Within talent assessment, there are three core pillars that we strive to achieve: Reliability, Validity and Fairness. These principles have long been accepted as the bedrock of robust assessment and something that all psychologists working in the field hold first and foremost. In more recent times, we have observed a rise in the importance of candidate engagement, speed of completion and cost reduction but those three core pillars remain undiminished. In this article we focus on Fairness.
Why does fairness matter?
Fairness resonates on multiple levels. Treating others fairly and justly is, to many, a basic value we should all be striving for, however, fairness also translates into positive organisational outcomes. Evidence shows that where diversity and inclusion feature as a higher priority, we see companies outperforming competitors (McKinsey), and in the case of public service for example, organisations are striving to reflect the diversity of the communities they serve. Overall, unfair decision-making can cause psychological, health-related and economic disadvantage to those on the receiving end.
How can we reduce bias and increase fairness?
Diversity of perspective.
When designing your assessment criteria and identifying ‘what good looks like’, it’s vital to have diversity of perspective. For example, when undertaking job analysis ensure that you have sufficient minority group representation and never be afraid to challenge. Define your success criteria carefully, being careful not to inadvertently reinforce existing biases by failing to challenge the existing status quo (which may be already flawed from a fairness perspective). Seek input from multiple perspectives to review your success criteria to prevent unintentional bias creeping in. For example, a Leadership competency defined as ‘demonstrating a bold, competitive nature, taking charge and exhibiting a commanding presence’, (whilst perhaps a slight exaggeration of what we have seen!), does demonstrate how easy it is to ostracise specific groups based on gender, cultural and even age-related differences and craft success criteria that will inevitably put certain groups at an immediate disadvantage.
Data driven decisions.
When designing or choosing an assessment, data is key. For any ‘off the shelf’ assessment, scrutinise supporting documentation to gauge any reported group differences and use that to guide your decision-making. When working with an assessment provider to develop a bespoke assessment, be sure to check how they plan to pilot the assessment and what their data collection and analysis strategy looks like. It is vital to collect the correct biodata so that initial analysis can examine for any potential adverse impact.
Review and adjust.
Working with your assessment provider, it is important to review your assessment to look for weak points in the process. If developing a bespoke assessment, this might be as detailed as pinpointing ‘problematic’ questions and removing or rewording them. It might be mean looking at specific competencies and reweighting them to minimise group differences, or perhaps reviewing assessor training and how assessors approach score allocation or how decisions are made. Armed with data, this process can lead to massive shifts in reducing the potential adverse impact in your assessment process.
Ongoing monitoring.
Ongoing monitoring is key to a fair assessment process, regardless of whether it is an off the shelf or a bespoke assessment. In the case of a psychometric assessments, this can include comparing mean scores to look at differences between groups, or auditing your pass marks to look at the proportion of people who pass from each of the groups of interest. Similarly with interviews and Assessment Centres, this kind of data is applicable too.
Get the latest insights on talent acquisition, candidate experience and today’s workplace, delivered directly to your inbox.
.webp)
.webp)
.webp)
