Developmental dyslexia is a common and serious specific learning disability affecting up to one in eight children in Hong Kong, with profound consequences for academic success and personal well-being. For children learning to read the visually complex Chinese script, the challenge is particularly intense. A key research challenge is a long-standing "chicken-and-egg" debate: are weaknesses in the brain's fast-processing visual system (Magnocellular-Dorsal deficits) a cause of reading difficulties, or merely a consequence of limited reading practice? This uncertainty has constrained theoretical progress, creating a major barrier to developing effective, evidence-based support for struggling readers beyond traditional linguistic-based methods.

This project is designed to resolve this debate by directly testing if training a core visual-attentional skill can causally improve reading. The approach is built on investigator's strong pilot data, which established a novel "Neurocognitive Cascade Model": a framework suggesting that the integrity of foundational visual processing impacts reading through a sequential chain reaction flowing from visual attention to working memory.

To test the model, investigators will employ a rigorous, assessor-blind randomized controlled trial with 100 Cantonese-speaking children with developmental dyslexia. Participants will be randomly assigned to either an intervention group that plays action video games, scientifically proven to enhance visual attention, or an active control group that plays a different type of non-action, engaging game. This design allows investigators to isolate the specific impact of the visual-attentional training. The investigators will track changes at four time points-before, during, immediately after the intervention, and at a 6-month follow-up-using a multi-level assessment battery, including objective neural signals of visual processing, cognitive skills, and real-world reading performance measured with advanced eye-tracking technology.

This research will deliver new knowledge that is both scientifically and socially impactful. It will provide the first causal evidence for a visual-attentional pathway to reading remediation in Chinese developmental dyslexia, fundamentally shifting theoretical understanding of the disorder from a single, phonological deficit to a more comprehensive, multi-component model. The findings will have a significant translational impact, paving the way for innovative, non-reading-based interventions and new early screening tools. Ultimately, this project will offer a new, evidence-based approach to help children overcome the specific neurocognitive bottlenecks that hinder their ability to learn to read, with the potential to improve educational outcomes for a large and underserved population.

Inclusion Criteria:

• Formal diagnosis of dyslexia by an educational or clinical psychologist using the Hong Kong Test of Specific Learning Difficulties in Reading and Writing (HKT-SpLD),39 with a literacy composite score at or below the 10th percentile.
• Non-verbal IQ score ≥ 85 on Raven's Standard Progressive Matrices.
• Normal or corrected-to-normal vision (both eyes logMAR 0.1 or better).

Exclusion Criteria:

History of neurological or psychiatric disorders (e.g., epilepsy, psychosis).

Upon providing informed consent from both caregivers and children, eligible children will complete the comprehensive Pre-Intervention (T1) assessment battery. The participants will then be assigned to either the Intervention or the Active Control group using stratified blocked randomisation. Participants will be stratified based on baseline reading severity (high vs. low) and comorbid Attention-Deficit/Hyperactivity Disorder (ADHD) symptoms (high vs. low). Within each stratum, a computer-generated algorithm will randomise participants in blocks of six to ensure a balanced allocation across conditions. All research staff conducting the assessments will be blinded to the group allocation.