Applying Naturalistic Decision Making to the Financial Industry: How People Make Decisions in Highly Complex, Real-World Environments
- By Anna-Marie Leslie
Psychologists often use experiments to test their theories, but in the field of complex decision making it is impossible to replicate real-world behaviour in the controlled environment of a lab. To address this problem, around 30 years ago a movement emerged to conduct field research to investigate strategies that people use to make high consequence decisions in difficult situations. The result of this research is naturalistic decision making (NDM).
NDM is a framework for studying how people make decisions and perform cognitively complex functions in situations marked by limited time, uncertainty, high stakes, team constraints, unstable conditions and varying amounts of experience. Gary Klein, a pioneer of this approach, began by working with firefighters to investigate what led them to take decisive life-saving actions. Klein built up timelines of events to identify the signals that experienced firefighters based their seemingly ‘intuitive’ decisions on, such as the sound of the fire, the location and intensity of the heat and the movement of the smoke. Next, Klein developed methods for capturing what the experts were thinking during these complex, critical events where they were under extreme pressure. That information was then used to design decision support systems for training less experienced people faster and more effectively.
Nowadays, the international NDM community is one of the most established decision research networks in the world and the same methods are being used across a wide range of industries. NDM is well established in organisations where the wrong decisions can result in fatalities such as first line emergency responders. It is also well used in complex operating environments such as the military, security and aviation, as well as in contexts where information is sparse and sometimes confusing such as the field of intelligence. In aerospace, NDM methods may be used in training for space missions and the Defense Advanced Research Projects Agency (DARPA) in the U.S. is thinking about how NDM methods could complement artificial intelligence. Even sports coaches are using the methods to improve teamwork and performance. The financial industry faces similar challenges to many of these organisational contexts, and as such, Nasdaq is an important part of the international NDM research community. Our Head of Behavioural Sciences (and Co-Founder of Sybenetix) Wendy Jephson was recently invited to give a keynote address at the International NDM conference, of which Nasdaq was a key sponsor. In conjunction with leading academics, we have also presented our research to the wider academic community at this conference.
A Step-Change in Tech Design
NDM methods are able to offer a step-change to the way we work as designers of technology. Traditionally when gathering requirements directly from the end user, technologists encounter crucial problems with both novices and experts: novices don’t know what they need their technology to do to enable them to become an expert, and experts often work so intuitively they don’t think about what, how or why they do what they do. Therefore, when people describe their actions, they omit important decision points. If we instead rely on policies and procedures, we find that they are typically written in isolation with what people think should be done. But they often don’t envisage the nuances encountered in daily life and the workarounds people do to complete specific tasks. Therefore, the failure to understand the issues that Users encounter, and how they navigate around them, results in poorly designed technology and inefficiencies and inconsistencies in outcomes.
At Nasdaq we take a different approach. Our behavioural scientists use NDM methods to work with experts to unpack all the factors that go into fast-moving, complex decisions where information is missing and data needs to be collected, collated and assessed before taking the next step.
In trade surveillance and compliance, millions of messages are transmitted in the capital markets every day. Some of them contain structured data such as orders, cancels and amends, while others contain unstructured data, such as electronic communications. Analysts are challenged to cut through all this noise and separate the many false positives from real cases of market abuse. Using NDM methods, our behavioural scientists at Nasdaq have captured not only the tasks that their clients perform – from reviewing alerts through to escalation and investigation – but also the complexity of those scenarios and investigations. The output of this analysis is a detailed blueprint of real-world behaviour and decision making as it happens. We have used this blueprint to design the best technological solutions to meet our users’ needs.
Understanding the full complexity of human decision making and combining it with in-depth analysis of the financial domain and organisations is the only way to design best-in-class products that truly meet users’ investigative needs and unleash the full capabilities of next gen AI.
Other industries have been using these methods for decades, and now is the time for financial firms to recognise the value of the research and embrace these methods as Nasdaq is doing as it continues on its path to Rewrite Tomorrow.
About the author:
Anna Leslie joined Nasdaq and the Behavioural Sciences team in 2018. The team combines expertise in behavioural science, financial domain knowledge and advanced analytics to bring diverse thinking and cross-industry experience to solve some of the biggest challenges in financial services. The team works closely with Alpha clients using behavioural science methodology to dive deeply into complex challenges. Together they are designing innovative technology for Conduct Risk that can deliver efficiency, consistency and greater resilience to organisations.
Prior to joining Nasdaq, Anna led a world leading multi-disciplinary team of behavioural scientists who provided expert assessment and consultancy advice to law enforcement and defence operations for the British Government. Anna has expertise in applying a broad range of behavioural science theory to complex investigations and security operations. Trained originally as a forensic psychologist specialising in criminal behaviour, she is now drawing on the breadth of behavioural sciences to solve complex Fincrime problems in context. Communications patterns, understanding relationships and what motivates them, interviewing and detecting deception are all areas of particular expertise. As well as working for Nasdaq, Anna is also a Research to Practice Fellow for the Centre for Research and Evidence on Security Threats (CREST) at Lancaster University, UK. Here she uses her deep understanding of both the behavioural sciences and the operational and investigation space to translate research into practical training and advice for security experts.