In this article for IFA Magazine, Louis Williams PhD FHEA, Head of Psychology & Behavioural Insights at Dynamic Planner, takes a deep dive into the use of risk profiling tools – including questionnaires – as he reminds us why having robust scientific methods to underpin the process is essential.
Advisers must consider what level of risk tolerance, capacity, knowledge and experience a client has in order to ensure they are taking an appropriate amount of risk, but also that the potential return of the investment fairly compensates for the risk being taken. This process is undeniably complex, and as risk tolerance is domain-specific, for example, risk-taking attitudes towards our finances can greatly differ to attitudes towards our health, it’s important that both the design and the content of questionnaire items used to assess attitudes are adequate.
Risk-profiling should be a multi-step process that starts with a psychometric questionnaire followed by general conversations that review these answers in more depth. Psychometrics is a scientific field that combines concepts from psychology and statistics. In using these questionnaires, it’s easy to think it is simply about the end product: the measurement of someone’s attitude, but the reason psychometric questionnaires are used is because developing a psychological measure is not simple, there is a scientific approach that involves psychological theory, data testing and statistical analysis.
The design
The variable being explored “attitude to risk”, is known as a latent variable, it cannot be directly observed, similar to personality and happiness. Therefore, clients are provided with items where responses to these can infer their willingness to take risk. For example, the Big 5, a measure of personality, consists of factors like extraversion and neuroticism with relevant items that have been tested and deemed valid for providing insight into these factors and the latent variable, personality. Dynamic Planner use a similar approach when it comes to measuring attitudes to risk.
When designing a psychometric questionnaire, scientists must test numerous items with a large cohort and use factor analysis allowing items to cluster into different factors dependent on how individuals have responded. This is an iterative process where items are removed if they are redundant or not useful for measuring the latent variable and each factor within the model. Finally, the individual items selected to be relevant as a measure are analysed, including their impact on the overall questionnaire.
On the surface, a questionnaire can appear to be well-designed, but does it measure what it is intended to measure? A risk tolerance tool might aim to measure an individual’s willingness to take risk when, in fact, it measures something entirely different, such as a client’s investment time horizon or their spending preferences. So, a great deal of thought needs to go into the format of the questionnaire and how it is presented to clients, but also the statical testing of the question set to help ensure it is specifically assessing what we want to.
The content
It is vital that a measure of attitude to risk assesses multiple dimensions, which further emphasises the need to use a psychometric approach. Academic research highlights three key factors to consider when assessing attitudes to risk; what motivates people to take risk (drivers), the personal characteristics that may prevent people from taking risk (constrainers), and the environmental factors that may enable risk-taking behaviours (enablers). Whilst considering these factors we must also be mindful of the ways in which attitudes can be formed and therefore explore these views with questions that allow clients to express their thoughts, behaviours and feelings associated with financial risk.
Dynamic Planner use a variety of items that gives insight into these factors and a client’s overall attitude to risk. Items that assess client’s self-identity when it comes to taking risk, that question their fear of missing out on potentially profitable investments. Items that present the prospect of larger gains but at the expense of taking greater risk, that capture client’s abilities to tolerate uncertainty, and help determine their emotions towards taking risk, whether positive or negative.
Some final considerations
A psychometric questionnaire is not just a set of questions. Unfortunately, questionnaires exist without valid measures, academic theory, data analysis and statistical models that can continue to be tested overtime. It’s essential that we use scientific methods and consider the development behind the questionnaires we use with clients, measures which have explored the manner in which different populations respond to items, understand how items cluster and work together to define multiple dimensions of the variable being measured, capture which items are not so relevant or are in fact redundant, and how items impact the overall measure that is being designed.
