A
Postdoc
Institute of Design and Visual Interaction
Isabela Zaine
As a psychologist rooted in Behaviour Analysis, I study how to make learning efficient, inclusive, and respectful of individuality. I shape teaching and technology around the learner — not the other way around. I apply this to reading, education and healthcare technology, working with multidisciplinary teams of psychologists, educators, health-care providers, computer scientists and designers.
Isabela Zaine
ScreenReads
Why do some children struggle to learn to read — and what can we do about it? At the Royal Danish Academy's Centre for Visibility Design, I am part of ScreenReads, a project developing personalised digital support for early and struggling readers. We combine eye-tracking (to understand how children actually look at text), motivational design, and computational models of how learning happens. Our work has two aims: to build tools that give teachers complementary insight into how their pupils are reading, and to empirically test how changing the visual features of text affects early readers' performance. The project partners with Danish higher education institutions, schools and private companies, and is funded by Innovation Fund Denmark.
These two aims come together in the two studies I currently lead. The first asks whether a machine can see reading difficulty the way a teacher hears it. The second asks whether changing the shape of a single letter can make it easier to recognise. Both start from the same premise: that small, well-targeted changes — in how we measure reading, or in how we draw letters — can make a real difference for the children who find reading hardest.
Can Eye-Tracking Spot Reading Struggles as Well as a Teacher Can?
When a child reads aloud, an experienced teacher can often hear exactly where they stumble — a word sounded out slowly, a line where they lose their place. But teacher assessment takes time and training, and it doesn't scale easily. Could tracking where a child's eyes linger on the page offer a faster, automatic window into the same struggles?
This study puts that question to the test. We compared an automatic eye-tracking measure — called "focus-words" — against expert human judgment of oral reading, across two groups of Danish schoolchildren in Grades 2 and 4. Rather than only checking whether the two methods agree on average, we looked closely at which kinds of reading difficulty each one catches, and which it misses.
Preliminary results suggest the two approaches are complementary rather than interchangeable. Eye-tracking reliably picks up moments of sustained decoding effort — where a child is genuinely working to sound out a word — but it can miss quick stumbles that a child catches and corrects on their own. Gaze-based tools may therefore give teachers a useful early signal, without replacing their trained judgment of how a child reads.
Could a Small Letterform Change Help People Tell "b," "d," "p," and "q" Apart?
Four of the most commonly confused letters in early reading — b, d, p, and q — are, geometrically speaking, the same shape: a vertical line and a circle, flipped or rotated. This confusion is common in children learning to read, and it can persist far longer in children with dyslexia. Type designers have long tried to solve it with "dyslexia-friendly" fonts that redesign the whole alphabet — but when researchers have tested these fonts carefully, the results have been largely disappointing, showing little real benefit for reading speed or accuracy.
One reason may be that these fonts change many things about the letters at once, making it impossible to tell which change, if any, is doing the work. This project takes a more focused approach: instead of redesigning the whole alphabet, we add a single small tail to just two letters — "d" and "q" — and test, letter by letter, whether that one change measurably improves how quickly and accurately readers recognise them.
We are running this test with typically developing children and adults, and with children and adults with dyslexia. The goal is to find out whether a small, well-targeted design change can meaningfully reduce a confusion that has frustrated readers and educators for generations.