Ben is the chief psychologist at Test Partnership, with extensive experience in consultancy and research. He writes extensively on many topics, including psychology, human resources, psychometric testing, and personal development.
Employee Testing and Selection: Best Practices to Improve Quality of Hire
When it comes to employee selection, the gulf between theory and practice is particularly wide. For decades now, occupational psychologists have known the most effective methods for screening candidates, but organisations refuse to use them. Instead, they rely on CV sifting and interviews exclusively.
Ironically, concerns about employee testing largely centre around validity and fairness, highlighting just how large the academic-practitioner divide is within talent acquisition. Ultimately, however, blame for this divide can be attributed to testing providers and psychologists who fail to disseminate information effectively. Instead, they hold this information closely, packaging it up as consultancy services instead of making it publicly available.
In this article, we will highlight the best practices for employee testing, sharing how to maximise the quality of hire with employee testing.
Which employee tests should you use?
The first and most important step is to choose which specific assessments to include. As a general rule, there are three types of employee tests that should be considered: cognitive assessments, behavioural assessments, and knowledge/skills assessments. Cognitive assessments look at a person’s ability to learn, solve problems, and make decisions, and are especially predictive of knowledge/skill acquisition.
Behavioral assessments look at a candidate’s attitude, disposition, and character, usually in line with a specific model of personality. Knowledge/skills assessments look at specific areas of expertise, such as software skills, theoretical knowledge, and best practices.
The choice between these three assessment formats depends on a few variables. Generally speaking, the more training and onboarding offered, the less knowledge/skills assessments matter. For example, for graduate schemes, their current level of skill is expected to be low, and significant training will be provided anyway, making knowledge and skills tests poor choices.
However, cognitive psychometric tests are likely to be very powerful, as the ability to learn and acquire skills is essential at this stage, more so than any other assessment format. Conversely, if you are hiring experienced freelancers or contractors, who will receive no training or onboarding, then knowledge and skills tests will be far more important.
Behavioral assessments add the most value in roles where either: A) culture fit is essential, or B) the role is relatively simple. Although interviews are the most common way that organisations assess culture fit, interviews are a poor indicator. Realistically, charisma and social skills are the main determinants of interview performance, not culture fit. T
his is particularly important for senior leadership roles, and we strongly recommend using behavioural assessments at this level. In relatively simple work, where there is little autonomous decision-making, skill acquisition, or complex problem-solving, the strongest predictors of performance will be behavioural. For example, the drivers of performance will be behaviours such as reliability, honesty, punctuality, and health and safety adherence, all of which are measurable using employee tests.
When should you use employee testing?
Theoretically, the ideal time to use employee testing is right at the start of the recruitment process. Employee tests are highly scalable and easily automated, allowing you to create high-quality shortlists with very little time investment or administrative burden.
However, in practice, not all organisations have the luxury of testing every single candidate at the very start of the process, nor do they have enough applicants to actually shortlist in the first place. Consequently, the ideal time to use employee testing will depend predominantly on the volume of applicants and their motivation to participate.
For roles where you expect a high volume of applicants, i.e. graduate, apprentice, and internship schemes, organisations can afford to use graduate selection tests early in the recruitment process. Not only does this make the process substantially more scalable and convenient, but the value you get from assessments is directly proportional to how selective you are. If you have 10,000 applicants, and you want to interview 100 of them, your shortlisted candidates will be of extremely high quality, as you are picking the very best. If you can lean into this advantage, we strongly recommend that you do so.
However, for senior management roles or hard-to-fill niche roles, you won’t have enough candidates to be highly selective, and you have to pay close attention to candidate attrition. Realistically, in these situations, the candidate typically has the power in the selection process, and may just deselect themselves if you ask too much.
In this instance, we recommend shoehorning in the assessments wherever you can, perhaps at the end of the process. Ideally, you would combine the assessments along with interviews and other exercises, creating a mini assessment centre. Candidates won’t deselect themselves in person, and they will feel thoroughly vetted if given a battery of tasks to complete.
How should you use the data to screen?
Collecting quality data at the right time is only half the battle, as you only get value from employee tests if you actually use that data for screening. Many organisations, who are either new to the process or are unaware of their utility, may be hesitant to actually screen candidates using their employee tests.
For example, some organisations use the results only in a tie-break situation, or purely to structure an interview thereafter, without actually screening anyone out. This represents an extremely inefficient approach which artificially reduces the quality of hire and should be avoided at all costs.
Given the predictive validity of well-designed employee tests, organisations are well advised to screen candidates out on the basis of poor behaviour. For example, if a candidate is genuinely scoring in the 1st percentile on cognitive ability tests, and they are applying for a highly complex role, there is no world where this candidate would perform well on the job.
Scoring in the 1st percentile on a cognitive assessment is substantially more telling of their future performance than completely failing in an interview, and should never be ignored. Similarly, scoring in the 99th percentile for cognitive ability is more indicative of future performance than acing 100 rounds of interviews, and should be recognised as such.
Additionally, using employee testing to structure interviews adds far less value than you would think. Realistically, the biggest advantage of employee testing is that they measure constructs which cannot be measured via interview, adding incremental validity.
By using the results to structure an interview, you simply give candidates a chance to talk their way out of poor results, reducing the effectiveness of the selection process. Instead, we recommend keeping these separate and distinct, avoiding cross-contamination.
Summary and Recommendations
Employee testing should be considered an essential part of the selection process, and is in many ways more important than interviews. Interviews are a high-touch, time-consuming, and expensive process, reserved only for the highest-potential candidates.
The heavy lifting of the recruitment process should instead be done by employee testing, representing a more scalable and evidence-based approach. When done well, interviews are only needed to distinguish good from great candidates, and never to distinguish terrible from decent candidates.
Ben is the chief psychologist at Test Partnership, with extensive experience in consultancy and research. He writes extensively on many topics, including psychology, human resources, psychometric testing, and personal development.
