something in the environment that impacts an individual’s performance (error).

Thus for those of you who are mathematically minded:

OBSERVED SCORE = TRUE SCORE + ERROR

True score – the individual’s ability and is always constant for a particular person
Observed Score – the score obtained by an assessment
Error – anything that may have impacted the individual’s performance on a test

By practising the tests, you are minimising the error and therefore increasing your observed score. With zero (not possible) error, your observed score should equal your true score.

Consistency

Have you ever wondered why test administrators read from standardised instructions in a verbatim manner? Again, this is to maintain consistency, which minimises error. If two candidates take the same test independently and are given different instructions; we cannot be confident that the difference in the scores that we observe is down to the different abilities and not down to the variation in instructions. One candidate may claim to have received poorer or better instructions than the other. This is why we have to maintain consistency.

Final advice

Simply getting scores over the 30th percentile is not enough! Some organisations may set cut-offs higher than this. Also, if at the final stage of an assessment centre you find yourself in a situation where the assessors have to make a decision between you and another candidate who has scored exactly the same as you in the assessment centre exercises – the job will most likely be offered to the candidate with the higher Psychometric test score. So make sure you get plenty of practise and try and obtain as high a score as possible in the real psychometric assessment.