Illustration

Simon has applied for a production manager position. One of the psychometric tests he did was a cognitive test which tested his General Mental Ability (GMA). GMA is an excellent overall indicator of cognitive or problem-solving capability and correlates strongly with success in managerial roles. The test consisted of 30 items or problems that Simon was required to solve. Simon managed to get 26 out of the 30 items correct. This is his RAW SCORE. Simon’s raw score of 26 was then compared to how well a large sample of managers did on the same test. The performance of these managers is shown in the table below. This is the NORM TABLE supplied by the test developer who compiled it in the process of developing the GMA.

Getting back to Simon

Remember that Simon’s performance of 26 (raw score) correct out of 30 items converted to a stanine of 7 (standard score) when compared to the Manager Norm Group. As a percentile, this stanine of 7 would place Simon at the 89th percentile, meaning that his GMA is better than 88% of other managers. No doubt Simon is quite an impressive candidate with respect to General Mental Ability. However, if Simon had achieved a stanine of 4, it would mean that his GMA is only better than 39% of other managers, but worse than 59%. Most managers would have better GMA than Simon and would be expected to out – perform him on problem solving tasks.

Therefore

Because psychometrics use normative metrics, expressed as stanines (a score between 1 and 9) or stens (expressed as a score between 1 and 10); we can confidently know whether that psychological capability is below average, average or above average compared to the norm group. This helps us to understand whether a candidate is weak, average, or strong with respect to the psychological capabilities required by a job. This is vital because psychological capabilities play such a key role in determining individual job performance.

Furthermore…

It enables us to make objective comparisons between 2 or more candidates being considered for the same job. Let us assume that Simon was competing with Lindiwe and Gordon for the production manager position. All three were assessed for their General Mental Ability.

• Simon achieved a stanine of 7 compared to the managerial norm group. This means he is above average and more capable than 88% of other managers to engage in problem solving.

• Lindiwe achieved a stanine of 6 compared to the managerial norm group. This means she is high average and more capable than 59% of other managers to engage in problem solving.

• Gordon achieved a stanine of 3 compared to the managerial norm group. This means he is below average and less capable than 76% of other managers to engage in problem solving.

Assuming that Simon, Lindiwe and Gordon are equal in all other respects.

1. Who do you think the best candidate is for the job?
   

2. Now let us assume that all three candidates achieved a stanine of 3 on the GMA. Would you select any of them for the position? If not, why not?

Answer to the first question:

Since:

• The candidates are equal in all other respects and,

• GMA (General Metal Ability) ability is such a key driver of job performance,

• It is obvious that Simon is the best candidate for the job.

Answers to the second question:

• It would not be advisable to select any of the three candidates if all three achieved a stanine of 3 on the GMA since all three have below average cognitive capabilities compared to other managers.

• Why risk appointing anyone of them when cognitive ability is a key driver of job performance and we know with confidence that all three of them are likely to be substantially less effective with problem solving than most managers, seeing as their performances on the GMA are weaker than 76% of other managers?

A word on samples and norm groups

This is how samples work. Let’s say you want to check the quality of a batch of 10 000 items that came off a production line. You would not attempt to test every item. You would select a statistically significant sample of items for quality testing which is large enough so that the results of that sample can be generalized to the entire batch.

This is exactly how psychometrics work. A candidate’s scores are compared to the norm group. The number of people making up the norm group is large enough to represent the entire population from which it comes from. So, if the norm group is for example the general working population, then we can OBJECTIVELY know how a candidate’s performance on a test squares up compared to the general working population. When using stanines or stens we can easily see whether a candidate is below average, average, or above average compared to the general working population. When we match the stanines or stens to percentiles, we can see where an individual ranks against the norm group – in this case the general working population (top 4%, or bottom 11% and so on).