Article · 12 minute read
By Amber Williams & Jake Smith – 23rd January 2025
The traditional 4/5ths rule has long served as a common benchmark for fairness in the application of employment assessments. The 4/5ths rule, also known as the 80% rule, is a statistical guideline to help identify potential adverse impact in talent management practices.
This rule compares the selection rates of different demographic groups at a selection stage with the aim of highlighting disparities. If the selection rate for any group is less than 80% of the rate for the reference group (often the majority/more privileged group), this can be taken as an indicator that there is adverse impact.
While the 80% threshold is a useful starting point, at Saville Assessment, we believe it should not be the ultimate goal in implementing fairer assessment practices.
Organizations and assessment providers should aim to ensure that candidates are not unfairly disadvantaged and that all candidates have an opportunity to perform at their best. By striving for greater fairness across all subgroups, businesses foster more ethical hiring practices and contribute to a more inclusive and high-performing workplace.
Bringing together our recent research and articles, here we explore the impact of marginal gains on increasing fairness in assessment practice — small, actionable steps that, together, create a fairer approach to selection and talent management. Specifically, this piece explores how combining assessments, leveraging advancements in Situational Judgment Tests (SJTs), and addressing stereotype threat can support companies to effectively apply marginal gains to optimize decision-making in talent while promoting fairness for all demographic groups.
While a single assessment can provide a strong prediction of future performance, adding multiple assessments that assess different attributes can enhance fairness in evaluating potential recruits and existing employees. Combining multiple assessment tools—such as cognitive ability tests, personality assessments, and SJTs— can help to mitigate adverse impact and increase fairness across diverse groups.
Real improvements in fairness can be seen for subgroups based on gender and ethnicity when we combine assessments. This is in contrast to using an aptitude measure in isolation, which can potentially result in adverse impact.
By evaluating candidates across multiple areas rather than focusing on a single aspect, this creates a holistic approach which provides a more comprehensive understanding of their strengths and potential. This broader evaluation can help uncover diverse talents and support fairer assessment and selection processes.
Combining assessments and distilling multiple data points into one score enhances the validity of talent decisions, maximizing the chance that the right people are placed into the right roles.
By modelling group differences and cut-off scores* based on differences seen in real data, Saville Assessment can provide advice on cut-off scores which go much further in mitigating against adverse impact.
*Cut-off scores are an established score that determines candidate success on an assessment. For example, Saville Assessment’s recommended cut-off score is the 31st percentile, meaning any candidate performing at or above this level can progress to the next stage of the hiring process.
The integrated approach of combining assessments and applying data-driven practices allows organizations to identify talent more effectively while increasing the ability to foster equitable opportunities for all candidates. Attempting to move beyond the benchmarks of the 4/5ths rule, it can demonstrate a commitment to not only meeting compliance standards but also striving for genuinely fairer hiring processes. As we continue to explore strategies that enhance fairness, the role of well-designed tools like Situational Judgment Tests (SJTs) becomes increasingly significant in shaping inclusive and equitable talent acquisition practices.
Situational Judgment Tests (SJTs) are effective tools for assessing potential decision-making skills. SJTs gauge an individual’s understanding of appropriate responses to different job-specific scenarios. We find the design format of this type of assessment plays a critical role in ensuring fairness. Our research shows that reducing extraneous cognitive load in the design of the SJT—the mental effort required to process irrelevant information—significantly improves outcomes for minority ethnic groups.
Research has found that that SJTs with a larger cognitive loading are associated with producing larger score disparities with protected minorities disadvantaged. Our single-item response format that presents one possible action at a time helps to reduce cognitive demands that have been associated with multiple-choice formats. Candidates are able to focus on demonstrating their understanding of how effective a course of action is, ahead of trying to compare multiple pieces of verbal information in working memory.
While the format is designed to be straightforward and accessible for candidates, this simplicity does not compromise the sophistication of the scoring process. Our SJTs utilize a smart, efficient scoring mechanism. This produces a reliable and easy-to-interpret overall score from a low number of items, by comparing candidate responses against those of experts in the role.
A collaboration with a retail company to design a simplified SJT resulted in fairer outcomes for candidates. Among 350,000 applicants, the selection ratio for candidates from minority ethnic and female subgroups both exceeded 1.0 relative to majority white and male subgroups, indicating that the assessment not only avoided disadvantaging protected groups but may have contributed to slightly higher proportions of these candidates progressing in the selection process.
By prioritizing simplicity in our SJT design through adopting single-item response formats we can reduce group differences and support organizations to deliver fairer assessments that reduce adverse impact while maintaining their predictive power. This seemingly small adjustment (marginal gain), delivers incremental improvements that can collectively drive greater fairness. By addressing disparities in assessment outcomes, this approach not only aligns with the 4/5ths rule as a fairness benchmark but also supports the broader goal of fostering equitable opportunities across all candidate groups.
Moving our view from assessment over to other marginal gains, it is important to consider other aspects of the assessment process that impact group performance and fairness in the assessment process. This helps to understand, beyond assessment design and methods, what other possibilities there are to achieve marginal gains.
Our research demonstrates that asking demographic questions after assessments can reduce stereotype priming. For instance, moving demographic questions to the end of tests increased completion rates from 30% to 70% among female and minority candidates, while reducing group performance disparities.
Removing unnecessary reminders of group membership during assessments helps alleviate anxiety and cognitive load, enabling candidates to perform to their potential.
Using language and scenarios that are culturally neutral and accessible ensures that assessments resonate with a diverse audience and avoid alienating specific groups.
Stereotype threat—the fear of confirming negative stereotypes about one’s social group—can undermine test performance, particularly for protected groups. Addressing areas of assessment that can impact this issue is an essential and easy step for creating a fairer assessment process.
Mitigating the effects of stereotype threat helps create assessment processes that better foster environments where all candidates have a fairer chance to showcase their potential and perform their best. These small yet impactful adjustments exemplify the principle of marginal gains—incremental improvements that collectively lead to significant progress in reducing adverse impact. By addressing unintentional barriers that may disproportionately impact protected groups, aligning with the goal of moving beyond compliance with the 4/5ths rule to achieve more meaningful fairness in assessments.
Building on these efforts, exploring the role of user interface (UI) design and user experience (UX) offers additional opportunities to enhance fairness in assessment practices.
Fairness in assessments isn’t only about the content—it’s also about the experience. User experience (UX) design principles, when applied thoughtfully, can reduce barriers and ensure equal access to opportunities. Strategies for UX-driven fairness include:
Designing assessments with clear navigation, intuitive formats, and compatibility with assistive technologies supports all candidates can engage equitably.
Incorporating feedback from diverse user groups during development helps identify potential biases and usability issues, fostering fairness in implementation.
Simple visual designs with minimal distractions enable candidates to focus on their responses, reducing the likelihood of external factors skewing outcomes.
Prioritizing user-centered design in assessment processes takes a crucial step toward reducing barriers and creating a fairer assessment process for candidates. By considering diverse user needs, simplifying interfaces, and minimizing cognitive load, we work towards making an assessment process more accessible and equitable. Beyond compliance with the 4/5ths rule, these innovations reflect a commitment to driving meaningful progress in diversity, equity, and inclusion within assessment design, delivery and experience; and thus talent management decisions.
At Saville Assessment, we are proud to collaborate with forward-thinking organizations that aim to exceed the 4/5ths ceiling of fairness and achieve fairer outcomes for all candidates.
Our work demonstrates that with thoughtful design, data-driven decision-making, and a commitment to fairness, it is possible to go beyond the 4/5ths benchmark. Below are examples of how tailored assessment solutions have helped organizations achieve fairer hiring outcomes with reduced adverse impact:
Assessment Solution: Combined Match 6.5 and Aptitude assessments.
Roles Assessed: Three graduate streams, each with a unique behavioral ranking system tailored to the role.
Outcome: By setting a screening cut-off at the 31st percentile, no adverse impact was observed for female or minority ethnic candidates across all streams. This approach ensured a fair and equitable hiring process, enabling diverse talent to advance without compromizing on the predictive validity of the assessment.
Assessment Solution: A custom Sales SJT combined with Match 6.5 for a robust and holistic evaluation of candidate suitability.
Role Assessed: Sales agents.
Outcome: Screening candidates up to the 82nd percentile using a combined score revealed no adverse impact across gender or ethnic groups, including Black, Asian, Hispanic, and Latino applicants. This innovative assessment approach not only supported fairness but also enhanced the organization’s ability to identify top-performing sales talent across diverse backgrounds.
These success stories highlight how combining precise assessment tools with thoughtful design principles can help organizations go beyond meeting the 4/5ths rule in assessment. Achieving these outcomes is not the result of a single change but a culmination of marginal gains—incremental improvements that, together, make a substantial impact. Whether by setting optimal cut-off scores, refining assessment design to minimize cognitive load, or embracing combined tools to evaluate multiple facets of a candidate’s potential, each step contributes to advancing fairness.
By striving for fairer outcomes organisations can set new benchmarks in inclusive hiring, transforming their talent processes to empower equity and inclusion. This in turn leads to more diverse teams, which can drive innovation, improve decision-making, and more accurately reflect the community or customer base the organization serves.
If you want to discuss anything you have read about in this article, or find out more about Saville Assessment’s services, talk to our expert team today.
Amber is a Managing Consultant at Saville Assessment. She is a Neurodiversity at Work Expert and has a passion for creating inclusive assessments.
Jake is Screening Solutions Manager at Saville Assessment. He has over 15 years’ experience developing psychometric tools and enjoys conducting research that drives the future of assessment.
You can connect with Jake on LinkedIn here.
While many organizations are establishing their stance, AI’s role in candidate assessment remains a topic of debate. In this article we look at five prompts for thought when considering using AI in this space.
19th Dec, 2024
In today’s fast-paced business environment, Talent Acquisition teams are under immense pressure to run efficient recruitment processes and acquire the right skills.
10th Dec, 2024
We look at the increasing importance of personalization in the assessment experience and the simple things you can do to make a real difference.
2nd Dec, 2024
© 2025 Saville Assessment. All rights reserved.