Jambo!

My name is Juma. I have a Master’s degree in Educational Technology and a Bachelor’s degree in Education Science, both focused in STEM education.

Currently I am a doctoral researcher in Mathematics Education at the University of Trieste (UniTS), Italy, since November 2023. I am supervised by Prof. Danilo Lewanski (University of Trieste) and Prof. Chris Sangwin (University of Edinburgh).

My research explores the intersection of mathematics education, educational technology, and data science, with a focus on computer-aided assessment systems such as STACK.

From January 2022 to December 2023, I worked at IDEMS International (UK), supporting universities in Kenya integrate online learning systems (STACK) in in STEM education.

From July 2020 to December 2021, during the COVID19 pandemic, I volunteered at the eCampus Maseno University (Kenya). Part of my tasks was supporting with instruction design, developing online learning resources, running workshops and trainings sessions on EduTech integration, and supporting online learning projects.

Before that, I taught Mathematics and Chemistry in several Kenyan schools, for 5 years, that is between 2016 and 2020.

Rethinking Assessment

Technology is changing how we teach and assess mathematics and STEM in general particularly in low-resource settings.

Even though e-assessment systems cannot grade every type of mathematical reasoning the same way a teacher would, they still offer other possibilities, such as interactive tasks, instant personalised feedback, and new ways to help students understand and build mathematical reasoning with little effort.

To illustrate, consider the following mathematical argument, which is a standard example illustrating that an irrational number raised to an irrational power can be rational.

A student might write something like {see the screenshots}:

At first glance, both versions appear similar. However, a closer inspection shows that the sub-arguments occur in different orders, certain steps are implicitly justified in one version but explicit in another, and the logical structure of the proof—case distinction, assumption handling, and conclusion—varies subtly across the two.

A teacher grading a large students would need to:

Look through each proof line by line to determine whether the logical dependencies are correct.
Check whether the student has actually established both cases, or has conflated them.
Verify whether the reasoning is valid even if the structure is unconventional.
Provide written feedback explaining which parts are sound, which are incomplete, and why.

This process can take several minutes per script—even for experienced markers—because mathematical writing is inherently open-ended. Students rarely express proofs in a single standard form; instead, they produce dozens of structurally distinct but potentially correct variants. As a result, manual grading at scale becomes slow, exhausting, and in somecases with lots of errors that go unnoticed.

Assume I make errors (to show the kind of feedback one is likely to get).

In summary electronic systems like STACK allow instructors to encode the mathematical structure of a task, like we see here the logical dependencies in a two-case proof—and automatically recognise many equivalent or reordered versions of correct reasoning. They also provide real-time feedback that helps students revise their arguments long before a teacher sees their work.

In other words, the aim is not to replace human judgement, but to rethink the workflow of assessment so that technology handles repetitive syntactic checking while teachers concentrate on the deeper pedagogy.

High School Highlights

University STEM Education Highlights

Highlights of various events for Undergraduate Mathematics Education