An Inclusive and Gamified Design Framework for Educational Serious Games Targeting Children with ASD: A Longitudinal Exploratory Study of Behavioral Patterns

Learning to read is a fundamental human skill, one that unlocks social connection and economic opportunity, yet it does not happen naturally. Unlike speaking, which emerges from our biology, reading requires the brain to rewire itself, repurposing visual circuits originally designed for recognizing objects into a system for decoding symbols. For children with Autism Spectrum Disorder, a condition defined by differences in social communication and repetitive behaviors, this process is often even more complex. They may struggle with the attention and memory required to learn, or find the sensory experience of a classroom overwhelming. In recent years, educators and technologists have looked to video games as a potential bridge. The hope is that the structured, predictable, and rewarding nature of games could make learning to read more accessible. However, creating a game that truly works for this specific group is difficult; it requires balancing a solid teaching method with design that respects how these children think and feel, all while keeping them motivated without causing anxiety.

A team of researchers in Brazil set out to build a solution that brings these three needs together. They developed a new design framework for educational games, one that combines a specific method for teaching literacy, a set of rules for making interfaces easy to understand, and strategies to keep children engaged based on their natural desire to explore. To test if this approach worked, they created a mobile game and ran a long-term study with six children, three of whom used the game while three others learned using traditional, non-digital methods. The researchers did not look for a magic cure or a guaranteed improvement in reading scores. Instead, they watched closely how the children interacted with the material over hundreds of sessions, measuring how quickly they responded and how steady their behavior remained.

The framework they built rests on three pillars. The first is a teaching method called PGL, which treats reading as a connection between pictures, sounds, and written words, avoiding complex grammar rules until the basics are mastered. The second pillar draws from guidelines designed for people with cognitive differences, ensuring the game is clear, predictable, and free from confusing distractions. The third pillar uses a model of human motivation that focuses on the "Free Spirit" type of player—someone who learns best when they have control, can explore at their own pace, and are not forced into direct competition. By weaving these three elements together, the researchers created a digital environment where a child could tap a screen to pop a balloon, revealing a syllable, and receive immediate, gentle feedback.

To see if this digital approach held up, the researchers conducted a study at a specialized clinic in Recife. They recruited six children between the ages of six and ten, all diagnosed with autism. Three children used the new mobile game, while the other three worked through the same lessons with a human instructor using traditional paper and pen. The study was not a quick test; it was a deep dive into behavior over time. The children in the game group completed nearly 1,300 sessions, while the traditional group completed over 2,000, with the total time spent ranging from a few hours to over twenty hours per child. The researchers focused their analysis on a single, simple activity within the game—a tapping task where the child had to pop balloons to match sounds—because it was the only part of the curriculum where they had complete, detailed records for every single participant.

The results offered a glimpse into how the children behaved, though the researchers were careful to frame them as preliminary signs rather than final proof. The children using the game showed a distinct pattern of interaction. They responded to the tasks faster, with an average reaction time of about 0.95 seconds, compared to the 1.51 seconds observed in the traditional group. More importantly, their behavior was more consistent. The children in the game group stayed in a "stable" state for the vast majority of their sessions, rarely showing signs of restlessness or distraction. In contrast, the children in the traditional group spent most of their time in a "moderate" state, with some showing periods of instability. The difference in reaction times was so complete that every single child in the game group was faster than every single child in the traditional group, a pattern the researchers described as a strong separation between the two groups.

However, the study also highlighted significant reasons to be cautious about these numbers. The way time was measured was different for the two groups. The game recorded every tap automatically with millisecond precision, while the traditional group's times were written down by hand by the instructor. This difference in measurement tools makes a direct, head-to-head comparison of the exact seconds risky, as the human recorder might have missed a fraction of a second or reacted slightly slower than the machine. Furthermore, the study involved only six children, a number too small to prove that the game caused the improvement. The children also had different medical profiles; some were taking medication that can affect reaction speed, and the researchers noted that these factors could have influenced the results just as much as the game itself.

Despite these limitations, the study demonstrated that the framework itself is viable. The game successfully translated a complex teaching method into a digital format without losing its core structure. The children were able to engage with the material for long periods, and the system captured detailed data on their behavior, proving that such a tool can be built and used in a real-world setting. The researchers concluded that while the game did not prove it could teach reading better than a human teacher, it did show that a child with autism could interact with a structured digital lesson with greater fluency and stability than in a traditional setting. The patterns observed suggest that when the design respects the child's need for predictability and control, the digital environment can become a place where they feel comfortable and focused.

This work does not claim to have solved the challenge of teaching reading to children with autism. It does not offer a guarantee that the game will improve reading skills, nor does it suggest that the game is a replacement for human instruction. Instead, it offers a blueprint for how to build such tools. By combining a proven teaching method with careful design for cognitive accessibility and a focus on intrinsic motivation, the researchers created a system that works in practice. The study serves as a proof of concept, showing that it is possible to create a digital learning environment that is not just accessible, but also engaging and stable for children who often find the world chaotic. The true value lies in the framework itself, a set of rules and principles that can guide future developers in creating games that truly meet the needs of this diverse population.