Chair at the Association for Business Psychology and Managing Director at the psychometrics firm Sten10
Share this content
Part of the
Brought to you by HRZone.com

People analytics – how candidate assessment tools can refine your recruitment process

5th Jul 2019
Chair at the Association for Business Psychology and Managing Director at the psychometrics firm Sten10
Share this content
People analytics for recruitment
portishead1/iStock
People analytics for recruitment

Looking to enhance your recruiting approach to help you find the right talent for your business? Candidate assessment tools and their use of data and people analytics could help you make better informed decisions about potential new hires.

In an attempt to better match applicants to job roles, employers spend a considerable amount of time and resources testing candidates using a whole range of psychometric and behavioural assessments (from personality questionnaires to written analysis tasks).

Each assessment produces data on the applicants, which employers rely heavily upon to help choose between stronger and weaker candidates.

The theory is that a high performance by a candidate on an assessment should equate to a high performance of that applicant in the actual job. Therefore going through a testing process and analysing the test data should ensure that the most well-suited person is chosen for the job.

Candidate assessment tools: what to look out for

There are many issues to consider when looking to implement a new candidate assessment tool, not least of which is the validity of the assessment.

The good news is that many tests on the market have been proven to have ‘assessment validity’ – the test effectively measures what it has set out to measure and so people’s test results are legitimate. But employers also need to make sure that the tests they choose are relevant for the job role that they are recruiting for. 

For instance, it is all too easy to think of using a simple personality questionnaire to choose between applicants, but a personality assessment may be less relevant than a work sample test when recruiting for a pilot or a surgeon. This is all determined by conducting proper qualitative and quantitative analysis up-front about qualities that top performers in a role share.

It is also worth considering if a test matches your company’s image or brand, and how a chosen test would compare to others in your industry. One would expect a tech start-up to test its applicants differently to a clothing retailer, so new techniques such as gamification may be accepted (or even expected) by some applicants, while it could be a turn-off for others.  

The key question to ask the firm supplying the test is: ‘what research can you provide that shows me this is a valid, fair way of assessing people’?

How can people analytics inform the design and refinement of the recruitment process? 

Assuming the recruiter has fully understood the skills and competencies required for the job role in question, the success of using testing in the recruitment process depends on a thorough understanding of:

  1. What the test data means

  2. How to apply this data to match the most suitable applicants to (each) job role. 

When these two elements are fully understood, a recruitment process can be designed and then, over time, refined to get the optimal outcomes. 

What can test data show? 

Test data can show proportions

People analytics at its most simplistic can look at proportions, such as the proportion of candidates that did well on the assessment or the proportion that failed. 

Imagine an assessment centre setting where a group exercise mimicked a scenario that may occur in the job that is being recruited for. At the end of the exercise the candidate would be allocated a score based on their performance.

The recruiter will have data on people’s scores across all of the assessments at the event and can see whether those who do well in the group exercise tend to do well overall at the assessment centre or not. This can then be used to spur interesting questions about the usefulness of certain assessments at the event and whether they add value or not. 

Test data can show correlations

Alternatively, test data can show correlation figures – the strength of the relationship between variables (for example between test performance and job performance or between age and performance in a certain exercise). The benefit of this type of data is that it can be used to investigate not only the validity of an assessment process, but also its fairness. 

Of course, these tests do not have to be used in isolation and often companies will use data gathered from a range of sources to help them inform and refine their recruitment process. One area that is really gaining traction in recruitment decisions is the use of selection algorithms.

What are the benefits of using algorithms in the recruitment process? 

Selection algorithms are highly complex forms of people analytics that use all sorts of data points to help with the recruitment process. They are often seen in tests such as mobile game-based assessments. There are claims that playing for as little as 20 minutes on certain games can generate large amounts of data on an individual, particularly how they use the information to solve problems. These scores are then used to develop candidate profiles for areas such as creativity, persistence, capacity to learn and ability to prioritise. 

There is lots of scope for growth in this area. The most valuable part of AI recruitment algorithms is that, when applied correctly, they can reduce subjectivity in the recruitment process by basing decision making on objective data.

What are the downsides to candidate assessment tools?

This sounds like a positive way forward in the use of people analytics in recruitment but there is always the flip side to consider. Any technology that relies heavily on data needs vast amounts of it to ensure its validity and its predictive qualities. This is often hard to come by. 

Plus, unless the data input is constantly monitored there is always the potential for biases in society to start being reflected in the algorithms themselves (for example performance differences according to socio-economic status).  

Finally, some may argue that it assesses (and uses) too much personal data and may be considered too invasive (for example videoing people’s facial features and eye movements to determine personality profiles).

Fundamentally, people are employing other people to work with them. While there are great efficiencies that can be gained from using algorithms, there is only a certain level that candidates will accept. You need to be careful not to underestimate the importance of human contact in the recruitment process.

The best advice for HR professionals looking to choose an assessment tool is to think carefully about what you want the test data to reveal and know how you are going to use it in your recruitment decision.

A recruitment tool should be there to help you save time, money and resources. For it to be worthwhile it must add value, so always think pragmatically. People analytics is hugely beneficial, but it has to work for you.

Replies (0)

Please login or register to join the discussion.

There are currently no replies, be the first to post a reply.