It is clear that psychological investigation has an important role to play in maths education. It is also a good example of how an interdisciplinary approach can be highly effective, involving educational practitioners and a range of sub-disciplines of psychology. Decades of research has taught us a lot, particularly in the context of mathematical cognition and emotional responses to maths. Here I touch on some of the things research has taught us about maths anxiety, along with suggestions for next steps.

__Maths as a stressor__

Many people enjoy maths. However, for many, it incites a negative emotional response. This is why, for a long time, researchers of the human stress response have used pressurised mental arithmetic tasks as a way of stressing people out. Ironically, timed, socially evaluative maths testing is commonplace in classrooms. Anecdotally (and from some qualitative research), many adults recount experiences they had as children involving maths teachers “putting them on the spot” or humiliating them for incorrect responses. Interestingly, such recollections will often fulfil the criteria for what cognitive psychologists call “flashbulb memories”. Recent research has involved investigating maths anxiety in children who have been in formal education for only a short time (e.g. Petronzi et al. 2018). Sadly, through explicit scoring and timing systems, being good at maths soon becomes associated with how fast one is at it; a child may have very low maths self-efficacy because they are surrounded by peers who happen to be very quick, ignoring the fact that a whole class might be quite fast at mathematical problem solving and some variation is a) to be expected, b) too much importance is attached to completing maths problems quickly. Thus, from a young age, children self-identify a hierarchical structure within their class. This paves the way for negative maths experiences. Arguably, assessment-heavy maths education systems are in need of an overhaul and psychological research will be integral to this.

__Cognition and emotion__

Maths anxiety is a good example of the coming-together of cognition and emotion. Much of the experimental or quasi-experimental work that has taken place has been grounded in theoretical models of working memory and attentional processes. However, little research has explicitly attempted to study the underlying mechanisms related to these models. For example, only a handful of studies have researched intrusive thoughts experience during mathematical problem solving (e.g. Hunt et al. 2014). Whilst methodologically tricky, this is feasible and it would be useful to understand the exact ways such thoughts operate. Similarly, more research should make use of objective measures of attention, e.g. eye-tracking technology, to assess the relevance of specific attentional processes in relation to maths anxiety and visually presented stimuli (maths problems, timers, instructions, etc), e.g. heightened vigilance, inhibition, sustained attention, shifting, and disengagement.

__Meta-cognition__

Recent studies have highlighted the relevance of meta-cognition in relation to maths anxiety, particularly in the context of performance. Meta-cognition can be thought of as “thinking about thinking” (Flavell, 1979) and relates to the processes by which learners plan, monitor, evaluate and change learning behaviours in accordance with a given task (Chauhan & Singh, 2014). Morsanyi et al (2019) provide an excellent discussion of maths anxiety and meta-cognitive processes in which they emphasise the impact of maths anxiety on learners’ judgements and decision-making on numerical reasoning and problem-solving tasks. This incorporates consideration of learners’ confidence and cognitive effort, which relates nicely to the concept of avoidance – something that is often associated with maths anxiety (e.g. Choe et al. 2019). The idea of maths anxiety being closely tied to effort was discussed by Skemp in 1971, in which he argued that if a student fails to find the correct solution to a maths problem, an anxious response may initiate greater effort to find the right solution. This, however, can be self-defeating, making understanding the solution even more problematic, thus creating a vicious circle. There is much scope to explore these ideas further and I would recommend the inclusion of measures of student motivation, given this has been shown to interact with maths anxiety to predict performance (Wang et al. 2015).

__Observational approaches__

Much applied research in maths education tends to be self-report based, which is great in terms of acquiring large amounts of data and identifying patterns. However, it sometimes misses more nuanced verbal and behavioural interactions. There is a small body of research that highlights the relevance of studying specific interactions between children and adults during maths learning, suggesting possible “emotional contagion”, whereby children model parents’ negative expressed emotion, e.g. frustration and anger (e.g. Else-Quest et al. 2008). With many parents finding it stressful when it comes to the home numeracy environment, e.g. helping with maths homework, parent-child interaction is something that needs to be explored further. Relatedly, some have argued that (perceived) teacher support is an important predictor of maths anxiety (e.g. Sultan et al. 2015). Research would benefit from observational and longitudinal approaches to investigate this further.

__Population considerations__

The most well-known studies into maths anxiety and attitudes have tended to be based in Europe or the U.S. However, a wealth of research has shown they are many similarities and differences in research findings between developed and developing nations. Where there are similarities between culturally (and socio-economically) different countries, e.g. in maths anxiety, we should not assume the same underlying predictive mechanisms are at play. Further research is required to consider the role of cultural norms and variations in factors such as motivation to predict maths-based behavioural and psychological outcomes. In populations that are culturally and ethnically diverse, e.g. the UK, such factors are certainly understudied. I am currently leading a multi-country survey-based study of Year 11 pupils and we are in need of more UK secondary schools, so please get in touch if you would like to discuss how to get involved.

__Teachers__

Most teacher training programmes fail to include a focus on relevant psychological features of maths education. For example, only minimal attention has been paid to the anxiety teachers experience when teaching maths. Some research has highlighted the relevance of maths anxiety in teachers, e.g. contributing to teachers’ motivation to support children (e.g. Trujillo & Hadfield, 1999), but also indirect links to poorer student performance as a result of stereotype threat (Beilock et al. 2010). It is important to highlight the subtle difference between maths anxiety on the one hand and anxiety towards teaching maths on the other: whilst related, a high score on one does not necessarily mean a high score on the other. For instance, one aspect of maths teaching anxiety is the worry associated with student maths outcomes (which may be related to external pressures placed on teachers) (Hunt & Sari, 2019). Research should focus on the factors that may cause, exacerbate, or reduce maths anxiety in teachers, encouraging greater interdisciplinary working between psychologists and those involved in teacher training.

__Interventions__

Researchers have tested several innovative strategies to help reduce maths anxiety (e.g. Brunyé et al. 2013; Park et al. 2014; Jamieson et al. 2016). However, there is scope to explore many more and I would urge researchers to engage more with schools and other educational institutions. Engagement with key stakeholders in the early stages enables a more partnership-based approach; educators gain insight into the rigours of the academics’ approach, whereas academics develop an appreciation of a fuller range of practical considerations and suitability of various designs and measures. Some suggestions for developing and testing targeted maths anxiety interventions include those that focus on: feedback (e.g. nature of feedback; confidence-based marking); instruction (e.g. pre-test instructions that induce a challenge rather than threat response); cross-subject learning (e.g. maths and physical activity); combined strategies (e.g. the ‘Maths and Me’ workbook approach I am currently working on with Kingsmead Pupil Referral Unit, Derby). That said, I would encourage educators to propose areas and groups that they believe should be targeted.

Thomas Hunt is an Associate Professor in Psychology at the University of Derby and leads the Mathematics Anxiety Research Group www.marg.wp.derby.ac.uk