Design and Evaluation of a Serious Game Prototype to Stimulate Pre-Reading Fluency Processes in Paediatric Hospital Classrooms

Abstract

Didactic digital tools can commence, enhance, and strengthen reading fluency in children undergoing long-term hospitalization due to oncology conditions. However, resources specifically designed to support rapid naming and decoding in Spanish remain scarce. This study presents the design, development, and evaluation of a game prototype aimed at addressing this gap among Spanish-speaking preschoolers in hospital settings. Developed using Unity through a design-based research methodology, the game comprises three narratively linked levels targeting rapid naming, decoding, and fluency. A sequential exploratory mixed-methods design (QUAL-quan) guided the evaluation. Qualitative data were obtained from a focus group of hospital teachers (N = 6) and interviews with experts (N = 30) in relevant fields. Quantitative validation involved 274 experts assessing the game’s contextual, pedagogical, and technical quality. The prototype was also piloted with four end-users using standardised tests for rapid naming, decoding, and fluency in Spanish. Results indicated strong expert consensus regarding the game’s educational value, contextual fit, and usability. Preliminary findings suggest potential for fostering and supplementing early literacy skills in hospitalised children. Further research with larger clinical samples is recommended to validate these outcomes.

1. Introduction

In the complex landscape of paediatric healthcare, educational interventions face distinctive challenges. This is particularly prominent within oncology wards, where maintaining children’s engagement and learning continuity is critical yet frequently disrupted. Game-based learning and gamification have emerged as powerful tools for fostering motivation and engagement, with a growing body of research demonstrating their efficacy across various literacy-related processes, educational stages, and inclusive learning environments [1,2,3,4]. One of the most notable examples of this are serious games. These digital games, distinct from gamification and general game-based learning approaches, are designed with educational, therapeutic, or training purposes in mind. Their most prominent characteristic is, integrating learning objectives within interactive and engaging game mechanics to promote active participation and skill development in specific contexts. Despite their potential benefits, their application in non-traditional settings, such as hospital classrooms, remains notably underexplored, especially in the context of paediatric oncology.

Recent studies have shown that neurological disruptions—such as those caused by developmental disorders, brain injuries, or central nervous system (CNS) conditions—can affect brain regions that are essential for developing reading fluency [5]. These disruptions are particularly prominent in children diagnosed with CNS-related conditions, where any changes to the tissue, including tumours, often impair neural areas responsible for critical reading components such as speed, accuracy, and prosody [6,7,8]. Consequently, this overlap between affected brain regions and those involved in reading fluency may result in long-term neurocognitive consequences that hinder both short- and long-term literacy outcomes, as seen in Figure 1.

These impairments pose a dual challenge—both clinical and educational—by disrupting executive functions that are essential for literacy development, such as processing speed, working memory, and sustained attention. The children’s learning trajectories are only further disrupted with additional barriers such as intensive medical treatments, hospital-induced stress, and the emotional toll of their illness. In this context, hospital-based classrooms must provide instruction that is not only flexible and individualised but also emotionally responsive to the needs of this vulnerable paediatric population.

Consequently, the intersection of paediatric oncology, serious games, and reading fluency represents a vital yet understudied frontier in educational technology. This notion is supported by a review of existing literature, which highlights a significant gap in research specifically examining the use of video games to improve reading fluency in hospital classroom settings—particularly among children receiving treatment for CNS conditions [9].

2. Background and Related Works

2.1. Pedagogy and Characteristics of Paediatric Oncology

Pedagogical practice in paediatric oncology is grounded in addressing the unique educational and emotional needs of children with cancer within hospital classrooms. This requires a highly adaptable approach that accommodates the varying stages of treatment, fluctuating physical health, and evolving emotional well-being of each child. Given the prolonged and often intensive nature of cancer therapies—frequently extending over months or years and involving intricate medical protocols—educational interventions must be flexible, responsive, and deeply personalised to support meaningful learning throughout the treatment journey.

Empathy plays a central role in this pedagogical model. The psychological strain of illness, combined with the disruption of normal routines, necessitates a learning environment that is not only academically supportive but also emotionally nurturing. Therefore, cultivating a warm, welcoming atmosphere is essential to promote both cognitive development and overall well-being, benefiting children and their families alike.

Within this framework, the ‘hospital classroom’ becomes a critical educational space [10]. These classrooms are designed to allow hospitalised or outpatient children to continue their academic journey and participate in recreational activities during treatment. Their key goal is to prevent educational marginalisation while facilitating smooth reintegration into mainstream schooling following recovery.

Paediatric oncology areas within these classrooms are typically located near or within the oncology unit and are designed to meet the specific needs of young patients. Attention to lighting, colour schemes, and child-friendly furnishings contributes to a calm, engaging, and emotionally supportive environment. Moreover, such areas are also often equipped with specialised educational technologies, such as computers, tablets, and interactive tools, alongside resources for emotional and psychological support. Collectively, these features work to sustain academic progress and promote positive learning experiences during a profoundly challenging period in a child’s life.

2.2. Enhancing Reading Fluency Through Video Games

Reading fluency is a foundational component of early literacy. Among its key predictors is Rapid Automatized Naming (RAN)—the ability to name familiar items quickly and accurately—which plays a crucial role in the development of fluent reading and overall literacy success [11]. As specialised reading circuits in the brain become automated, cognitive resources are freed for higher-level processes such as reading comprehension [12,13]. Building on this, numerous studies emphasise that reading fluency not only facilitates accurate word recognition but also serves as a strong predictor of future reading achievement [14,15,16,17,18,19,20,21,22,23,24]. In particular, RAN has been identified as one of the most reliable indicators of subsequent reading fluency development [25,26,27,28].

Digital games have gained recognition as promising tools for supporting literacy by offering highly interactive, engaging environments that can motivate and challenge learners. However, despite their growing application in education, only a few are specifically designed to enhance reading acquisition. In particular, a review of 18 studies found only one that directly addressed reading acquisition through gameplay, underscoring the lack of attention this domain has received in serious game research [29,30].

One notable example of this is The Adventures of Amaru, which integrates word decoding and vocabulary development through narrative gameplay, character engagement, and a balanced reward system following the completion of mini-games [31]. Additional examples reflect a growing awareness of games’ educational potential, although the field remains relatively unexplored. For instance, Tradislexia aims to enhance reading speed and accuracy in children with dyslexia [32], while a separate preschool-focused game has shown greater improvements in early reading outcomes compared to traditional textbook-based learning [30]. Moreover, and in conflict-affected regions, significantly enhanced decoding and fluency in Arabic among Syrian refugee children aged 5 to 10 has been noted following engagement with Antura and the Letters [33].

In the Netherlands, improved reading fluency and motivation in Dutch-speaking primary students was noted following student’s interaction with Letterprins and Reading Race [34,35]. Likewise, the employment of ELAN in France demonstrated measurable gains in phonological awareness and grapheme–phoneme correspondence [36]. Positive outcomes were also reported following the use of other notable tools such as a therapeutic video game designed for children with special learning difficulties [37]. Similar results were reported following the use of the ABC Adventure virtual reality game, developed for early childhood reading in Early Childhood Education, leading to significant improvements in reading skills [38]. A significant improvement in core literacy skills was also observed after just three months of using Escribo Play, a Portuguese app featuring 20 mini-games, including a 68% increase in reading proficiency and a 48% increase in writing skills [39].

Comparable initiatives have also emerged in other regions—for example, in Mexico, a game was developed specifically to enhance reading comprehension among third-grade students [40]. Additionally, Saving the Word was designed to enhance reading abilities in primary school learners [41], and Isla Secreta (Secret Island) was created to support reading and writing development in first and second grade students [42]. In Spain, Galexia targeted fluency in children aged eight and older with dyslexia [43], PetitUBinding intervention program reported fluency gains in early readers [44], and Leobien supported first-grade reading comprehension, prioritising progress over initial outcomes [45].

Despite some advancements in the field, the current literature reveals a notable scarcity of serious games specifically designed to support reading enhancement, particularly in initiating and developing reading fluency. Within the context of Early Childhood Education (ECE), the focus of the present study, only a handful of recent developments have targeted early reading acquisition [30,38,39]. Among these, just one game has been explicitly designed to improve reading fluency, and it is intended for children aged eight and above [43]. Moreover, it is also worth noting that only two games, one digital and one virtual reality-based, have been developed with the specific aim of fostering reading in five-year-olds [30,38]. In particular, Escribo Play, while effective in supporting general reading acquisition through a series of mini-games, was designed for Portuguese-speaking learners in mainstream classroom settings and does not specifically target fluency [39]. As for Spanish-language tools, Galexia is the only game focused exclusively on reading fluency, but it is intended for children from eight years old onwards [43].

In addition to the scarcity of serious games specifically designed to target distinct subsets of literacy skills, a recurring critique in the field highlights the simplistic design of many educational games, especially 2D point-and-click formats, which often depend on basic response mechanics and lack opportunities for meaningful exploration or cohesive storytelling. These games, often described by the metaphor “broccoli with chocolate,” fail to integrate gameplay with educational content in a way that fosters authentic engagement [46]. Notably, when game mechanics fail to align with learning objectives, both the instructional effectiveness and motivational appeal of these games are significantly reduced. As a result, there is a noticeable lack of serious games targeting reading fluency in ECE. This gap not only highlights the need for more specialized game development but also reveals a promising and underexplored domain—one with significant potential for academic inquiry and innovation.

2.3. Serious Games in Paediatric Healthcare Education

The integration of ICT resources and video games in paediatric healthcare education, particularly within oncology-focused hospital classrooms, remains notably under-researched. Despite growing interest in game-based learning, few studies have examined the potential of digital games to support early reading fluency in clinical educational settings [5]. This gap is especially pronounced in relation to young children with oncological conditions, where tailored game-based interventions for initiating reading instruction are virtually absent.

Nevertheless, some research has explored the broader educational and emotional potential of video games in paediatric hospital contexts [47,48,49]. This includes projects such as SAVEH, VIDEM, and SALUD-in, which aim to develop digital games that promote social interaction, collaboration, and cognitive stimulation among hospitalised children [48]. Specifically, examples like Mundo Isla and Tango.H exemplify this approach, focusing on activities such as cooperative learning, physical engagement, and cognitive exercises through interactive platforms [47].

The emotional and psychological benefits of immersive technologies have also been highlighted in studies on virtual reality, which has been shown to reduce distress and anxiety among hospitalised children and adolescents [50]. Similarly, platforms like Minecraft have been employed to facilitate role-playing and creativity, enhancing emotional well-being through playful, open-ended learning experiences [51]. Another emerging strategy involves involving children in co-creating their own video games, a practice shown to increase motivation, autonomy, and satisfaction [52].

While the use of video games as educational tools in hospital classrooms is not extensively explored, existing research provides a beneficial foundation for developing investigations that combine the didactic potential of video games with their educational value to develop reading skills [53], particularly for initiating reading and improving fluency and in such unique learning environments.

2.4. Serious Games Prototypes Evaluation

Designing a serious game is a multifaceted process that requires a careful integration of pedagogical goals, technological capabilities, and game design principles. Throughout this process, a clear educational objective is essential to effectively guide the design strategy from the onset [54,55,56]. Moreover, despite the rising prominence of serious games in education, many still lack a solid theoretical foundation—an oversight made more apparent by literature reviews that point out the importance of engaging narratives, intuitive controls, and purposeful gameplay in sustaining user engagement and reinforcing instructional content [57]. Yet, numerous games incorporate overly complex exploratory elements that risk diluting learning outcomes [58]. In contrast, clearly defined objectives and accessible interfaces have been shown to enhance user attention and educational focus [59].

As a key driver of learner-centred design, prototyping is one of the most crucial elements in game development. It enables the creation of preliminary versions that allow developers to refine gameplay and instructional features through demonstration and usability testing—ultimately enhancing the game’s value for the end user through iterative improvement [60,61]. Similarly, prototyping serves several purposes in educational game design, from it facilitates early evaluation of core game elements, provides insights into user interaction, and supports the assessment of key aspects such as usability, engagement, and content alignment.

Prototypes range from low-fidelity models (e.g., paper sketches) used for conceptual validation to high-fidelity versions that closely resemble the final product. While low-fidelity prototypes allow for rapid iteration, high-fidelity prototypes offer detailed, realistic gameplay experiences that enable nuanced evaluations of functionality, aesthetics, and pedagogical integration. However, developing such advanced prototypes is time- and resource-intensive, a challenge reflected in the present study, which employed a fully functional, forward-designed high-fidelity prototype. Prior research using tools like the Fun Toolkit suggests that prototype fidelity strongly influences user perception, particularly among school-aged users [62].

Numerous studies and review articles have proposed diverse evaluation heuristics and criteria for assessing educational games [63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84]. Many studies [85,86,87,88,89,90,91,92,93,94,95] and review studies [57,96,97,98,99,100,101,102,103,104] have reported the use of a wide diversity of evaluation aspects focusing on educational games. Despite this, there is no universally accepted framework or scale for evaluating serious game design contexts [78,99,101,103,104], particularly within hospital classrooms. This lack of standardisation reflects the complexity and diversity of serious game evaluation. Moreover, general usability criteria are often insufficient in educational settings, prompting researchers to advocate for pedagogical usability considerations that address learning goals and learner characteristics [105,106].

Some studies have even proposed tailored design heuristics for digital games intended for children with language delays, further illustrating the need for population-specific evaluation approaches [107]. Yet, most assessments of educational software continue to rely either on conventional engineering metrics or on purely pedagogical indicators, an analytical limitation noted across multiple comparative studies [108,109].

Given the absence of validated scales for assessing serious games or high-fidelity prototypes in hospital classroom environments, we developed a custom evaluation tool for this study: the Questionnaire for the Quality Evaluation of the Serious Game Prototype Design within Hospital Schools. This instrument was derived and adapted from existing validated frameworks [76,78,80,81,86,87,88,96,98,105,110,111,112], which were selected and mapped based on their relevance to serious game design evaluation and research information in the qualitative phase, all in line with the principles of exploratory sequential mixed methods designs. Therefore, our approach aligns with a common practice in the field, constructing ad hoc tools informed by established heuristics and data, particularly in contexts where no validated instruments exist for the target environment.

This study presents a novel contribution by designing and evaluating a serious game prototype aimed at supporting rapid naming, decoding, and early reading fluency in paediatric oncology hospital classrooms. The primary research objective was to capture and assess expert perceptions of the prototype’s design quality, with particular emphasis on its suitability for preschool-aged children undergoing cancer treatment. To address this aim, the following sections outline the methodological framework and instructional design (Section 3), present and integrate findings from both qualitative and quantitative phases (Section 4), and discuss the results, limitations, and broader implications (Section 5).

3. Materials and Methods

3.1. Instructional Serious Game Design

In accordance with the pedagogical framework outlined for game-based learning, the game design integrates playful and academic activities essential for fostering interactive and meaningful learning in formal education settings [113]. This approach aims to immerse students in an engaging educational experience that merges enjoyment with essential literacy outcomes. Central to the educational framework is the development of phonological awareness, which numerous studies identify as foundational to literacy acquisition [114,115,116,117,118,119,120,121,122,123]. This skill is supported through the use of technology and game-based methods specifically tailored for younger learners [124,125,126,127].

Additionally, the game’s design is informed by vocabulary research, particularly the strategy of repeated exposure to target words. This approach—integrated into gameplay mechanics—encourages multiple interactions with selected vocabulary items. Such repetition has been shown to enhance reading acquisition by fostering deeper language engagement and improving comprehension [128,129,130,131,132,133]. Moreover, repeated reading practices, implemented within immersive digital environments, also play a crucial role in enhancing reading fluency and accuracy [16,134,135,136,137,138,139,140,141,142]. These practices are only further enriched by incorporating technological learning environments in early childhood education and the use of serious games [143,144,145,146].

Building on this foundation, the vocabulary enrichment strategies employed in the game align closely with contemporary recommendations in early childhood education. These guidelines advocate for focused instruction on 3–5 high-utility words per week, coupled with explicit teaching of word meanings. Moreover, the game’s activities are closely designed to promote contextualised word usage, a practice shown to enhance both vocabulary retention and the ability to apply new words effectively [147]. On top of this, the game’s pedagogical structure also includes phonological and syllabic awareness. More specifically, it follows an integrated method that includes multiple phonological and syllabic components [148,149], complemented by a developmental sequence that progresses from larger to smaller phonological units in line with the natural evolution of phonological skills [150].

The game’s instructional model is grounded in an eclectic approach, drawing on research that supports the effectiveness of holistic and mixed methodologies in enhancing reading outcomes [42,151,152,153,154]. In this context, eclectic refers to the intentional integration of techniques from various instructional models—including rapid naming, syllabic decoding, and visual word recognition—tailored to address the neurocognitive needs of the target paediatric population.

Central to the game’s pedagogical delivery are guided instruction and explicit teaching, both of which are rooted in well-established educational practices. These methods emphasize breaking down tasks into manageable steps, modelling desired skills, providing scaffolded support, delivering timely and constructive feedback, and ensuring repeated opportunities for practice [155,156,157,158,159,160]. Among these, explicit teaching has shown particular efficacy in literacy interventions and remains especially critical in early reading development [18,161].

Aside from explicit teaching, the game design also incorporates principles of active recall and cognitive load management. Drawing from retrieval-based learning, and the concept of desirable difficulties, the game supports memory retention and transfer [162,163,164]. It also applies cognitive load theory [165], by ensuring clear user guidance and minimising extraneous load—practices supported by recent research on serious games in literacy and educational training [166,167].

Multimedia learning principles are embedded throughout the game to enhance comprehension and retention [168]. The design also integrates spaced practice, as recommended by cognitive psychology literature, to promote long-term learning [169,170,171]. Finally, scaffolding strategies, grounded in research on cognitive development and educational games, facilitate knowledge transfer across contexts and support skill consolidation and retention [172,173,174,175,176].

3.2. Game Structure: Characterisation of the Video Game Yuki’s Adventures: Hidden Words

Name: Yuki’s Adventures: Hidden Words.

Knowledge Area: Reading literacy: Decoding and fluency.

Topic: Enhancing decoding and initial reading fluency in Spanish through interactive gaming among paediatric oncology patients.

Taxonomy: Prototype of a serious/educational game. Edu Game or Edutainment [177]. It is a linear game, focused on basic learning and the development of early skills through mini-games and elements of action, adventure, card challenges, strategy challenges, puzzle games, sports games, and simulation games [178,179,180]. Moreover, it also features balanced exploration, helping prevent overstimulation from visually striking components that could distract from the learning objectives [58].

Methodology used for the design: Integrative methodological model [181].

Purpose: To facilitate the development of foundational reading skills in young oncology patients, leveraging the motivational aspects through a serious game adapted specifically to their needs.

Target audience: Preschool children aged 5–6 from oncology areas with CNS conditions.

Objectives: Initiate and improve word recognition, rapid naming, decoding, early fluency and prosody as a predictor of reading success when starting Primary Education.

Environment: Interactive adventures in virtual scenarios designed for children’s engagement.

Game rules: The player begins in a rural village, taking on the role of a character who has forgotten all the knowledge related to reading. The journey unfolds as a solo adventure starting in World 1. Upon awakening in a village house, the character interacts with surrounding objects and environmental elements, each representing high-frequency words. Recognising these elements enables the character to “remember” them, earning items that can be used in subsequent settings or sold in an in-game shop.

Following this initial stage, the character enters a nearby dungeon, where they must complete a series of reading-based challenges (Figure 2). Successfully navigating the dungeon opens a door leading to a forest. Before progressing further, the character encounters the game’s antagonist, Zote, a cosy ghost-like figure who steals reading memories from others.

The forest, known as Bosque Luminaria (Figure 3), represents the second stage of the learning journey. Here, the character interacts with forest elements to develop skills such as recognising capital letters, decoding syllables, and reading full words. This world includes a variety of challenges: engaging in dialogue with two characters, solving a witch’s riddle, locating a hidden treasure chest, fishing, using virtual lenses to uncover concealed elements, and navigating a labyrinth. Upon completing these tasks, the character reaches a hut containing a computer that presents the final challenge of this world. Completing it unlocks the door to World 3.

World 3 (Figure 4) is a vibrant, hyper-casual city where the character attends school each morning and receives guidance from a virtual teacher. In the afternoon, the player visits the club library, where instruction becomes more explicit, covering letters, syllables, corresponding vocabulary, and complete sentences. The character also explores various city environments, including a sports centre, a beach, and a theme park, completing context-based tasks tied to literacy goals.

All challenges are designed to foster engagement and immersion, promoting word recognition and reading within a meaningful and motivating narrative. Mini-games are embedded throughout the gameplay and target various literacy levels, including whole-word, syllable, letter, and phoneme-based tasks.

Following core instructional segments, the game offers a series of playful wrap-up activities (Figure 5). While these are primarily recreational rather than pedagogical, they are designed to provide children with moments of enjoyment and emotional relief. These light-hearted segments serve as positive emotional pauses, a particularly important consideration in hospital settings, where supporting emotional well-being is integral to the learning experience.

Challenges: The game’s levels are designed with increasing complexity to promote progressive learning. Core instructional principles such as scaffolding, spaced repetition, and active recall are embedded throughout gameplay. Children receive real-time feedback on their advancement via a dynamic progress map (Figure 6), which visualises their journey and reinforces a sense of achievement.

Gamification mechanics are incorporated to support both intrinsic and extrinsic motivation. Upon completing each world, players earn a corresponding digital sticker, which is stored in a rewards album (Figure 7), and receive a tangible prize from a teacher or caregiver. This dual-reward system is intended to enhance engagement through visible and meaningful accomplishments—particularly in clinical contexts where sustaining attention and motivation is vital. While the motivational impact of specific reward types was not empirically tested in this pilot, informal teacher feedback indicated positive reception; future research will explore age-related responses and the effectiveness of adaptive reward mechanisms.

To support individualised instruction, the prototype features a learning evidence log that registers correct and incorrect responses for each task completed by the child (Figure 8). This interface, currently implemented at an initial level for demonstration purposes, presents structured feedback by game world, task completion status, and item-level performance, allowing hospital-based teachers to identify specific areas of difficulty and adjust instruction accordingly. While the current prototype offers a simplified display layer, the final system architecture is being designed to support advanced learning analytics through the integration of usage records encoded as the current .json data. This will be made operational via a middleware layer deployed on Heroku (https://www.heroku.com/) and linked to a MongoDB database (https://www.mongodb.com/) for the persistent storage of interaction data. This infrastructure will enable the systematic collection of fine-grained information on session duration, user behaviour, and response latency to reading-related stimuli, thereby establishing a robust foundation for the temporal analysis of emergent fluency indicators within gameplay. Visual and colour-coded elements enhance interpretability, facilitating its use in clinical-educational settings. This feature supports targeted reinforcement and fosters vocabulary consolidation tailored to each learner’s progress and instructional needs.

Furthermore, the system incorporates a spaced repetition mechanism in the menu, as illustrated in Figure 9, that retrieves previously incorrect responses and reintroduces them later through flashcard-based activities, reinforcing retention and supporting long-term acquisition.

In addition, mini-games can be revisited independently through a dedicated selection menu (Figure 10), which reinforces retrieval-based learning and spaced practice. This feature allows players to access specific reading challenges at any time, supporting memory consolidation and transfer through repeated, distributed exposure.

Interface Design and Accessibility Features: The game interface is designed to be child-friendly and intuitive, featuring visually appealing elements in a hyper-casual, cartoon-like style to sustain engagement. The gameplay is based on simple touchscreen interactions, allowing players to perform core actions such as moving in four directions, selecting objects, or dragging items during in-game tasks. Given the clinical context, accessibility was a key principle in the design process. Namely, the prototype was specifically adapted for five-year-old children with CNS tumours undergoing oncological treatment, whose neurocognitive profile may involve reduced working memory, slower processing speed, and fluctuating attention spans. To accommodate these characteristics, the game employs minimal cognitive load interfaces, large visual targets, and one-touch navigation compatible with desktop simulating touch-based interaction through mouse input in anticipation of its future deployment on tablets (see Figure 11).

Visual design features were intentionally developed to create cosy, high-contrast environments in warm tones, combined with smooth animations and child-friendly visuals. These elements were designed to minimise visual fatigue and foster a sense of emotional comfort and security for young users in hospital settings, particularly boasting bright, affectively supportive colour palettes as depicted in Figure 12.

The game’s usability is further promoted by the addition of supportive elements such as manual navigation, constant in-game narrator guidance, and in-game elements helpers (Figure 13). While advanced personalisation settings were not implemented at this prototype stage, final development aims to include additional accessibility tools, such as scalable text, colour adjustment, or audio support.

The settings menu was designed to allow users to adjust key technical parameters relevant to gameplay performance and user comfort. These include screen resolution, windowed or full-screen display, overall game quality according to hardware capacity (managed via level-of-detail optimisation), volume control, and interaction mode selection, enabling either keyboard-based navigation or point-and-click control with a mouse or touchscreen. These options were included to accommodate diverse testing environments and devices used by participants during evaluation. As shown in Figure 14, the current interface already incorporates basic accessibility-focused adjustments aimed at supporting varied user needs. Although these features were positively received by expert reviewers, the final version is intended to include additional features including scalable text, manual brightness adjustment, alternative contrast modes, and expanded visual customisation tools.

3.3. Learning and Game Mechanics Alignment

A key strength of the developed prototype lies in its deliberate harmonisation of educational objectives with game mechanics—an essential principle in serious game design. Organized into three uniquely themed worlds, the game’s prototype is carefully structured to support literacy development, guiding players from foundational word recognition toward early stages of reading fluency [53]. Moreover, each game element is intentionally designed to be both engaging and pedagogically meaningful, ensuring entertainment and instruction are seamlessly integrated throughout the gameplay experience.

Each in-game activity reflects the eclectic instructional method, with gameplay mechanics directly mapped to specific pedagogical strategies. These activities are not only interactive but are also grounded in well-defined educational purposes, an essential characteristic of serious game frameworks. This alignment ensures that learning experiences are meaningfully embedded within the gameplay, fostering a coherent environment in which educational theory and interactive design are inextricably linked. Furthermore, the storyline and learning content were purposefully integrated across the game’s three narrative environments, each corresponding to specific stages in the reading acquisition process. This coupling was grounded in pedagogical models such as Bloom’s taxonomy and the Learning Mechanics-Game Mechanics (LM-GM) framework [54], and further validated by expert feedback, which highlighted the narrative’s coherence and its capacity to support progression through literacy tasks. Detailed mapping of game levels, learning objectives, and didactic strategies is provided in Appendix A, Appendix B and Appendix C.

3.4. Evaluation and Validation Methods

This study employed an exploratory mixed-methods approach, combining qualitative and quantitative data to evaluate the serious game prototype. Central to the evaluation was the expert judgment technique, a widely recognized method in educational research for assessing technological resources, including serious games [182,183]. By leveraging expert insights, the study ensured and rigorous and informed appraisal of the prototype’s educational and design quality. A critical aspect of this method involves the selection and qualification of experts, addressed here through the Expert Competence Coefficient (CCE). According to the classification criteria:

• A coefficient between 0.8 and 1.0 indicates high competence.
• A coefficient between 0.5 and 0.8 indicates a medium level of competence.
• A coefficient below 0.5 indicates a low level of competence.

To ensure methodological rigour in the expert validation process, the Expert Competence Coefficient (CCE) was applied, calculated using the formula K = ½ (Kc + Ka). This coefficient is widely used in educational and technological evaluation studies to ensure the qualification of participants in expert judgement processes. The first component, Kc (knowledge coefficient), was derived from the expert’s self-assessed level of expertise in relation to the topic on a scale from 0 (no knowledge) to 10 (full mastery).

The second component, Ka (argumentation coefficient), was calculated based on the expert’s assessment of the influence of different sources on their knowledge and judgement. Each source was assigned a weight depending on the degree of influence indicated: theoretical analysis carried out by the expert (0.10 low, 0.20 medium, 0.30 high), practical professional experience (0.20 low, 0.40 medium, 0.50 high), and other sources, including national and international literature, awareness of developments abroad, and personal intuition, which each contributed a fixed value of 0.05. The Ka was obtained by summing the weights corresponding to the expert’s chosen levels across all sources.

Therefore, the final CCE value results from averaging Kc and Ka, resulting in a coefficient ranging from 0 to 1. In this study, only experts with a CCE equal to or higher than 0.8, indicating a high level of competence, were considered for inclusion in the validation process.

For the qualitative evaluation (QUAL), a focus group session was conducted, supplemented by individual interviews with experts who met the CCE threshold. These sessions aimed to capture in-depth insights into the design quality and educational alignment of the prototype. The focus group was composed of six teachers from paediatric oncology hospital classrooms in Andalusia, the Canary Islands, and the Region of Murcia. Participants were selected based on their extensive experience with children aged 3 to 13 diagnosed with CNS tumours. The group met online for 52 min and provided feedback on the prototype’s pedagogical basis, visual and auditory design, clarity and accessibility of content, alignment of objectives and challenges, and overall usability in hospital settings.

For the quantitative evaluation (QUAN), an ad hoc questionnaire was developed and subjected to a pilot study, which confirmed its strong content validity, clear factorial structure, and high internal consistency. To validate the instrument, a group of ten experts specializing in Educational Technology and Game Design assessed the relevance and clarity of each item using the expert judgment technique. The result, measured through Aiken’s V coefficient, showed that all items exceeded the threshold value of 0.8, demonstrating a high degree of clarity and pertinence.

To further examine the questionnaire’s structure, an Exploratory Factor Analysis (EFA) was conducted, revealing a robust three-factor structure that accounted for a significant proportion of the variance. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was 0.822, and Bartlett’s test of sphericity was significant (p < 0.001), supporting the suitability of factor analysis. Reliability was subsequently evaluated using both Cronbach’s Alpha and McDonald’s Omega, with each factor demonstrating strong internal consistency: Factor I (α = 0.926, ω = 0.928), Factor II (α = 0.851, ω = 0.861), and Factor III (α = 0.815, ω = 0.814).

Convergent validity was demonstrated by Composite Reliability, with values exceeding 0.70, and Average Variance Extracted (AVE) values above 0.40 [184]. Moreover, discriminant validity, assessed using the Fornell-Larcker criterion, showed that the square root of the AVE for each factor was greater than its correlations with other factors: Factor I (0.744), Factor II (0.753), and Factor III (0.733). These results confirmed the distinctiveness and structural validity of the instrument.

The final questionnaire included 19 items distributed across three constructs: (1) quality of design in the adaptation of content to context, (2) pedagogical quality of design, and (3) technical quality of design. The instrument was administered online via Google Forms and utilised a five-point Likert scale with the following options:

• MN = Inadequate/Strongly Disagree;
• N = Deficient/Disagree;
• R = Regular;
• P = Good/Agree;
• MP = Very good/Strongly agree.

To preliminarily explore the potential impact and feasibility of the game prototype, a flexible twelve-week intervention (February–April 2025) was conducted. This pilot implementation was carried out in a hospital classroom with four children undergoing oncology treatment. The prototype was integrated into the teacher’s regular pedagogical planning. Sessions were brief, lasting between 5 and 10 min, and were distributed across up to three moments during the day. The frequency and timing were adapted to each child’s condition and the rhythm of the hospital classroom, aiming to explore the potential and contextual relevance of the game in a real paediatric oncology setting. Participants used the game following the provided guidelines, with scheduling adjusted to classroom routines. The initiative originated from a teacher in charge of an oncology hospital classroom. Participation was voluntary and included informed consent from both families (through a collaborating association) and the teacher. The intervention was assessed using the Rapid Naming Test (TDR—Test de Denominación Rápida in Spanish), the Basic Instrumental Aspects in Language and Mathematics test (PAIB-1—Prueba de Aspectos Instrumentales Básicos en Lenguaje y Matemáticas), and the validated Scale of Reading Fluency in Spanish (SRFS).

4. Results

The serious game prototype evaluated in this study was developed using Unity (version 2022.3.3f1) and incorporates key learning mechanisms to support literacy development. Scaffolding is achieved through the progressive sequencing of tasks across the three themed worlds, while spaced repetition and active recall are embedded through the structured revisits of previously learned graphemes, syllables, and words—both within the Reading Club and through retrieval-based mini-games. The game was designed for low- to mid-range devices to ensure accessibility and feasibility across evaluation contexts. Minimum hardware specifications included the following: Windows 10 operating system, Intel Core i3 processor (or equivalent), 4 GB RAM, integrated GPU (e.g., Intel HD Graphics 4000), and 1 GB of available storage.

For testing purposes, all expert reviewers accessed the downloadable build on desktop computers, while the hospital-based pilot was implemented on a 2-in-1 convertible device (tablet-laptop hybrid) operating in tablet mode to simulate the touchscreen tablets intended for clinical use in hospital classrooms.

Following the technical deployment of the prototype, qualitative analyses were conducted using Atlas.ti software (Atlas.ti v.24, ATLAS.ti Scientific Software Development GmbH, Berlin, Germany) was used for all qualitative data analyses. Qualitative findings indicated that hospital-based teachers anticipated strong emotional engagement and perceived the serious game as a potential motivator for children to explore and participate actively. According to both the focus group with six hospital-based teachers and the individual interviews conducted with 30 design experts, the game was considered to be well-aligned with the intended learning objectives and appropriate for the hospital context. The expert sample included five specialists from each of six key fields involved in serious games design of this nature: educational game design; educational technology; psycholinguistics and reading processes applied to digital environments; computer engineering, programming and development; hospital education and health-related technologies; and accessibility in reading-focused games.

In the focus group discussion (N = 6), participants were invited to share their experiences with the game and assess its design, narrative, pedagogy, and challenges in relation to the intended educational outcomes. There was broad consensus that the game was well-suited to the needs of children undergoing cancer treatment. Specifically, the participants agreed that the objectives, activities and tasks, theoretical and didactic underpinnings, language, content, and presentation were all appropriate for the paediatric oncology context. From the perspective of hospital-based teachers, the alignment of the game with the clinical and emotional trajectories of paediatric cancer patients was regarded as one of its key strengths, contributing not only to literacy development but also to children’s social, emotional, and curricular growth.

Participants also highlighted the game’s visual and musical aesthetics as particularly effective in capturing children’s attention. The technological nature of the tool was seen as an advantage over traditional paper-and-pencil methods, increasing learner engagement and making the learning experience more enjoyable. Additional strengths cited included the game’s intuitiveness and ease of use, its accessibility and inclusivity, and the originality and challenge level of the embedded activities.

Nonetheless, several design and technical challenges emerged during the evaluation. Teachers reported difficulties in exiting the game and returning to the start menu and noted that the time required to access interactive segments or mini-games could be improved. These points of feedback were addressed in the subsequent iteration of the game prototype.

In addition to the focus group with hospital teachers, individual interviews were conducted with the 30 design experts previously described, representing complementary domains involved in serious game design development, with gender representation being relatively balanced. These experts affirmed the strength of the theoretical foundation underlying the prototype, its contextual suitability for paediatric hospital settings, and its potential to foster intrinsic motivation among young patients. They also praised the integration of playful and educational elements, the clarity and simplicity of the narrative, and language deemed developmentally appropriate for the target audience and the overall usability of the interactive environment. The visual and auditory quality of the game, along with its feasibility for implementation in real hospitalisation conditions, were also highlighted as key advantages.

Areas identified for further development included the need to enhance technical performance, diversify the range of tasks, and provide clearer, more accessible instructions to support user autonomy. These suggestions were implemented in the final version of the video game to improve functionality, educational effectiveness, and alignment with the hospital learning environment.

Concerning the QUAN investigations, 274 experts (N = 274) participated in the evaluation (

Table 1. Descriptive statistics for demographic data (N = 274).

Variable	Category	n	%
Gender	Female	92	33.6
	Male	176	64.2
	Non-binary	2	0.7
	Gender-fluid	1	0.4
	Prefer not to say	3	1.1
Age group (years)	<30	36	13.1
	30–39	84	30.7
	40–49	110	40.1
	50–59	37	13.5
	≥60	7	2.6
Country of residence	Spain	204	74.5
	Argentina	22	8.0
	Colombia	15	5.5
	Mexico	15	5.5
	Venezuela	5	1.8
	Ecuador	5	1.8
	Chile	2	0.7
	Peru	2	0.7
	Cuba	1	0.4
	Guatemala	1	0.4
	Dominican Republic	1	0.4
	Uruguay	1	0.4
Primary workplace	University	144	52.6
	Video-game production company	112	40.9
	Non-university	11	4.0
	Training company	7	2.6
Primary professional activity	Professional video-game designer	126	46.0
	Lecturer	110	40.1
	Lecturer/Researcher	38	13.9
Expert competence coefficient (CCE)	Mean (SD)	0.871 (0.058)	–

), with over 50% of participants being male and middle-aged. Participants received access to the serious game prototype through a downloadable build provided via Google Drive, which they explored prior to completing the online questionnaire. All statistical analyses were conducted using SPSS software (SPSS v.27, IBM Corporation, Armonk, NY, USA).

This study also explored participants’ views of serious game design (

Table 2. Descriptive statistics for the serious game prototype (N = 274).

Dimensions	M	SD
D1. Quality of design in the adaptation of content to context	4.28	0.396
D2. Pedagogical quality of design	4.16	0.444
D3. Technical quality of design	4.27	0.433

), with the experts’ responses in the distributed instrument being overly positive and consistent (Cronbach’s alpha = 0.86) across all the examined constructs.

Furthermore, item-level scores were analysed within each of the three evaluated dimensions.

Table 3. Mean evaluation and standard deviation carried out by the experts across items of the dimension: “Quality of design in the adaptation of content to context”.

Quality of Design in the Adaptation of Content to Context	M	SD
The prototype’s content is based on a solid and coherent theoretical framework.	4.44	0.604
The language used in the prototype is tailored to the target student audience.	4.34	0.649
The content delivery facilitates the intended learning outcomes.	4.15	0.673
The prototype responds accurately and as expected to user interactions.	4.16	0.729
The auditory and visual representations employed are aesthetically pleasing.	4.23	0.646
The prototype’s design offers a reasonable solution to the specific needs of the students.	4.34	0.683
The prototype design is adapted to the specific location and time of use.	4.34	0.690
All necessary factors for ensuring successful game implementation in hospital classrooms have been considered in the design.	4.21	0.651
The prototype is designed for use in the intended space, considering the learning times of the target students and the resources available in hospital classrooms.	4.34	0.655

presents the mean values and standard deviations for the dimension titled “Quality of Design in the Adaptation of Content to Context.”

The mean rating for all items within this construct was M = 4.28 (SD = 0.396), with individual item scores ranging from 4.15 to 4.44. These results reflect a high overall assessment of the prototype’s quality in adapting content to the specific educational context. Furthermore, the experts indicated that the prototype is grounded in a robust theoretical framework and makes effective use of language and methodology that are well-aligned with the needs of the target learner population and the characteristics of the hospital-based learning environment. Additionally, the game demonstrated adequate responsiveness and featured well-balanced auditory and visual aesthetics. Its development also accounted for key factors necessary for successful implementation within the intended educational context.

The highest-rated item was the theoretical soundness and internal coherence of the prototype (Item 1; M = 4.44, SD = 0.604), followed by its contextual and temporal adaptability (Item 7; M = 4.35, SD = 0.690). The lowest score within this dimension was assigned to the prototype’s capacity to facilitate the intended learning outcomes (Item 3; M = 4.15, SD = 0.673). Nevertheless, this score still reflects a clearly positive expert perception of the didactic methodology employed.

Overall, the aforementioned results suggest a high level of agreement among respondents, as demonstrated by the relatively low standard deviations. The strongest consensus was observed for the item on theoretical coherence (Item 1), while slightly more variation in responses was found for the item addressing the prototype’s responsiveness to player actions (Item 4; M = 4.16, SD = 0.729).

Table 4. Expert assessment of the items of the dimension “Pedagogical quality of design”.

Pedagogical Quality of Design	M	SD
The content is narratively diverse enough to stimulate learning.	4.09	0.605
The prototype has specific, didactic, and relevant objectives for educational use.	4.39	0.667
Relevant and sufficient activities and challenges are presented.	4.13	0.650
The prototype elicits or helps elicit students’ interest.	4.09	0.611
The prototype enables user assistance. It provides training levels that facilitate learning the game mechanics.	4.10	0.597

presents the mean scores and standard deviations for the dimension “Pedagogical Quality of Design.”

The mean rating for all items within the Pedagogical Quality of Design construct was M = 4.16 (SD = 0.444), with individual item scores ranging from 4.09 to 4.39. These results reflect a high perceived pedagogical quality in the design of the serious game prototype. According to expert evaluations, the learning objectives were considered appropriate, the instructional activities relevant and sufficient, and the user support mechanisms effective in facilitating learning. Additionally, the content was viewed as varied and engaging, with the potential to stimulate learners’ interest.

The highest-rated elements within this dimension were those related to the game’s educational objectives and instructional activities. Experts regarded the objectives as specific, pedagogically grounded, and relevant for the target audience (Item 11; M = 4.39, SD = 0.667). Similarly, the activities were rated as both pertinent and sufficient for achieving intended learning outcomes (Item 12; M = 4.13, SD = 0.650). Conversely, the lowest-rated aspects were the variety of content (Item 10; M = 4.09, SD = 0.605) and the prototype’s capacity to generate learner interest (Item 13; M = 4.09, SD = 0.611). Despite these lower ratings, the scores remain within a high and favourable range.

Low variability in responses across most items suggests a strong consensus among experts. Notably, there was high agreement regarding the level of training and support provided by the game (Item 14; M = 4.10, SD = 0.597). The greatest variation in expert opinion occurred with respect to the adequacy and relevance of the educational objectives (Item 11; SD = 0.667), though the mean rating remained the highest within the construct.

Table 5. Expert assessment of the items of the dimension “Technical quality of design”.

Technical Quality of Design	M	SD
The prototype operates properly in terms of accessibility, stability, and load speed (technical performance), without significant delays in user response.	4.31	0.631
The controlled units (elements and interface) react to user actions.	4.28	0.662
The prototype allows the user to interact easily and adaptively, providing a pleasant feedback experience.	4.13	0.565
The prototype’s content and its components can be evaluated.	4.41	0.716
During game development, information on its design can be obtained for improvement.	4.20	0.736

reports the mean scores and standard deviations for the dimension “Technical Quality of Design.”

The technical quality of the prototype was likewise positively assessed, with an overall mean of M = 4.27 (SD = 0.433). Individual item means ranged from 4.13 to 4.41, indicating consistently strong expert approval. Experts affirmed that the prototype demonstrated adequate technical performance, responsive interaction, intuitive controls, and satisfying feedback mechanisms. They also emphasised the importance of the prototype’s evaluability, noting that the system architecture and content components lend themselves well to assessment and future improvements. In this context, evaluability refers to the extent to which the prototype’s structure, functionalities, and pedagogical elements are clearly defined, observable, and measurable, enabling systematic assessment and iterative refinement. Furthermore, the clarity of the information provided about the design was seen as a strength, supporting the potential for iterative refinements based on user feedback and contextual requirements.

The highest technical ratings were associated with the evaluability of the prototype’s content and components (Item 18; M = 4.41, SD = 0.716), followed by its accessibility, stability, and overall system performance (Item 15; M = 4.31, SD = 0.631). The lowest-rated technical aspect, though still favourably evaluated, was the ease of interaction with the prototype to receive pleasant feedback (Item 17; M = 4.13, SD = 0.565).

Although responses across this construct remained relatively consistent, variability was slightly higher than in the previous two dimensions. Greater dispersion was observed, particularly in items related to the availability of information for prototype improvement (Item 19; M = 4.20, SD = 0.736) and the evaluability of content and components (Item 18; M = 4.41, SD = 0.716).

Taken together, these findings indicate that the prototype’s content is appropriate and relevant for the hospital classroom context. The educational approach is well-structured, coherent, and tailored to the developmental and emotional needs of young learners in clinical settings. Furthermore, the technical design demonstrates a high level of robustness, contributing to a smooth and accessible user experience. Collectively, these results validate the prototype’s promise as an effective educational resource in hospital-based learning environments.

In addition to analysing expert responses for each construct, correlations among the three dimensions of the questionnaire were also examined using Spearman’s rank-order correlation coefficient. The analysis revealed statistically significant positive correlations among all constructs (p < 0.001). The strongest association was observed between content-context adaptation and technical quality of design (r = 0.63), followed by the correlation between content-context adaptation and pedagogical quality of design (r = 0.52). The correlation between technical and pedagogical quality was also significant, though more moderate (r = 0.45). These relationships are visually illustrated in Figure 15.

These correlation patterns suggest that expert ratings of contextual adaptation were strongly linked to perceptions of both pedagogical and technical quality. Additionally, positive evaluations of pedagogical design were often accompanied by favourable assessments of technical performance. The observed differences in correlation strength highlight the centrality of contextual alignment in expert evaluations, indicating that well-contextualised design contributes meaningfully to both pedagogical coherence and technical functionality.

The experts who participated in the quantitative phase of the study represented two primary professional backgrounds: individuals from the educational field (e.g., professors and researchers) and professionals from the video game design sector. Given the differing domains of expertise, it was hypothesised that their evaluations of the prototype might also diverge, particularly with regard to aspects most aligned with their respective areas of specialisation. For example, it was anticipated that designers might be more critical when evaluating technical features, while educators might apply more stringent criteria to the assessment of pedagogical elements.

To examine potential differences in evaluation between the two groups, the Mann–Whitney U test was conducted, appropriate for ordinal data and non-parametric distributions. Bonferroni correction was applied to adjust for multiple comparisons and reduce the risk of Type I error.

The results revealed no statistically significant differences in the item-level ratings between the two groups (p > 0.05 across all comparisons). As shown in

Table 6. Mean and standard deviation (in parentheses) per item and group analysed.

Construct	Item	Activity
		Professional Game Design (n = 126)	Teaching/Research (n = 148)
Quality of design in the adaptation of content to context	1	4.524 (0.547)	4.372 (0.642)
	2	4.397 (0.608)	4.284 (0.681)
	3	4.246 (0.666)	4.074 (0.671)
	4	4.095 (0.774)	4.216 (0.686)
	5	4.151 (0.633)	4.291 (0.652)
	6	4.460 (0.615)	4.236 (0.722)
	7	4.444 (0.627)	4.264 (0.732)
	8	4.310 (0.600)	4.128 (0.683)
	9	4.444 (0.573)	4.243 (0.706)
Pedagogical quality of design	10	4.159 (0.557)	4.027 (0.638)
	11	4.429 (0.638)	4.365 (0.692)
	12	4.190 (0.678)	4.081 (0.623)
	13	4.119 (0.560)	4.061 (0.652)
	14	4.167 (0.603)	4.047 (0.587)
Technical quality of design	15	4.357 (0.613)	4.270 (0.645)
	16	4.325 (0.617)	4.250 (0.699)
	17	4.198 (0.537)	4.074 (0.584)
	18	4.429 (0.709)	4.385 (0.724)
	19	4.294 (0.658)	4.122 (0.790)

, both educational experts and video game designers provided consistently high ratings across all dimensions of the questionnaire. This convergence in responses despite professional background differences further reinforces the perceived overall quality and coherence of the serious game prototype.

Taken together, the results of the quantitative evaluation indicate a highly favourable appraisal of the prototype across its three core dimensions: content-context adaptation, pedagogical quality, and technical quality. Importantly, these positive assessments were consistent regardless of the respondent’s area of expertise. That is, professors and researchers were not disproportionately critical of pedagogical aspects, nor were game designers more exacting in their assessment of technical components.

After integrating and triangulating these findings with the qualitative data, it can be concluded that the educational game prototype, Yuki’s Adventure: Hidden Words, demonstrates strong contextual alignment, pedagogical soundness, and technical robustness. These characteristics collectively support its potential application within paediatric oncology classroom settings. Specifically, the game offers promising benefits for fostering rapid naming, initial decoding, and automatization in the recognition of syllables, words, and basic sentences, core components in early literacy development for hospitalised learners.

The mixed-methods integration was a central phase in the evaluation of the prototype, consistent with the exploratory sequential design adopted for this study. The integration process was not limited to surface comparison but involved a systematic cross-analysis to identify convergences, logical divergences due to iterative development, and expansions that enriched the interpretation of results. To facilitate this process, a joint display strategy was employed, aligning analytical dimensions across both strands using MAXQDA 24 software (MAXQDA 2024, version 24.4.1, VERBI Software, Berlin, Germany). Qualitative findings from expert interviews and the focus group guided iterative refinements of the prototype, and these changes were subsequently validated through quantitative ratings. This developmental logic resulted in patterns of partial convergence, logical divergences by development, and confirmatory expansion across constructs.

Three joint displays were constructed, as can be seen in

Table 7. Summary of mixed-methods integration types by construct.

Construct	Integration Outcome	Nature of Integration
Contextual Adaptation	Converging and expanding	Qualitative insights guided design refinement (e.g., language, flexibility, aesthetics). Likert ratings confirmed adequacy. Iterative adjustments resolved early concerns.
Pedagogical Quality	Logical divergence + confirmation	Experts identified areas for improvement (challenge variety, progression); iterative adjustments led to converging positive ratings.
Technical Quality	Logical divergence + confirmation and expansion	Experts initially identified technical issues (bugs, navigation). These informed refinements were later validated by high usability and performance ratings. Integration revealed expansion and convergence.

, corresponding to the study’s core constructs: contextual adaptation, pedagogical quality, and technical quality. These displays demonstrated high coherence across data strands. Experts’ initial concerns during the qualitative phase, such as clarity of instruction, progression, or visual consistency, were addressed in the redesign, and high quantitative ratings (all means above 4 on a 5-point Likert scale) confirmed the success of these adjustments.

The integration process yielded the following outcomes:

• Convergence: Both data strands confirmed the strengths of the prototype in terms of contextual relevance, ease of use, and motivational appeal. For example, the prototype’s alignment with paediatric oncology conditions (e.g., short sessions, reduced load) was consistently praised across methods.
• Expansion: Qualitative narratives provided depth to quantitative trends. While visual and auditory aesthetics received high scores, experts elaborated on specific tensions (e.g., background music monotony, visual overstimulation) that informed future improvements.
• Divergence by development: In areas such as feedback mechanisms or instructional clarity, qualitative critiques prompted design revisions. These refinements were later validated by positive quantitative ratings, illustrating a cycle of development-confirmation that aligns with Design-Based Research principles.
• Cross-construct coherence: The analysis revealed interdependencies among the three core constructs. For instance, high technical quality was necessary to deliver pedagogical content effectively, and the contextual adaptation of the interface supported both usability and didactic effectiveness. Adaptation to context emerged as the central axis reinforcing both pedagogical and technical quality.
• Meta-inferences: The integrated results validated the relevance of Design-Based Research as a methodological framework for creating digital learning tools tailored to clinical-educational environments. The iterative integration of expert input contributed to a high-fidelity prototype evaluated as pedagogically sound, technically functional, and contextually appropriate for hospital classrooms.

Additionally, the analysis revealed strong interdependence among the constructs. For instance:

• The technical performance of the prototype was directly linked to its contextual feasibility in hospital settings. Stability and ease of interaction were essential to avoid frustration in vulnerable learners.
• The pedagogical strength of the game depended on its contextual adaptation. Tailored pacing, language, and representation of content were crucial to addressing the cognitive and emotional needs of children undergoing cancer treatment.
• Technical and pedagogical dimensions were mutually reinforcing. Experts noted that intuitive interaction design supported educational understanding, while visual and auditory coherence facilitated cognitive engagement.

Finally, this phase enabled the formulation of further meta-inferences about the success of the prototype:

• The presence of multiple logical divergences by development confirmed that initial qualitative criticisms triggered meaningful improvements, validated later through high Likert ratings (all means > 4).
• The most consistent point of integration was the adaptation to the paediatric oncology context, which acted as a central pillar in enhancing both pedagogical relevance and technical feasibility.
• The concept of confirmation with expansion emerged repeatedly: quantitative validation was complemented by rich qualitative insight that explained why the design was perceived as effective.

This integrated analysis highlighted how iterative refinement, grounded in expert feedback, resulted in a final prototype that is pedagogically, contextually, and technically sound. The mixed-methods process validated the logic of the design-based research (DBR) framework, where expert-informed adjustments during the qualitative phase were corroborated by positive quantitative evaluation scores. Additionally, the analysis also confirmed the multidimensional strength of the final prototype, which validated the prototype’s pedagogical, contextual, and technical adequacy for use in hospital-based pre-fluency interventions targeting young children with central nervous system tumours.

In light of the highly positive expert evaluations and the lack of statistically significant differences across professional backgrounds, an exploratory assessment of the prototype’s potential impact and feasibility in real-world clinical-educational settings was considered appropriate. This initiative was encouraged by a teacher working in an oncology hospital classroom in the Canary Islands, who recognised the educational promise of the prototype and advocated for its integration into daily instructional practice. Notably, this application took place despite the prototype still being under development and not yet finalised.

Given the early educational stage of the participants (ages 5–6, enrolled in the final year of Infant Education) and the limited availability of validated literacy assessment tools designed for this age group in the Spanish context, careful methodological consideration was required for instrument selection. In Spain, the official Infant Education curriculum promotes a global, exploratory introduction to literacy, with formal reading instruction typically beginning in Primary Education. As such, administering comprehensive standardized reading tests at this stage would be both pedagogically inappropriate and psychometrically unreliable.

Moreover, the exploratory nature of this pilot, implemented as a single-group post-test design, was not intended to establish causal relationships or generate generalizable findings. Instead, it aimed to offer preliminary insights into the prototype’s educational potential. To achieve this, three validated instruments were selected for their alignment with foundational literacy skills: the Rapid Naming Test (TDR), the reading subtest of the Spanish Test of Basic Instrumental Aspects in Language and Mathematics (PAIB-1), and section 3A of the Spanish Reading and Writing Test (LEE), focusing on basic sentence-level reading. These tools specifically targeted rapid naming, initial decoding, and early indicators of fluency, thus aligning with both developmental appropriateness and curricular expectations.

In addition, fluency performance was assessed using the Scale of Reading Fluency in Spanish (SRFS; Escala de Fluidez Lectora en Español, EFLE) [185]. The TDR measured rapid automatized naming, widely acknowledged as a strong predictor of reading fluency. The PAIB-1 captured syllable and word-level reading skills in Spanish-speaking preschoolers, while the selected LEE items assessed sentence-level decoding. The SRFS was applied across both PAIB-1 and LEE tasks, offering a multicomponential fluency evaluation encompassing reading speed, accuracy, prosody (including volume, intonation, pauses, and phrasing), and an additional component (reading quality) to provide a holistic perspective. Each component was rated using a four-point scale (1 = lowest, 4 = highest), with standardised descriptors outlined in Appendix 1 of the EFLE scoring scale [185].

Four children (1 male, 3 female) participated in this pilot implementation, selected according to strict inclusion criteria: age (5–6 years), enrolment in preschool, no prior formal literacy instruction, ongoing treatment for oncology conditions, and diagnosis involving the central nervous system.

As presented in

Table 8. Rapid Naming Test (TDR in Spanish) Results.

Participant	Objects	Colours	Numbers	Letters	Level (Estimated Percentile)
P1	49″	44″	39″	41″	High (≥p90)
P2	65″	69″	42″	48″	Medium (~p50)
P3	67″	72″	44″	47″	Medium (~p40)
P4	51″	46″	42″	43″	High (p80–90)

, results from the Rapid Naming Test indicated that two participants (P1 and P4) exhibited high naming speed (≥80th percentile), while the remaining two (P2 and P3) demonstrated moderate performance, falling between the 40th and 50th percentiles. The administered subtests included object and colour naming, standardised for this age group, and letter and number naming, which were used qualitatively and interpreted using first-grade reference norms due to the participants’ pre-literacy stage.

In the PAIB-1, designed to assess isolated word reading, three out of four participants performed within the medium range on all indicators (direct score, centile, standardised score, and T score). One participant (P3) fell into the low range, indicating difficulties in early reading processes (

Table 9. PAIB-1 Reading Scores.

Participant	Direct Score	Percentile	Standardised (Score)	T Score	Interpretation
P1	23	35 (medium)	7 (medium-high)	44 (medium)	Medium
P2	21	25 (medium)	6 (medium-high)	42 (medium)	Medium
P3	16	15 (low)	4 (medium-low)	38 (low)	Low
P4	22	30 (medium)	6 (medium-high)	43 (medium)	Medium

).

For the assessment of sentence-level reading fluency, two sentences from Section 3A of the LEE (Test de Lectura y Escritura en Español) were selected. Although the LEE is standardised for students in Grades 1 to 4 of Primary Education, the chosen items were specifically selected for their brevity and syntactic simplicity, rendering them suitable for the developmental level of the participants, who were five-year-old children undergoing oncology treatment who had not yet commenced formal literacy instruction.

Given these developmental considerations, the administration of full subtests or longer passages would have been inappropriate, both pedagogically and methodologically. Moreover, since the game prototype was still in its early stages and not designed to produce significant gains in complex reading comprehension, these simplified sentences served as controlled indicators of emergent sentence-level fluency. To complement this, word-level reading was independently assessed using the PAIB-1 test, which includes isolated lexical items appropriate for learners at the pre-reading stage.

The adaptation of the LEE—limiting it to two brief sentences—was both ethically justified and methodologically necessary to ensure developmental validity, reduce cognitive burden, and maintain alignment with the study’s exploratory aims.

Based on the application of the SRFS-EFLE (Escala de Fluidez Lectora en Español) across performance on the PAIB-1 and LEE assessments, 75% of participants (3 out of 4) achieved an overall fluency score equal to or above the group mean (M = 3.0). These students (Students 1, 2, and 4) obtained ratings of 3 or higher in at least three of the four assessed components: reading speed, accuracy, prosody, and reading quality. In contrast, Student 3 showed below-average performance, particularly in speed, prosody, and reading quality, which is consistent with their lower scores on the PAIB-1 subtest, as detailed in

Table 10. Reading Fluency Components—SRFS-EFLE.

Component	P1	P2	P3	P4
1. Speed	4	3	2	3
2. Accuracy	3	3	3	3
3. Prosody	3	3	2	3
4. Reading Quality	4	3	2	3

and

Table 11. Descriptive Statistics—SRFS-EFLE (N = 4).

Component	Mean (M)	Median (Md)	Mode (Mo)	SD
1. Speed	3.00	3.00	3.00	0.82
2. Accuracy	3.00	3.00	3.00	0.00
3. Prosody	2.75	3.00	3.00	0.50
4. Reading Quality	3.00	3.00	3.00	0.82

.

Given the limited sample size (N = 4), which reflects the exploratory nature of the pilot within a hospital classroom setting, the analysis was restricted to descriptive statistics, without inferential testing. The results suggest a generally favourable trend in early reading fluency development following the intervention. Specifically, three out of four participants achieved scores at or above the group mean in overall initial fluency, as measured by the SRFS-EFLE scale.

The strongest performance was recorded in the accuracy component, where all participants received an identical score (M = 3.0, SD = 0.00), indicating a high and consistent ability to decode words or sentences correctly. In contrast, greater variability was observed in the components of reading speed and reading quality (SD = 0.82). While not designed to address effectiveness, this pilot study aimed to explore feasibility and inform final development; subsequent research should, therefore, incorporate larger samples and control group designs where feasible.

5. Discussion, Conclusions and Future Research

This study presented the instructional design, development, and evaluation of a serious game prototype aimed at fostering rapid naming, decoding, and reading automatization in preschool-aged children undergoing long-term hospitalisation in oncology settings.

The game was grounded in an extensive three-year process of literature review and iterative development, informed by evidence-based pedagogical practices, principles of game design, and the experiential insights of hospital educators, alongside experts in Educational game design; educational technology; computer engineering, programming and development; psycholinguistics and reading processes applied to video games; hospital education and health with technology; and accessibility in reading games.

Drawing on evidence-based practices from paediatric oncology classrooms, the resulting prototype was designed to merge playful engagement with targeted literacy outcomes. Its evaluation revealed that the game aligns well with the educational and clinical needs of its intended context. This alignment was further validated through highly favourable feedback from a diverse group of experts—including preschool hospital teachers and game design experts and professionals—who praised its pedagogical coherence, usability, and contextual relevance. The convergence of insights from these varied perspectives lends strong credibility to the qualitative assessment and reinforces the prototype’s potential as a valuable supplementary tool for supporting instructional goals in paediatric hospital classrooms.

Quantitative results mirrored this positive perception: across all three evaluated dimensions (contextual adaptation, pedagogical quality, and technical quality), mean scores ranged from 4.16 to 4.28 out of 5, indicating uniformly high levels of approval among 274 expert respondents. These data corroborate the prototype’s potential to bridge gaps in stimulating pre-reading fluency processes under clinical and developmental constraints.

Although the post-test pilot sample was limited to four children—reflecting the exploratory nature of the study—the primary aim was not to evaluate effectiveness, but rather to assess the feasibility of implementing the prototype within real hospital classroom settings and to inform its future development. Preliminary findings suggest that the game may support early gains in reading fluency, particularly in naming accuracy and basic decoding skills. These encouraging results highlight the need for further research with larger and more diverse clinical samples to evaluate the game’s generalizability and its potential for measurable educational impact.

Importantly, the small pilot cohort was not pre-planned but emerged from a contextual opportunity, encouraged by a teacher who saw value in the game’s educational potential even at a pre-final stage. Ethical considerations, especially the participants’ vulnerability due to age, health condition, and treatment status, necessitated a cautious, non-intrusive exploratory approach, consistent with the British Educational Research Association (BERA) Risk–Benefit Guidelines. As such, the study was exploratory and descriptive in nature, with a focus on validating design quality rather than establishing statistical significance in learning outcomes. Although children’s subjective experiences and motivational indicators were not systematically assessed using formal instruments, informal teacher feedback indicated a generally positive reception, with boys showing particular enthusiasm. This anecdotal input, though not empirically analysed, offers initial insight into engagement and will inform future research.

Furthermore, this approach facilitated the extraction of evidence-informed design principles while simultaneously identifying key areas for improvement—such as accessibility, interface usability, narrative coherence, and motivational appeal—all of which have been implemented in the more recent version of the prototype. Building on these insights, this research calls for a re-examination of traditional educational design paradigms, advocating for pedagogical innovation through interactive and inclusive technologies. In doing so, it also surfaces important epistemological, methodological, and ethical considerations that future studies must address to establish a rigorous framework for developing serious games in sensitive clinical contexts.

In addition to its methodological contribution, this study presents a serious game whose design embodies a set of distinctive features that clearly differentiate it from existing literacy-oriented educational games. First, unlike general-purpose applications, the prototype was conceived specifically for preschool children undergoing oncological treatment for CNS tumours, a population that frequently exhibits neurocognitive vulnerabilities in rapid naming, processing speed, and fluency. The game is not intended as a vehicle for formal reading instruction; rather, it serves as a pre-reading cognitive-linguistic training tool, focusing on foundational processes such as syllable segmentation, rapid naming, and decoding automatization, which are critical precursors to fluent reading. Importantly, these tasks are not presented in isolation, but are embedded in a high-fidelity, story-driven role-playing environment that substitutes traditional drill-based approaches with quests, virtual classrooms and interactive scenarios. This integration of pedagogy and gameplay is supported by an eclectic instructional model that aligns each learning objective with specific mini-games and narrative mechanics (see Appendix A, Appendix B and Appendix C). Furthermore, the aesthetic design—including neutral characters, adaptive pacing, and emotionally safe audio-visual elements—was intentionally developed for hospitalised contexts, where children often face attentional variability, fatigue, and emotional vulnerability. Finally, the prototype underwent expert validation involving both clinical-educational professionals and game design specialists, ensuring its feasibility and relevance in paediatric oncology settings. Taken together, these characteristics position Yuki’s Adventures: Hidden Words as a context-specific, theory-informed, and clinically sensitive innovation that extends current research in educational technology and provides a foundation for future studies on fluency development in vulnerable paediatric populations.

Building on these, the primary contribution of this work lies in the elaborate design of an adaptable, pedagogically robust serious game that harmonises educational content with engaging gameplay for children facing extraordinary learning barriers. The resulting design patterns offer a valuable reference point for educators, researchers, and developers seeking to create meaningful, skill-transfer-driven serious games in similarly constrained or high-stakes learning environments.

Beyond its practical design value, this study also advances current literature in several key areas. First, it addresses a significant gap by designing and evaluating a serious game tailored to hospitalised preschool children with CNS tumours—an underrepresented population in educational game research. Second, it offers an initial validation of the preliminary prototype’s design, grounded in principles of neuroeducation, inclusive pedagogy, and accessible game design, all tailored to a clinical-educational context. This validation serves as a foundational step towards developing a fully functional final version of the game, ensuring pedagogical coherence, contextual relevance, and usability are achieved prior to broader implementation. In this context, if a fully developed final version of the game were pursued, a 24-week implementation period—double the 12-week prototype—would be more suitable to foster deeper engagement and sustained learning. Third, the study makes a methodological contribution by applying a Design-Based Research model within a highly sensitive clinical setting, integrating expert validation with exploratory pilot testing. These elements set the work apart from previous studies and provide a replicable framework for advancing serious games design for vulnerable learners’ population.

In light of these contributions, the present study provides both an immediate educational resource and a foundation for future advancements in the design of serious games for vulnerable learners. Specifically, this study not only offers a validated tool tailored to the needs of children in paediatric oncology but also proposes a methodological model for developing and evaluating serious games in complex learning contexts. In doing so, it highlights the transformative potential of educational games when they are grounded in learner-centred pedagogy, guided by empathy, and developed with scientific rigour. At the same time, it challenges the traditional view of video games as mere entertainment, reframing them instead as powerful cognitive and emotional mediators within fragile and discontinuous educational environments such as hospital classrooms.

Within this scope, the present study did not integrate extended user-centred data such as systematic child feedback or long-term gameplay analytics. While these forms of evidence would certainly enrich final development, their collection lies beyond the objectives of the present exploratory phase. Instead, the prototype offers a conceptual and technical foundation that can serve as a reference point for subsequent projects or collaborations in which user-centred evidence might be incorporated more extensively.

Moreover, although no further iterations are planned at this stage, as the final prototype design has been completed, validated, and the exploratory research line concluded, the study has laid a strong foundation and produced a validated design and design principles that can inform the development of the full-scale serious game in the future, contingent on adequate funding and resources. Looking ahead, research exploring learner motivation and engagement in similar hospital-based serious games could benefit from validated, developmentally appropriate instruments tailored to preschool populations. The Leuven Scale of Involvement [186], for instance, provides a structured observational framework for assessing behavioural engagement and emotional well-being, while the Fun Toolkit [187] offers child-friendly self-report methods such as the Smileyometer and Fun Sorter to capture affective responses to digital learning. These instruments are well aligned with the developmental and cognitive profiles of hospitalised preschoolers and could be applied effectively in clinical-educational contexts in future studies within the field.

Evaluating motivation and engagement in this highly specific population—five-year-old children hospitalised with CNS tumours—would also require an ethically sensitive and contextually appropriate design. A complementary concurrent triangulation mixed-methods approach [188], prioritising quantitative measures while drawing on qualitative observations for additional nuance, could offer a robust framework. Such an approach would acknowledge the multifactorial nature of engagement, which is influenced not only by the game itself but also by mediators such as teacher facilitation [189], environmental conditions, and the impact of illness [190].

Ultimately, this study marks an important step forward while also opening several avenues for future research. Building on the present findings will require studies with larger samples, control groups, and both baseline and post-intervention assessments using validated, age-appropriate instruments, complemented by longitudinal follow-up. In parallel, integrating in-game analytics, AI-based scaffolding, and speech recognition will be crucial to enable real-time adaptation and support personalised learning trajectories. Finally, examining the transferability of this framework to other sensitive educational contexts could extend its social impact and enhance its broader systemic relevance.