Measure what matters  | Thomas.co

Measure what matters hero

In this candid interview, Thomas International’s Head of Science Jayson Darby gives his predictions for the future of HR, and discusses how psychometric assessments can support the drive for increased diversity, and explains how HR can coexist peacefully with Artificial Intelligence (AI).

In your view, what have been the most significant changes in the HR industry in recent years? 

Just before the pandemic hit, we saw a few prevailing trends across the HR specialism; increasing focus on diversity and inclusion and an attempt to build that into processes was already a theme; remote working was already a trend. One of the long-term demographic trends observed in many countries is that people are having ‘portfolio’ careers. People are living and working for longer, so health is becoming more important. There is more lifelong learning and learning and development throughout people's careers. Technological development resulted in greater automation and interconnectedness between people. People were learning skills to do jobs that don’t even exist yet. All of these macroeconomic trends were creating challenges for HR and putting it in the spotlight.  

Then the pandemic happened and what many were expecting to be decade long trends became the default for everyone overnight. One of the main challenges that we heard from HR practitioners was mandatory remote working through lockdowns. Hybrid and flexi-working have been more successful for teams that were already established. People who already had existing relationship were able to work together because they already had an understanding of each other’s working styles. Now that we are moving to a more blended approach to remote work, we have seen challenges with onboarding and knowledge becoming siloed withing organisations.  

What challenges do you think HR professionals will need to overcome in the future? 

One of the major challenges that I see HR needing to solve is finding the right blend between flexibility and hybrid working, both of which have limitations. HR professionals are having conversations around the best ways to bring people together, and use physical space. Is it for onboarding, training, or just for industry events? A key challenge for HR is onboarding new people and helping them to integrate into new teams remotely. Everyone has a different approach to work and motivation. There are differences in how people cope with adversity and challenges. If they speak to their teams frequently, people are going to understand that.  
 
What we do as occupational psychologists is bring these insights to people more quickly. So if you are trying to support a team now, and you are concerned about how team mates might adapt to a team structure or role that is remote or hybrid, we can provide insight into how individuals like to be motivated and how you can help them be successful, which basically shortcuts the months of observation that you would need to make otherwise. Ultimately that is the role of occupational psychology; bringing the value of psychology to HR practitioners and Learning and Development Specialists to help people make decisions quicker, and help people be happier and more productive at work.  

What does the future of work look like?  

When we look ahead to the future of work, I may be biased because I'm a business psychologist, but I think that we are only going to do more, more quickly, in a more complex environment. So I think that it would be prudent for all HR practitioners leaders to consider the tools they have available to help them navigate through these changes. Even if this pandemic ends today, it has changed the working world by proving that in many roles, people can work effectively on a remove highly remote basis, giving us far greater scope for flexibility in the workplace. I think we will see greater focus on measuring success based on outcomes and productivity, rather than hours present. Flexibility has many benefits, including helping working parents, people with chronic illness and geographically remote regions engage in the economy more effectively. 

The world has changed very rapidly because of the pandemic and it is the role of HR to support people through these changes. One of the most progressive tools with which we can equip ourselves is psychological science. Psychometric assessments can provide engaging experiences that allow people to learn more about themselves as professionals. Understanding how to motivate and manage your people is never going to be simple, but if in less than ten minutes we can tell HR professionals about how to motivate and support their teams, that is going to give them an advantage, regardless of the challenges they are facing. We may not know what lies ahead, but we do know how to help people be more effective they face adversity.   
 
Following the pandemic, people will expect more flexibility, but hybrid work is not going to be universal. Some people will enjoy the changing working practices and others will struggle with them, so I think the next challenge facing HR will be how to support flexibility for that range of responses. It would take hours and hours to figure out how each of us responds in different environments and works under pressure, and workplace psychology can give HR leaders a head start. In essence, that is what Thomas has been trying to do for the last forty years. Since the eighties, there has been huge technological advancement, and Thomas has been trying to bring the value of science through our software to help HR professionals make better decisions and provide personalised support. 

Diversity and inclusion have become increasingly prominent on companies’ agendas. Do assessments like Thomas’ General Intelligence Assessment (GIA) help or hinder the motion for more diversity?  

Thomas’ fluid intelligence assessment is registered with the British Psychological Society, so we can be confident that it does not discriminate against any protected characteristics. When it comes to neurodiversity, following best practice in the use of assessments is very important. We need to be cognisant of making reasonable adjustments. Dyslexia for instance is on a spectrum. There could be many people with mild dyslexia whose GIA scores would never be impacted, but someone with more severe dyslexia might require reasonable adjustments to their score. Whilst the assessment itself is unbiased, a business also has to be aware of the process that wraps around it. A company that signposts that they are accommodating and will consider reasonable adjustments for physical and mental factors that might affect candidates’ access to work is going to have the best possible outcomes. If a company lacks this signposting, or the process itself doesn’t give candidates the opportunity to make reasonable requests, that’s where the risk lies.  

As an assessment provider, Thomas sits in the centre of a Venn Diagram in which we are focused on ensuring our assessments don’t discriminate, and also on ensuring that we provide good advice and recommendations to our clients so that their recruitment practises are also fair. Especially when hiring at scale, a very subtle difference can be magnified, which is why there are such high standards for psychometric testing, and why organisations like the British Psychological Society conduct reviews of psychometric assessments. Yet while the assessments themselves may be unbiased, there are still demographic factors that are hard to correct for. When you think about women in tech for example, there is a systemic, educational issue. Are young girls even directed towards studying the courses that they need in that profession?  

The reason that organisations look for more objective tools on which to base decisions is because it is humans who are by nature biased. I think it is very unlikely that any organisation is purposefully, maliciously trying to discriminate. Bias is more often unconscious. Companies don’t realise that the tools and methods they use could be disadvantaging people. This is why it is so important to measure the important characteristics, which are that predict job success, rather than just looking at CVs. It's about levelling the playing field. The advantage of using psychological assessments is that if they are designed well, they measure what really matters, not superficial factors. For example, measuring numeracy skills will predict success in a role like accountancy. Since demographic differences don’t affect ability, your candidate pool should reflect the general population.  

Should HR professionals be concerned about Artificial Intelligence? 

It’s about ethical AI, because people talk about ‘Artificial Intelligence’, but the machines were programmed by humans often using logic learned from the real world, and the real world is biased in many ways. However, machine learning can also help people make better decisions. Take the example of an experienced judge who is well versed in case law. Machine learning can process so much more information, which can inform the judge’s decision. Now we could say that the model has still been created by a person and chose how to prioritise that information, so there is a challenge there. Things aren’t perfect, but I think as with any new technology there are ways to build models in an ethical way. One of the things that we are exploring as Thomas is enhancing the experience of our content by using machine learning to validate the accuracy of our reports. We constantly refine and test the model using the feedback we receive to provide HR professionals with better recommendations. But ultimately, we don’t take the decision away from them. Progressive providers are stress testing any automation that they build into their system to make sure that it is defensible. We are always exploring whether a person is likely to be disadvantaged.  

The danger is when machine learning is used for automated decision-making. You hear horror stories about this at industry conferences. For instance, a video interviewing technology used facial recognition technology that was less accurate for non-white faces, because it had been trained on predominantly white, male faces. There may have been no malicious intent there, but was a very negative outcome. There is a danger of automating decisions that should rest with a human being. This is especially true when important life decisions are involved, such as people’s access to education or employment. HR professionals are aware that they have a duty to be fair when making employment decisions.  

There are also questions about losing the human touch. If you had too strict a model, it may say, ‘don’t take this person forward they don’t meet the criteria’, but a person may say, ‘I think they’ve got great potential’. A human might want to consider whether they have the ability to invest in training and supporting this person, whereas an AI might say ‘no’. An important consideration for anyone developing this technology is the strictness of these cut-off points. A good example is our star rating system, which provides recommendations for goodness of fit. We don’t automatically move people forward or not, we just present information about who has a better match with the characteristics that relate to success in a role, and allow the HR professional to explore a wider set of information with candidates. We have to be ‘eyes open’ about the limitations of technology.