Humanising statistics: how emotionally intelligent teaching can transform mathematics anxiety

Each year, as a new academic term begins, thousands of university students will walk into statistics and quantitative research methods lectures and classes carrying a powerful emotional burden. For many, these quantitative courses trigger several negative emotions such as fear, apprehension, anxiety, self-doubt, panic and avoidance. These emotional responses often shape how students engage with learning long before any formulae and concepts are introduced by lecturers.

The anxiety that statistics and quantitative research methods courses can generate for some students is a serious issue, particularly because these courses are core components of degree programmes around the world and play a central role in higher education.

There are ways, however, that educators can support students in overcoming their concerns, with research showing that emotionally intelligent teaching can reshape these experiences, helping students build confidence, resilience and a more positive relationship with these mandatory quantitative courses.

My research with university students studying statistics and quantitative research methods suggests that mathematics anxiety is shaped less by objective learning experiences than by how students interpret those experiences. This insight informed the development of an educational intervention centred on emotionally intelligent teaching, which is the focus of my recent book.

Why are so many people anxious about mathematics?

Anxiety linked to engaging with quantitative work reflects a broader phenomenon: mathematics anxiety. Mathematics anxiety can trigger intense negative emotions and physiological reactions when students are faced with quantitative work. These reactions disrupt their attention and working memory, making it harder for students to process information and apply it to approaching problems effectively.

Students may begin to believe that they cannot understand the concepts being introduced, leading to mistakes in their work that generate feelings of shame, embarrassment and helplessness that subsequently prevent them from asking questions and seeking timely help.

Those who suffer most can end up avoiding lectures, classes and coursework submission, which creates further gaps in their knowledge as they miss out on feedback on their formative assessments. Unfortunately, this ultimately leads to underperformance or even failure in these quantitative courses.

Although students may feel that their anxiety is due to a lack of understanding, research shows that mathematics anxiety is not limited to those with weaker mathematical backgrounds nor is it linked to a lack of mathematical skills – in fact, many high-performing students report experiencing it.

The consequences of such anxiety, however, can be far-reaching, leading not only to individual underperformance but also to avoidance of quantitative work. This contributes to the global shortages of data analysis and quantitative skills that are increasingly essential across disciplines and professions.

Across the OECD countries, the percentage of students identified as mathematically anxious increased from 31 per cent in 2014 to 39 per cent in 2024 (see OECD, PISA, 2023, 2024). This can discourage individuals from entering quantitative fields, contributing to the quantitative skills shortage highlighted by the World Economic Forum (WEF) and the UNESCO Institute for Statistics (UIS).

Despite the rich literature of more than seven decades and several interventions developed by researchers to help mathematics anxiety sufferers, there remains a need for further research.

The maths anxiety gender divide

Female students continue to report higher levels of mathematics anxiety than male students (OECD, PISA, 2023, 2024). These patterns also have implications for equity, contributing to continuing gender imbalances in science, technology, engineering and mathematics (STEM) education and careers. According to the WEF and UIS, women make up only 35 per cent of STEM graduates, and approximately 29 per cent of the global workforce in STEM fields are women.

A common misconception is that fear of statistics and quantitative methods reflects a lack of ability. Additional revision sessions alone may not address the emotional barriers associated with mathematics anxiety. If we want students to succeed in a data-driven world, we need to rethink not only what we teach in statistics, but how we teach it.

This calls for pedagogical practices that promote inclusion and equity by addressing students’ diversity. Educational success depends not only on intellectual understanding, but also on students' emotional experiences of learning. Efforts to widen access, improve student wellbeing and strengthen quantitative skills will remain limited unless they explicitly acknowledge the emotional dimensions of learning. Addressing mathematics anxiety is therefore not only an educational concern, but a strategic priority for higher education systems.

Paradigm shift: from experience to interpretation of learning

Research suggests that mathematics anxiety is shaped more by students’ subjective interpretations of their learning experiences than by their actual experiences.

Students do not simply respond to quantitative content in isolation. They interpret it through a lens shaped by past learning experiences, interactions with teachers and peers, and internalised beliefs such as '‘I am not a numbers person’’. Over time, these interpretations form powerful unfavourable narratives that influence behaviour more broadly.

These narratives can become self-reinforcing. Students who expect to struggle may disengage and hesitate to ask questions, which in turn reinforces the belief that they cannot succeed. This is a classic example of a self-fulfilling prophecy: expectation shapes behaviour, and behaviour confirms that expectation.

In statistics and quantitative methods lectures and classes, this cycle is particularly problematic. Quantitative learning requires sustained engagement, willingness to make mistakes and the confidence to seek clarification. When anxiety disrupts this behaviour, learning is obstructed at multiple levels.

Mathematics anxiety: when learning becomes emotionally charged

Research suggests that mathematics anxiety does not simply affect performance. It changes how students experience learning itself.

Anxiety narrows attention. Students who become preoccupied with worry, shame or embarrassment often struggle to focus on the statistical concept being introduced. Their attention shifts from learning to monitoring potential failure, making even straightforward tasks feel daunting.

Anxiety silences engagement. When students fear being judged, they are less likely to ask questions, contribute to discussions or seek help. Difficulties that could otherwise be resolved quickly therefore remain unaddressed.

Anxiety encourages withdrawal. As uncertainty grows, students may avoid lectures, delay assignments or disengage from coursework altogether. This reduces opportunities to learn and receive feedback, increasing the likelihood of falling behind.

Together, these processes can turn temporary uncertainty into sustained underperformance. Crucially, however, they are influenced by the learning environment and are therefore open to change.

Humanising statistics: a research-informed pedagogy

Humanising statistics means recognising that learning is not a purely cognitive process but also emotional and social. Drawing on insights from psychology, sociology, neuroscience and education, this approach recognises that students' responses to quantitative learning are shaped by cognitive, emotional and social processes. It involves designing teaching approaches that acknowledge students' feelings, support their confidence and foster meaningful engagement.

At the heart of this approach is emotionally intelligent teaching. This does not require abandoning rigour or lowering expectations. Rather, it involves creating conditions in which students feel able to engage fully with challenging material.

Students' responses to statistics are shaped not only by the content being taught, but also by the meaning they attach to previous learning experiences. These meanings develop through interactions with teachers, peers and the wider learning environment. When those experiences are supportive, students are more likely to view statistics as accessible and worthwhile rather than threatening.

Positive interactions can begin to shift students’ unfavourable narratives. When students feel respected, heard, supported and able, they are more likely to reinterpret statistics not as a threat but as a tool they can learn to use. These approaches are not based solely on intuition. They emerged from a theory-driven, student-informed pedagogical intervention that was empirically tested with university students studying statistics and quantitative research methods.

What does humanising statistics look like in practice?

Humanising statistics is not an abstract ideal. It can be implemented through relatively simple but powerful teaching practices.

1. Building positive interactions

Eye contact, open body language and responsiveness to students’ cues can make a significant difference. When teachers are attentive to signs of confusion or anxiety, they can intervene early and appropriately.

2. Encouraging dialogue

Creating a classroom culture where questions are welcomed reduces the stigma associated with not understanding. Dialogue helps students realise that mistakes are a normal part of learning. Reinforcing the idea of viewing mistakes as learning opportunities can be an effective approach to promote a compassionate learning environment.

3. Providing constructive feedback

Feedback plays a crucial role in shaping students’ beliefs about their abilities. Highlighting progress, rather than focusing only on errors, helps students recognise their own development and potential.

4. Normalising anxiety

Explicitly acknowledging to students that some anxiety is normal and may even enhance performance makes a positive difference to their attitudes to quantitative courses. Educators should convey that statistics and quantitative methods are challenging subjects but accessible to everyone. This shifts the focus from ‘‘cannot’’ to ‘‘can’’, reducing feelings of shame, embarrassment and inadequacy.

5. Connecting to real-world meaning

Framing statistics as a tool for understanding real-world problems helps students see its relevance. When students engage with meaningful data and contexts, anxiety is often replaced by curiosity and interest.

6. Encouraging collaborative learning

Research indicates that students learn well through group work, which also helps them develop the core skills identified in the WEF Future of Jobs Report 2025 . The global shortage of these skills has an adverse impact on economic growth. Collaborative learning can create a dynamic and interactive environment and help students develop these skills, which are expected to be crucial by 2030 according to the report. It can promote interaction and inclusion, helping students to move away from negative narratives and towards more constructive beliefs about their abilities.

Beyond the classroom: institutional and policy implications of tackling mathematics anxiety

When students' learning experiences are enhanced, their interpretations can change too. Over time, this can break the cycle of anxiety and help students develop not only confidence but also resilience when encountering unfamiliar quantitative challenges. They become more willing to attempt problems, engage in discussions and persist through difficulty. Confidence grows not from removing challenge, but from successfully navigating it.

This shift has important implications beyond the classroom. Students report experiencing positive emotions about statistics, including curiosity, interest, intrigue and enjoyment. Statistics and data literacy are essential skills in modern society, shaping everything from policy decisions to everyday choices. Helping students engage confidently with these skills is therefore not only an educational priority but a societal one.

While individual lecturers can make a meaningful difference, mathematics anxiety cannot be addressed through classroom practice alone. It is a systemic issue that requires attention at institutional and policy levels. Universities need to move beyond viewing quantitative support as remedial provision and instead embed inclusive approaches within curriculum design, assessment frameworks and staff development. At the same time, sector bodies and policymakers must recognise mathematics anxiety as a barrier to student participation, progression and graduate outcomes in an increasingly data-driven world.

Building a more inclusive quantitative future

At the start of a new academic year, there is an opportunity to set a different tone for statistics education. Rather than accepting anxiety as inevitable, we can design learning environments that actively moderate it.

Humanising statistics is about recognising students as whole individuals whose emotional experiences shape learning. When educators respond to students’ emotions, learning becomes inclusive, effective and empowering.

Fear of statistics is not fixed. With the right approaches, it can be transformed into engagement, confidence, resilience and even enjoyment. The goal is not to remove challenge from quantitative learning, but to help students develop the confidence and resilience needed to engage with challenge constructively.

Statistics is not inherently intimidating. Yet for many students it becomes associated with fear long before they enter a university lecture room. If we want to build a more quantitatively capable and inclusive society, we must pay as much attention to students' emotions as we do to the curriculum. The challenge is not simply to teach statistics better, but to transform how it is experienced.

Subscribe to LSE Research for the World

Interested in LSE research? Sign up to receive our newsletter: a bi-monthly digest of the latest social science research articles, podcasts and videos from LSE.

Subscribe