A Statistically Significant Reduction in Math Anxiety (p-value < 0.05)

Categories: Uncategorized

By Kirsten Landsiedel, Biostatistics

Teaching Effectiveness Award Essay, 2024

If you have ever taken a math class before, you likely have a readily available answer to the question: “Do you consider yourself a math person”? Perhaps someone answered this question for you, you labeled yourself one way or another, or your identity and appearance narrowed your choices down to one. Longitudinal Data Analysis is a graduate-level course in statistical methods for the analysis of repeated measures data structures. As a theory-based course often geared towards applied researchers, I have noticed that many students come into the course with a lot of “math anxiety” (Google it!). So what can we, as teachers, do to help dismantle the decades-old barriers that hinder students’ learning? 

The answer is complex. Our self-perception in mathematical spaces is deeply ingrained in institutional, social, and political challenges that can’t be resolved overnight. We can, however, work to create a learning environment that feels more accessible, inviting, and inclusive. 

I used to think that a great teacher needed to be an expert in their field, someone who had all the answers, all the time. Now, I believe that relatability and vulnerability are the true trademarks of a great teacher. I try to be honest with my students about my own journey and struggles through math and stats. Even as a biostatistics PhD student, it takes me a long time to decipher mathematical notation and learn new concepts. I get frustrated, I struggle, I fail, and I try again. I encourage my students to voice their struggles and confusions. I remind my students, and myself, that there is no such thing as a “math person” or “non-math person.” Learning math is much more like learning to play soccer. If you show up to practice consistently, you will get better. 

From a pedagogical perspective, I implemented the following techniques to help ease math anxiety: intuition, images, and implementation. (1) Mathematical (and statistical) intuition is the most valuable and most overlooked part of the learning process. Instead of introducing expectations as integrals, why don’t we begin by introducing an expectation as the “balancing point” of a distribution and tell our students to picture a distribution perched on top of a seesaw – how could you position the distribution such that the seesaw is perfectly level? The pivot points to the expectation. Words can, ironically, make all the difference when teaching math. (2) Images: If you start with formulae, you’ll lose your audience. Start with drawing a picture, and the abstract becomes tangible. (3) Implementation: The book “Building Thinking Classrooms” suggests that group-based problem-solving activities facilitate much higher rates of “active learning” than traditional lectures. In response to this, I transformed the lecture-style labs into collaborative group coding activities. I try to create groups that are small enough such that all voices are heard but large enough to capture a diverse range of student talent and experience. As un-intuitive as it might sound, students learn more when their teacher talks less. I believed that all these tactics would work in theory. But like any good statistician, I needed data to confirm. After two anonymous surveys and many more conversations with my students, I can truly say that there is a statistically significant difference in student comprehension when these methods are applied (p-value < 0.05).