Érase una vez: An Exploratory Study of a Therapeutic Game for Enhancing Verbal Fluency in Children

Abstract

Serious games represent a promising pedagogical approach by combining therapeutic and educational goals through play. This study examined the potential of a culturally adapted version of Érase una vez, a narrative-based card game originally developed in Brazil, to foster and assess narrative competence in school-aged children. Twenty typically developing children (ages 8–9 and 11–12) participated in both individual written and group oral storytelling tasks. Usability perceptions were assessed through questionnaires and the System Usability Scale, while linguistic performance was analyzed using quantitative and qualitative approaches. Statistical methods and computational linguistic tools were applied to measure lexical, syntactic, and pragmatic features. Results indicated high engagement and satisfaction, with over 95% of participants reporting positive experiences. Narrative productions revealed significant age-related differences: older children generated longer and syntactically more complex stories, while younger groups produced simpler structures with lower lexical variety. Group narratives reflected classroom-level effects, with sixth graders achieving greater cohesion and creativity. Findings support the integration of Érase una vez as both a pedagogical and therapeutic tool. Its playful and flexible format promotes motivation, reduces performance anxiety, and elicits rich, naturalistic language samples. Despite limitations of sample size and design, results encourage further exploration of serious games in educational and clinical contexts.

1. Introduction

Verbal fluency tasks are widely used in clinical and neuropsychological assessments to evaluate language and executive control and to detect learning-related difficulties (Shao et al., 2014). Beyond its diagnostic value, verbal fluency is closely linked to broader communicative abilities, including narrative competence, as both rely on lexical retrieval, semantic organization, and executive processes such as inhibition and cognitive flexibility. In this sense, verbal fluency can be understood not only as a measure of isolated linguistic performance but also as a key component of structured narrative production.

In recent years, therapeutic games have emerged as engaging tools for both assessment and intervention in this domain. These approaches have been applied across a range of clinical and educational contexts, including depression, bipolar disorder, ADHD, and substance use, showing promising outcomes that are, in some cases, comparable to those of established treatments (Dewhirst et al., 2022; Barnes & Prescott, 2018; Maurin et al., 2020; Schoneveld et al., 2018; Merry et al., 2012). However, much of this work has focused on either engagement or clinical outcomes, with less attention paid to how game-based interactions may specifically elicit and structure linguistic and narrative processes.

Therapeutic games typically rely on three interrelated components. First, the ludic element, rooted in play, supports learning as well as cognitive, social, and creative development (Huizinga, 2020). Second, gamification involves the use of game elements in non-game contexts to enhance engagement, learning, and problem solving, often through features such as points, badges, and rewards (Deterding et al., 2011; Kapp, 2012). Third, serious games apply these principles within structured systems, such as board or card games, to achieve specific goals (Kapp, 2012) and can be implemented in both digital and physical formats. Their effectiveness is often strengthened when grounded in theoretical frameworks such as cognitive behavioral therapy, which emphasizes goal-oriented strategies, feedback, and behavioral reinforcement, elements that align closely with gamified environments (Beck, 2021; Stallard, 2019).

Within this framework, different types of game-based interventions have shown positive effects on language-related outcomes (see

Table 1. Examples of therapeutic games.

Game	Authors	Description	Targeted Skills	Key Findings
Play Therapy (Child-Centered/Group Play)	(Danger & Landreth, 2005)	Free or guided play sessions integrated with therapeutic goals	Expressive language, receptive language, social communication	Improves expressive and receptive language; supports fluency development indirectly
Digital Game-Based Therapy (Apps & AI tools)	(Du et al., 2023; Utepbayeva et al., 2024)	Mobile apps and AI-based interactive games used in therapy sessions	Speech production, engagement, repetition, phonological skills	Moderate improvements; high engagement and motivation enhance outcomes
Technology-Enhanced Interactive Systems (e.g., social robots)	(Georgieva-Tsaneva et al., 2023)	Game-like interaction with robots or conversational AI systems	Communication skills, turn-taking, engagement	Promising results for improving communication via interactive scenarios
Arts-Based/Creative Play Approaches	(Ifenatuora et al., 2021)	Use of drama, storytelling, music, and visual arts in interactive formats	Vocabulary, pronunciation, expressive fluency, narrative skills	Enhancement of verbal expression and communication through immersive contexts
Structured Language Games (e.g., chain message games)	(Khoiriyah & Purwanta, 2025)	Rule-based verbal interaction games involving repetition, sequencing, and message passing	Speech fluency, articulation, coherence, confidence	Strong improvements in fluency, communication response, and speaking confidence
Early Intervention Play-Based Programs	(Regis et al., 2018)	Structured early therapy combining play with language stimulation techniques	Overall language development, fluency foundations	Early stimulation leads to significant improvements in communication skills

). Structured language games have demonstrated improvements in fluency and confidence (Khoiriyah & Purwanta, 2025), while play therapy supports expressive and receptive language, indirectly enhancing fluency (Danger & Landreth, 2005). Technology-based approaches, including digital tools and social robots, have shown moderate but promising effects, largely driven by increased engagement and interaction (Du et al., 2023; Utepbayeva et al., 2024; Georgieva-Tsaneva et al., 2023). Cognitive–linguistic games contribute by strengthening underlying processes such as memory and executive functions (Acosta-Rodríguez et al., 2022; Soltani et al., 2022), whereas arts-based approaches enhance expression through immersive and creative contexts (Ifenatuora et al., 2021). Finally, early intervention programs suggest that combining play with targeted language stimulation can lead to significant improvements in communication skills (Regis et al., 2018; Banu et al., 2023).

Despite these advances, there remains a gap in understanding how therapeutic games can be designed not only to enhance engagement but also to systematically elicit and analyze narrative production and its underlying linguistic components. In particular, the integration of narrative-based tasks with computational linguistic analysis offers a promising yet underexplored approach for capturing subtle variations in verbal fluency within ecologically valid contexts.

The present study addresses this gap by introducing a narrative-based therapeutic game designed to elicit structured verbal production in children. By combining principles of gamification with narrative prompts and computational linguistic indicators, this approach aims to explore how game dynamics can support and reveal processes related to verbal fluency. Within this exploratory framework, the study seeks to contribute to the development of more ecologically grounded and analytically sensitive tools for assessing and fostering language abilities in childhood. In this context, the present article introduces Érase una vez, a game developed to work with participants’ narratives through the creation of unique stories during gameplay and that has been used to assess verbal fluency in independent readers between the ages of 8 and 12. Its format has promoted engagement and sustained attention during administration, which enhances the quality of the assessment.

Érase una vez

Érase una vez is a card game that divides its content into six different categories, which represent the elements that will be present in the stories being created. These categories are: Character, referring to the main character of the story; Ally, who supports the main character; Adversary, the antagonist of the story; Place, representing where the story may unfold; Object, which includes items used in the narrative; and Surprise, which introduces a twist element, something that breaks the linearity of the story and requires the player to handle an unexpected situation. The game is played in turns, and in each turn, a die is rolled to determine which element will be added to the story. There is no required order or mandatory inclusion of specific elements. Each story is built using the elements determined in each round. In the first round, the die is rolled twice so that two cards from different categories can be drawn to initiate the story. From the second round onward, the die is rolled only once, and a single new element is added (see Figure 1).

The turn changes when the facilitator decides that the story has been sufficiently developed with the current elements, or when they judge that the addition of a new card will be therapeutically beneficial. The game ends when the story naturally moves toward a conclusion or when the time allotted for the activity runs out. In either case, the facilitator informs the participant(s) that the story should begin moving toward its ending, so they can develop a planned conclusion for the activity.

The main aim of this study was to examine the potential of the Spanish adaptation of Érase una vez, originally developed in Brazil and adapted to Spanish in the current study, as both an educational and therapeutic tool for fostering narrative competence in children. By incorporating six distinct story elements (Character, Ally, Adversary, Place, Object, and Surprise) the game provides a structured yet flexible framework that stimulates creativity, verbal fluency, and cognitive organization to be tested in the Spanish context. The turn-based mechanics, guided by dice rolls, introduce randomness that challenges players to adapt their narratives, encouraging problem solving, planning, and coherence in storytelling. Furthermore, the facilitator’s role in pacing the activity ensures that stories evolve with both developmental and therapeutic value, while the concluding phase promotes reflective closure. In this way, the overarching objective was to assess whether this playful, gamified format could enhance children’s linguistic expression and engagement while also serving as a sensitive instrument for evaluating cognitive–linguistic development.

2. Methodology

2.1. Participants

Given the exploratory nature of the study, the sample consisted of 20 typically developing Spanish children (ages 8–9 and 11–12 years) enrolled in a private bilingual school (for Spanish and English language). They were divided into four groups (n = 5 per group) based on age. Participants came from middle–high to high socioeconomic backgrounds, with no history of school absenteeism or academic issues. Teachers selected students with the aim of reflecting average classroom performance, intentionally avoiding extreme cases. While this approach was intended to enhance ecological validity, it may introduce selection bias, as the sample was not randomly drawn and may underrepresent variability in linguistic performance. Informed consent was obtained from parents or legal guardians.

Inclusion criteria included: (a) Spanish as a first language, (b) belonging to the specified age groups, and (c) average academic performance. Exclusion criteria included: (a) language or learning disorders, (b) dominant non-Spanish bilingualism, or (c) lack of signed consent.

In addition to the main sample, a panel of expert evaluators was included for qualitative assessment of the instrument. This panel consisted of one speech and language therapist, two primary school teachers, and one psychologist, all affiliated with the same educational center. These professionals contributed their expertise to qualitatively evaluate the content, clarity, and appropriateness of the game.

2.2. Materials

The intervention was implemented using a physical Spanish-adapted version of the Brazilian card game Era uma vez (2022). In the individual written activity, students used standard school materials, while for the group oral task an automatic speech transcription application was employed. Additionally, all participants and facilitators completed a personal experience questionnaire and an adapted version of the System Usability Scale (Brooke, 1986; Sevilla-Gonzalez et al., 2020), following the adaptation reported by Ahulló et al. (2022). Given the exploratory nature of the study, behavioral indicators derived from task performance (e.g., word production during narrative generation) were also examined as indirect markers of engagement and task involvement rather than as measures of usability.

For data analysis, statistical tools included SPSS (v. 28) and JASP version 0.9, with descriptive and inferential tests (e.g., one-way ANOVA, t-tests, and correlations, among others), while linguistic data were processed using Python libraries (spaCy, NLTK, pandas) to extract metrics such as lexical density, linguistic markers related to emotional expression (named emotional load), and computational indicators associated with creativity (named creativity). Text transcripts were normalized and analyzed with a pretrained Spanish spaCy model for tokenization, lemmatization, sentence segmentation, and part-of-speech tagging. Because the data involved child Spanish, automated output was supplemented with rule-based corrections and manual review for repetitions, truncated words, and other nonstandard forms. NLTK was used to calculate lexical statistics, and derived measures included lexical density, emotional load, and creativity, defined here as a computational proxy combining lexical diversity, low-frequency word use, and variability in word combinations.

2.3. Procedure

The study was approved by the Ethics Committee of competence (UCV/2023-2024/063). Sessions were conducted in the school library, with one researcher and one observer per group. Each session included two phases:

(i)

Individual written task: Each child rolled a die to draw category-based cards and created a story incorporating all cards. Stories were transcribed without orthographic correction.

(ii)

Group oral task: Participants constructed a collective story by turn-taking, with real-time transcription via dictation software.

Each session lasted approximately 30 min (written) and an additional 20–30 min (oral). Instructions were clearly explained, and assistance was provided only to resolve comprehension doubts.

The game materials were linguistically adapted using a back-translation method (Hambleton, 1993) from Portuguese to Spanish, followed by expert validation to ensure semantic, cultural, and psychological equivalence.

2.4. Data Analysis

A mixed-methods approach was applied to address the predefined research questions and hypotheses. Quantitative analyses focused on usability and linguistic performance, while qualitative analyses explored participants’ open-ended feedback to contextualize and complement the quantitative findings. Linguistic performance was operationalized across four dimensions derived from the hypotheses: morphology, including the use of adjectives and verbs; syntax, including sentence length and use of conjunctions; lexicon, including token count and lexical density; and pragmatics, including indicators related to emotional load and creativity.

Given the exploratory nature of the study, constructs such as creativity and emotional load were operationalized through computational linguistic indicators rather than direct measures. These indicators should therefore be interpreted as proxies associated with these constructs, rather than as definitive assessments. Although this approach allows for fine-grained and ecologically valid analysis of language production, its construct validity and generalizability remain limited and warrant further validation in future research.

Statistical analyses were conducted with consideration of the exploratory design and small sample size.

To test Hypothesis 1, which examined differences in linguistic performance between early and independent readers, independent-samples t-tests and Mann–Whitney U tests were conducted on individual narratives. Parametric tests were conducted because several variables showed approximately normal distributions, and these analyses allowed comparability with the previous literature. However, considering the limited sample size and the potential instability of distributional assumptions, complementary non-parametric analyses were also performed as a more conservative approach.

To address Hypothesis 2, concerning the relationship between linguistic features and creativity in group narratives, descriptive analyses were conducted to examine variations in creativity-related indices as a function of linguistic predictors. These analyses were complemented by Pearson correlations and effect size estimates to explore the strength and direction of the observed associations, in line with the hypothesized relationships. Given the exploratory nature of the study and the limited sample size, the findings should be interpreted cautiously, as these linguistic indicators represent computational proxies rather than direct measures of creativity.

3. Results

The results are organized into three main sections: (1) usability analysis, (2) individual narrative performance, and (3) group narrative performance.

3.1. Participants’ Perceptions of the Game

SUS perceptions of the board game were examined across three groups: students (n = 20), experts (n = 2), and teachers (n = 2). In terms of satisfaction, 73.9% of participants reported that they “liked it a lot,” 21.7% “liked it,” and 4.3% remained neutral. The most frequently mentioned emotional responses were fun (29.6%), nervousness (18.5%), joy (14.8%), excitement (11.1%), and concentration (11.1%). Regarding session length, 52.0% found it adequate, 44.0% short, and 4.0% long (average estimated duration: ~30 min). The activity was rated as “very valid” by 66.7% and “valid” by 33.3% of participants.

Qualitative feedback showed high enthusiasm among students (e.g., “I loved it,” “I want to play more”) alongside some reported difficulties in maintaining narrative coherence. Teachers highlighted the game’s capacity to assess verbal fluency and emphasized the need for clear instructions. Experts noted its usefulness for evaluating coherence, planning, and verbal comprehension and suggested adding verb-tense-switching cards and multiple difficulty levels. All groups recognized the tool’s utility for developing both oral and written expression.

3.2. Individual Narrative Production

Students produced a mean of 35.35 words (SD = 2.08), experts 34.50 (SD = 0.71), and teachers 35.50 (SD = 4.95). Before conducting independent-samples t-tests, assumptions of normality (Shapiro–Wilk), homogeneity of variances (Levene’s test), and independence were examined. However, due to the limited sample size, these inferential analyses should be interpreted with caution.

As presented in

Table 2. Descriptive statistics of linguistic components in individual narratives.

Component	Group 1 (M)	Group 1 (SD)	Group 2 (M)	Group 2 (SD)
Tokens	85.50	37.89	294.00	95.13
Adjectives	4.70	2.26	16.00	5.44
Verbs	14.50	5.99	40.70	10.13
Conjunctions	4.70	1.83	10.20	1.62
False Starts	2.10	0.74	3.80	1.03
Emotion	2.50	0.71	3.60	0.84
Creativity	4.70	1.16	5.70	1.06
Word Frequency	1.13	0.05	1.25	0.07
Lexical Density	0.573	0.015	0.564	0.008
Sentences	5.40	2.01	12.70	3.59
Words per Sentence	15.60	2.32	22.90	2.23
Subordinate Clauses	1.80	1.03	4.40	0.97

, significant differences emerged across most linguistic variables (p < 0.001), including tokens, adjectives, verbs, conjunctions, and sentence count, suggesting differences in narrative complexity and linguistic production between groups. To reinforce the stability of these findings, complementary non-parametric analyses using the Mann–Whitney U test were also performed, yielding largely convergent results. The creativity-related index approached statistical significance (W = 28, p = 0.08), whereas lexical density did not reach significance (p = 0.147).

Correlation analyses (Appendix A) showed that most linguistic features were significantly interrelated, with lexical density as an exception, only correlating with word frequency. When disaggregated by group, correlations weakened or disappeared, suggesting that group membership influenced the observed patterns.

3.3. Group Narrative Production

Figure 2 shows linguistic components by grade (3rd vs. 6th). Sixth-grade students demonstrated greater complexity, particularly in verb use, subordination, sentence length, and conjunctions.

Third-grade students produced shorter, syntactically simpler narratives, with varying emotional content. Narratives collaboratively produced by groups were analyzed according to classroom and grade level. Group 6 consistently generated the most complex and extensive productions, with the highest counts for adjectives, verbs, conjunctions, sentence length, and subordinate clauses. Group 3 also performed well in creativity but demonstrated lower syntactic complexity.

4. Discussion

The findings of this exploratory study contribute to the growing body of literature supporting the integration of gamified approaches into educational and therapeutic contexts. The adapted version of Érase una vez, examined here as a pilot study, appears to provide a structured yet flexible environment that may facilitate narrative production in school-aged children. Grounded in contemporary frameworks that emphasize the importance of embedded–explicit instruction for language development (Gillam et al., 2014; Pinto et al., 2018), the present work offers preliminary evidence supporting the role of narrative-based interventions in both classroom and therapeutic settings (Lindgren, 2023).

One of the most salient aspects of the findings is the game’s capacity to engage students and encourage spontaneous verbal production. Its playful nature, combined with narrative prompts and a structured yet adaptable format, may reduce performance-related anxiety often associated with formal language tasks (Rubegni & Paolini, 2010), thereby fostering more naturalistic language use. This is consistent with previous research suggesting that game-based environments can elicit complex discourse and support more ecologically valid assessments of language abilities (Botting, 2002; Demir et al., 2013; Spencer & Petersen, 2020). The high levels of reported enjoyment and willingness to continue using the game further support its potential as a motivating context for sustained linguistic interaction (Li et al., 2023; Rizvić et al., 2020).

Furthermore, the game’s multi-role structure, in which participants alternate between storytelling, listening, and evaluative roles, may promote cognitive immersion and support narrative construction processes. This aligns with principles of transformational play and immersive learning (Barab et al., 2010; Hwang et al., 2025), as well as with models of game-based learning that emphasize the integration of narrative, mechanics, and task structure to enhance both engagement and learning outcomes (Chapman & Rich, 2018; Lane & Prestopnik, 2017).

The observed differences between emerging readers (8–9 years) and independent readers (11–12 years) across multiple linguistic indicators, including lexical production, syntactic complexity, and narrative structure, suggest that the task may be sensitive to developmental variation in language abilities. These findings provide preliminary support for the study hypotheses and are consistent with theoretical accounts linking verbal fluency and narrative competence to underlying cognitive and executive processes. From a professional perspective, educators and language specialists perceived Érase una vez as a flexible and engaging tool with potential applications in both assessment and intervention contexts. Similarly, psychologists and speech-language specialists involved in the study highlighted its potential usefulness for identifying difficulties in planning, coherence, and syntactic organization, in line with previous work emphasizing the importance of ecological validity in narrative assessment (Amorim et al., 2020; Kalogiannakis et al., 2021).

With regard to the linguistic indicators associated with creativity and emotional expression, the findings should be interpreted cautiously. In this study, these constructs were operationalized through computational linguistic features, such as lexical diversity, variability in word usage, and affect-related language patterns. These indicators were intended as exploratory proxies derived from observable narrative characteristics rather than as validated psychometric measures of creativity or emotional functioning. Accordingly, the observed results are better understood as reflecting variations in expressive language use and linguistic complexity that may be related to creative or affective dimensions, rather than as direct evidence of underlying cognitive or emotional processes. Further research using larger samples and validated external measures is necessary to examine the robustness, validity, and generalizability of these indicators. In addition, the game appears to support multimodal, collaborative, and challenge-based learning processes (Kaya & Erçağ, 2023; Mahzan et al., 2020), which are known to foster intrinsic motivation, a key factor in language acquisition. Its adaptable structure allows for application across different levels and contexts, potentially supporting both academic and socio-emotional development through shared storytelling experiences (Arán Filippetti, 2011).

Nevertheless, several limitations must be acknowledged. The small sample size (n = 20), together with the homogeneous and non-random selection procedure, limits the generalizability of the findings and may reduce variability in performance. The cross-sectional design also precludes conclusions regarding long-term effects. Moreover, the reliance on computational proxies for complex constructs such as creativity and emotional expression introduces important constraints related to construct validity. Statistical results should be interpreted cautiously, as small samples may lead to unstable estimates and potentially inflated effect sizes.

Future research should address these limitations by including larger and more diverse samples, adopting longitudinal designs, and incorporating validated measures of language, creativity, and emotional processing (e.g., phonological awareness, processing speed, comprehension). Further development could also involve digital adaptations of the game, the integration of adaptive features, and the use of more advanced linguistic analyses, such as transformer-based natural language processing models, particularly in clinical populations (e.g., ADHD, ASD, DLD).

In conclusion, the adapted version of Érase una vez represents a promising exploratory tool for eliciting and analyzing narrative production in children. While the present findings provide preliminary support for its potential educational and therapeutic applications, they should be interpreted as an initial step toward the development of more robust and validated approaches to language assessment within gamified environments.