When selecting or developing staff, choosing which assessments to use isn’t as straightforward as it may seem. Different assessment methods - such as ability tests, personality questionnaires, integrity tests, situational judgement tests, face-to-face interviews and assessment centres - are all good predictors of performance. Yet each comes at a cost. So how should you choose which of them - or which combination of tests - is best for you?
Let’s imagine you work for an international hostel chain and you want to hire 100 bartenders for new hostels in Cuba and the Caribbean. The role itself is relatively clear-cut. Most open-minded young people who speak reasonable English and Spanish could suit the job and be proficient in the role after two weeks of training. However, your organisation has had bad experiences in the past with unreliable individuals who have either been found drunk at work, stolen cash or been reported for sexual harassment. So there is an understandable need for caution in the selection process.
An external HR consultancy has proposed three different option packages for assessing your candidates: an online integrity test (£8,000); an online integrity test combined with a structured interview (£16,000) and an assessment centre (£24,000). Which of these options should you choose?
The good news is that mathematical formulas exist that can help you to solve this problem, by calculating the cost per hire and your return on investment. If you can estimate certain details such as how many applications you’ll receive, the percentage of employees who have been ‘bad hires’ in the past and how much a ‘bad hire’ costs your organisation, you can then make a simple calculation that will reveal the most cost effective solution.
Quality of hire
However, it’s not just about the cost of assessment. Best practice selection involves choosing the right individuals for the job. The best way to achieve this is to conduct a business impact study to correlate the results of your proposed assessments against the ‘success criteria measures’ of employees who are already productive in the role.
A 2016 working paper by US psychologists Frank Schmidt and John Hunter summarises the practical and theoretical implications of 100 years of research in personnel selection. It highlights the ‘validity’ of 31 different procedures for predicting job performance, as well as the validity of combinations of these procedures. The authors outline the relationship between assessment methods and performance outcomes - and they conclude that a higher score in an assessment predicts a proportional increase in performance.
Often, a key challenge for employers is establishing the ‘cut-off level’ of their assessments. This is the benchmark score over which candidates will pass to the next stage of the selection process - and under which they’ll be excluded. Logically, the higher you set this score, the fewer successful candidates you’ll have. So how do you set a score that is appropriate?
Again, the good news is that practical models are available which can help you to evaluate this. You’ll need to consider a ‘base rate’ of how many candidates are likely to be suitable for the job. If all (or most) of your candidates are suitable, you wouldn’t need a rigorous selection process, as whoever you choose would be a good performer. You can then identify how many good performers will be selected (depending on the base rate, the validity of the selection instrument/s and your benchmark assessment score). You can even calculate the percentage of rejected applicants who would have been bad performers.
Choosing which criteria to measure
Performance should be determined using objective metrics from existing staff, such as sales achievement, ‘percentage of targets achieved’, customer satisfaction ratings or 360 degree feedback ratings. If this objective performance data is not available, manager performance ratings - or other performance criteria - can be used. Ideally, you should use a mix of different dimensions such as interpersonal (how the employee interacts with others), operational (how they approach a task and the outcomes produced) and motivational (their level of drive and engagement). If your existing employees then complete the proposed assessments, you can analyse the results and determine which of the assessments - or which combinations - most accurately predict their job performance.
Undertaking a business impact study such as this will help you to identify the specific assessments that will have the greatest predictive validity for your roles. In other words, you’ll be able to accurately predict the job performance of your candidates, based on their assessment results. You’ll even be able to identify the percentage of revenue that they’re likely to generate.
Different types of business impact study exist, so if you’re considering this option it’s best to get professional advice on which of them will best meet your needs. The great advantage of these studies is that they provide evidence that your chosen assessments are adding value to your selection process - and that they’re providing you with a tangible return on investment.
Mats Englund is Analytics Director at international talent measurement and assessment specialist cut-e, which is part of Aon’s Assessment Solutions group. cut-e and Aon undertake 30 million assessments each year in 90 countries and 40 languages.