Introduction

Speech Sound Disorder (SSD) is one of the most common communication disorders among children, and 10–15% of preschool children experience it1.

Children with SSD have problems in perception, articulation/motor production, and phonological representation of speech segments (consonants and vowels), and also experience difficulty in phonotactics (syllable and word shapes) and prosody (lexical and grammatical tones, rhythm, stress, and intonation)2.

Surveys have shown that more than 40% of SLPs cases are children with SSD disorder3. If this disorder is not treated, children may face literacy, emotional, psychological, social, and occupational problems in the future during adulthood4,5,6.

This disorder must be recognized and treated before children enter the school to avoid the abovementioned issues. However, access to treatment is not always possible for patients. Several factors may prevent patients from benefiting from appropriate treatment, including unavailability or lack of service providers such as SLPs in some areas, long duration of treatment and rehabilitation process, high costs of speech therapy sessions, long distance between the patient residence and rehabilitation centers, and fatigue of patients during the trip to centers7.

On the other hand, it may not be possible to completely treat speech disorders with several therapy sessions in rehabilitation centers. At the same time, the parents could play an active role in their children’s treatment process and help them with the exercises given by SLPs at home to speed up their recovery. However, performing therapeutic sessions at home may also cause some challenges, including: (1) if the exercises are not attractive and fun, they will be left half-finished, and (2) children may need to receive appropriate feedback on how to do the exercise correctly. As a result, the treatment process may not be appropriately completed8.

Various technologies can be used to solve the mentioned challenges. Today, various technologies are used in different fields of health, such as telemedicine, registry, digital games, telehealth, virtual reality, and augmented reality9,10,11,12,13,14.

One of the technologies that can help solve problems in the field of speech therapy is the use of digital games. By providing an attractive environment, they can increase children's motivation to continue the treatment and provide appropriate feedback during the treatment process. Today, games are used in various fields of medicine, and many advantages and disadvantages have been mentioned.

The drawbacks and negative results of using games in healthcare include (1) the potential of these games to cause anxiety in some people, (2) loss of game-related features such as enjoyment and motivational capacity due to excessive attention to the serious contents of these games, (3) lack of proper design of games, (4) time-consuming process of determining scenarios and game design, and (5) lack of compatibility of the game levels with the capabilities of the target group. Despite these disadvantages, studies have also investigated the applicability and benefits of the games in speech therapy15.

A systematic review study that investigated the use of digital games in children's speech therapy revealed that this technology could improve satisfaction and increase the motivation and attention of children in the speech therapy process11. This study also showed that 48.15% of the games could be used at home under the supervision of parents11. Therefore, the present study aimed to design a digital smartphone-based game to help treat children with SSD at home.

This article consists of four sections. The first section introduces the topic and motivates the research problem. In the second section, the game development method is explained, in which the game document, the game's design and development, and the game's evaluation are presented. The game design and evaluation results are given in the third section. The fourth section also discusses the main results of the developed game.

Methods

In various sources, different stages have been considered for the development process of serious games, but based on the System Development Life Cycle, four stages of Analysis, Design, Development, and Evaluation can be assumed16,17. Each of these four main stages includes the sub-stages given in Fig. 1. Each of these sub-steps is described in detail below16.

Figure 1
figure 1

System development life cycle for games.

Full size image

During the game analysis stage, various activities are carried out to define the objectives, requirements, and content of the game. It involves identifying any existing issues and establishing the purpose of developing the game. Subsequently, the necessary content, requirements, and tools are determined based on these objectives. The game design process, as the second step, is an iterative one that commences with the development of a game design document. This document outlines crucial elements like mechanics, characters, storyline, and music. Following the completion of the document, a game prototype is crafted, and any identified issues are resolved. This iterative approach ensures that the game design is continually refined, focusing on enhancing the User Experience (UX).

In the next step, which is the development of the game, game programming is performed, and the developed sample is tested to determine its quality and identify and solve problems. Moreover, during this stage, if each part of the game is developed separately, they are integrated.

The final step involves evaluating the game to assess whether it aligns with the established primary objectives. In this step, different evaluation methods can be used to test the game and get feedback from users. Additionally, essential activities in this stage involve maintaining the game to identify any issues, implement updates, and make improvements if necessary.

Study overview

This developmental study was conducted to design and evaluate the usability of a serious game in speech therapy. The target users for this digital game are preschool children with Specific Speech Disorders (SSD), aged between 3 and 6 years old. The designed game is a supplementary tool to help the process of therapeutic exercises at home under the supervision of parents. "Ava" is the name of the game designed in this study. The Ava game is designed in the Farsi language as a smartphone game on the Android platform (the most common available platform in Iran), which teaches consonants, syllables, words, and sentences across four levels.

It's important to note that all stages of this game were conducted under the guidance and supervision of a speech-language pathologist. This study was conducted in 3 stages: (1) clarifying information needs, (2) designing and developing the game, and (3) evaluating the usability of the designed game. The study comprises three principal stages, as elaborated below.

First stage: determining the information requirements

This stage aimed to identify the consonants, syllables, words, and sentences that should be included in the "Ava" game for children. To determine the consonants to be taught in this game, a survey was conducted, and SLPs were asked to decide based on the following:

(1) The children included in the study were aged between 3 and 6 years old, a period characterized by sequential changes in articulation development. (2) Developing a comprehensive tool that can be utilized by children between the ages of 3 and 6 years.

After determining the consonants, six syllables related to each consonant that should be taught to children were determined. Next, the words corresponding to each consonant were identified to ensure they were appropriate for children aged 3 to 6 years. Farsi literature and studies were reviewed at this stage to determine the words. Nonetheless, studies not conducted in the Farsi language were excluded from consideration, given that the game in this study was designed in Farsi. Therefore, related Farsi books in this field were reviewed, and the words suggested for the age range of 3–6 years were compiled in a questionnaire.

The extraction of words was conducted using two methods: (1) extracting words from Farsi books and texts, including the Farsi vocabulary book of the first grade in elementary school, Farsi Version of the Diagnostic Evaluation Test of Articulation and Phonology, Frequency Dictionary: According to a Written Corpus of Today's Farsi Language, Illustrated Dictionary of Consonants, and (2) Additional words were included in the questionnaire based on recommendations from speech-language pathologists. These words were selected according to the different speech and linguistic characteristics and levels, including phonological, phonetics, lexical, semantic, grammatical, word frequency, difficulty, and the ability to depict the words.

Specific linguistic and speech characteristics were taken into account when selecting words that would be suitable for preschool children. Acknowledging that children's speech and linguistic abilities vary within the 3 to 6 age range, a diverse array of words was selected to account for this variability during the development process.

In cases where a word proves to be developmentally challenging for a child, the therapist has the option to skip that word and proceed to the next one. Following the literature review, a questionnaire was designed and sent via email to SLPs from various regions across Iran. In the questionnaire, SLPs were asked to determine the appropriate words for the age group of 3–6 years. Also, an open-ended question was designed in the questionnaire, and the participants were asked to suggest any other words they considered necessary.

The questionnaire presented words for each consonant in three different positions: The words in the questionnaire for each consonant were divided into three positions: the beginning, middle, and end of the word. Each consonant had 30 words listed in the questionnaire (with 10 words for each position). The responses to the questionnaires were gathered through the Delphi method conducted in two rounds. Following the word selection process, five focus group sessions were held with five SLPs. During these sessions, the words determined for each consonant were given to the SLPs on a worksheet, and they were asked to make three sentences with the words determined for each consonant. During this phase, the sentences were initially created with the least semantic, grammatical, phonological, lexical, and phonetic complexities while also using the fewest number of words possible. To present sentences compatible with preschool children with different speech and language abilities, in the subsequent sentences, gradually, the difficulty level of the sentences increased. Furthermore, because producing consonant clusters is challenging in Farsi, initially, sentences containing words with consonant clusters were not used.

Second stage: designing and developing the Ava game

This stage was carried out in the following two main steps.

Writing the Game Design Document (GDD)

The game design document was written after determining the information requirements of the Ava game. Before starting to write the GDD, all the games in the field of speech therapy were examined by the research team, and the GDD was written based on the results of this examination.

This document consisted of six chapters, with a brief overview of the titles and contents of each chapter provided below:

Chapter 1: Introduction of the game

  • The goal of game development

  • Game style

  • Game character

  • Target group

  • Child's age

  • Game mood and platform

  • Game name

Chapter 2: Core mechanics

  • Core mechanics

  • The main tasks of the gamer in the game

Chapter 3: Sub-mechanics

  • Sub-mechanics of the game

Chapter 4: Assets

  • Sound

  • Music

  • Dialogues

  • Sound effects

  • Animations

  • Particle effects

Chapter 5: Progress tables and prizes

  • Prizes the child gets after completing a stage of the game

  • Progress table

Chapter 6: Description of each stage of the game

  • A detailed explanation for each stage

  • A list of required assets for each stage

  • A list of dialogues of each stage

  • A list of objectives for each stage

  • Result of each stage

After the GDD was written, it was reviewed and revised by SLPs, medical informatics specialists, graphic designers, and game developers.

Design and development of the Ava game

During this phase, the complete game design document was sketched on paper by a graphic artist. This step was taken to prevent redundancy while allowing the research team to modify the scenarios early in the process. The wireframe was drawn based on the designed sketches. Balsamiq Mockups software was used to design the wireframe. Then, with the help of Photoshop software, the environment, the character, and the components of the Ava game, as well as the animations and effects, were designed and developed.

The Ava game was developed using Java in the B4A programming environment.

The speech recognition system and its conversion to text (Speech to Text system) were used to teach the words and sentences, which was done with the Google voice recognition system.

In the system development phase, the C# language was used to organize, compress, and produce images, animations, and speech. The designed game was 2D and online.

Two methods for improving articulation in children with speech sound disorders were incorporated into the design of Ava's game. The prime used therapeutic approach was based on the Van Riper traditional treatment approach, in which the focus was to correct an individual phoneme and extend it into syllables, words, and sentences containing the aimed phoneme in different positions18, which were introduced to children at every stage through continuous auditory and visual stimuli. In addition, an associated operant conditioning approach, including auditory modeling and cueing after each incorrect articulation, and also visual and auditory positive reinforcements in correct productions in a regular way, were considered for the children.

These hierarchical treatment principles help the child, the family, and the therapist to proceed with the treatment systematically and with predetermined levels of difficulty. Also, the visual and auditory appeal of the game motivates the child to continue the treatment until learning the target sound in different sentences and generalizing it into daily conversations.

The accuracy of the productions was determined by the speech recognition system and its conversion to text (Speech to Text system).

To make the game more comprehensive, especially for children whose problem is mostly phonological, the second approach, which was a phonological intervention approach using minimal pairs, was also considered. During the initial and subsequent stages of the game, phonological awareness was visually taught to children through gameplay activities. Our purpose in the first two stages of the game was that children could distinguish differences between their incorrectly produced phoneme and the aimed correct phoneme and generalize it to the syllable level.

In an effort to improve children's phonological awareness, two mini-games were incorporated in the first two stages, utilizing both visual and auditory elements.

Third stage: usability evaluation of the designed game

The usability of the designed game was evaluated using the Nielsen checklist19, the PSSUQ standard questionnaire (Post-Study System Usability Questionnaire)20, and by observing and surveying children. The details of this evaluation are given below.

Heuristic evaluation with Nielsen's principles

The heuristic evaluation of the designed game was done by six experts using ten Nielsen’s principles. Table 1 shows the 10 principles evaluated.

Table 1 Nielsen’s usability heuristics.
Full size table

This evaluation was done in two stages: the navigation phase and the analysis phase. The experts engaged in gameplay during the navigation stage to become well-acquainted with the game's user interface. In the analysis phase, each expert initially conducted an independent comparison of the game's user interface with Nielsen's principles, marking the items that were not adhered to in the checklist. Then, the researcher examined the completed checklists, and similar problems were removed. A consolidated list of problems was compiled based on the individual assessments conducted by the experts. Finally, the severity of identified problems based on the frequency of risk exposure, the impact of the problem on the user experience, and the persistence of the problem were determined, with the severity range of 0 to 4 for each problem (0 = no problem, 1 = cosmetic, 2 = minor, 3 = major, and 4 = catastrophe). To determine the final severity of each identified problem, the severity of each problem was summed up and divided by the number of evaluators who had identified that problem.

Evaluation of the “Ava” game with the PSSUQ standard questionnaire

The PSSUQ standard questionnaire was used by speech-language pathologists to evaluate the designed game's usability. This questionnaire is used to check user satisfaction with 16 questions and three sub-scales:

  • Overall satisfaction: Questions 16

  • System Usefulness (SYSUSE): The mean scores of questions 1 to 6

  • Information Quality (INFOQUAL): The mean scores of questions 7 to 12

  • Interface Quality (INTERQUAL): The mean scores of questions 13 to 15

All questions in the questionnaire were scored from 1 to 7 based on the Likert scale (strongly agree to strongly disagree). This questionnaire had already been investigated for reliability, and the results show high internal consistency and an overall alpha coefficient of 0.9721. In this study, an evaluation using the PSSUQ questionnaire was conducted by 22 SLPs.

In order to compare the mean score of each of the sub-dimensions of the PSSUQ questionnaire between men and women and also at different educational levels, the normality of data distribution was checked using the Shapiro–Wilk test. The mean score of System Usefulness, Interface Quality, and Information Quality dimensions was normal in men and women. Independent sample t-test was used to compare the mean of these sub-dimensions. Mann–Whitney test was used to compare the mean score of question 16 between men and women due to the non-normality of the mean of this variable. The mean score of the System Usefulness, Interface Quality, and Information Quality sub-dimensions was not normal in the educational levels, and the Kruskal–Wallis test was used to compare the mean of these sub-dimensions.

Usability evaluation of the “Ava” game by children

Following the evaluation by experts and speech-language pathologists, the Ava game was provided to ten children with SSD, chosen through a convenient method, to assess the game's usability. In this evaluation, the children played the “Ava” game, and the evaluators observed and recorded the children's reactions. In addition to observing the children's reactions during the game, some questions were also asked of the children regarding the game's quality. Since the age range of the children was 3–6 years, it was not possible to use the Likert scale, and the children might not understand the scales. For this reason, emojis (, , , , ) were applied.

It should be noted that the purpose of carrying out this part of the study was only to evaluate the usability of the game, the reactions, and the opinions of children (as the primary users) about the quality of the game. Therefore, investigating the effect of the game on children's speech sound disorder is not included in the objectives of this study. Accordingly, during a 30–45-min treatment session for each child with SSD, the above-mentioned items regarding the game were evaluated.

Before children can use a game, it must be validated. A game can be validated for children in different ways:

  1. 1.

    Involving experts in the game development process.

  2. 2.

    Investigating the usability of the game by experts as well as children.

  3. 3.

    Considering the content of the game according to the age of the children.

  4. 4.

    Ensuring that the game content aligns with ethical and safety standards.

  5. 5.

    Validation of game content.

  6. 6.

    Assessment of game difficulty levels considering the age of the children and their individual abilities.

  7. 7.

    Ensuring that the game content is compatible with cultural sensitivities.

As explained in different stages of game development, we considered the above in this study.

Ethics approval and consent to participate

The study was conducted in accordance with the Declaration of Helsinki and approved by a local ethics committee in Iran, namely Ethics Committee of the Tehran University of Medical Sciences (ethics approval number: IR.TUMS.SPH.REC.1400.189). Verbal informed consent obtained from all the participants included in the study. Also parents/legal guardians gave informed consent for children to participate in the study.

Results

The results of the three stages of developing and evaluating the "Ava" game are detailed below:

Stage 1: determining information requirements

During the survey stage, Speech-Language Pathologists highlighted that the "Ava" game requires all 23 consonants of the Farsi language (/b/, /p/, /t/, /dʒ/, /tʃ/, /h/, /χ/, /d/, /r/, /z/, /ʒ/, /s/, /ʃ/, /f/, /ɢ/, /k/, /ɡ/, /l /, /m/, /n/, /v/, /j/, /ʔ/) for educational purposes, and none were omitted.

Also, six syllables were considered for each consonant by combination with the six vowels of the Farsi language; therefore, 138 syllables were considered. In the word section, a total of 564 words suitable for the age group of 3–6 years were determined based on a survey through the utilization of a questionnaire. Also, 30 SLPs with a mean age of 34.83 ± 7.75 years participated in the survey to determine the words. Most of the participants in the survey were women (N = 18, 60%).

Sentences were determined in five focus group sessions. The mean duration of the focus group sessions was 90 min. The mean age of the participants (5 SLPs) in the focus group sessions was 33.8 ± 5.97 years, and 60% (N = 3) were men. Three sentences were determined for each Farsi consonant. In total, 69 sentences were determined for the 23 consonants of the Farsi language. These 69 sentences were shown to children as animations in the 4th stage of the game, and they were asked to guess the sentence related to each consonant. In these focus group meetings, besides choosing the sentences, there was a discussion about how to illustrate each sentence as an animation. The sentences were sequenced from easy to difficult considering their semantic, grammatical, phonological, lexical, and phonetic complexities.

Stage 2: design and development of the Ava game

Sketches were created on paper according to the GDD. Then, wireframes were designed based on the drawn sketches. An example of the designed wireframes is shown in Fig. 2.

Figure 2
figure 2

Wireframe designed for the Ava game.

Full size image

Based on the designed wireframes, the Ava game was designed in four stages: teaching sounds, syllables, words, and sentences (Fig. 3).

Figure 3
figure 3

The main menu of the designed game.

Full size image

In the first stage (teaching the sounds), 23 consonants are taught to children and subsequently, a mini-game is integrated for additional practice. In the second stage (teaching syllables), children are first asked to choose the desired consonant, and then six syllables related to that consonant are taught. Apart from teaching, a mini-game is also designed to provide children with additional practice. (Fig. 4).

Figure 4
figure 4

Mini game designed to teach syllables.

Full size image

In the third stage, related to teaching and practicing words, we tried to teach about 30 words for each consonant. The target consonant was present in 10 words at the beginning, 10 in the middle, and 10 at the end. At this stage, the child is first asked by the broadcaster to choose the consonant, and after choosing the consonant, the broadcaster again asks the child to choose the position of the consonant in the word (beginning of the word, middle of the word, or end of the word). By choosing the position of the consonant in the word, the picture of the first word is shown to the child, and the broadcaster asks the child to guess the name of the word.

The child makes a guess and speaks the word by tapping on the microphone. Then, the speech recognition technology recognizes it, converts it to text, and matches it with the word name.

If the word is pronounced correctly, the honey will be given as a prize, and the child will be encouraged audibly. However, if the word is not pronounced correctly, the broadcaster says the correct word related to the picture and asks the child to repeat it. This process continues until the child can pronounce the correct word (Fig. 5).

Figure 5
figure 5

Teaching words in the third stage of the game with the help of pictures.

Full size image

To prevent disappointment, if children do not pronounce the words correctly, no negative points will be given to them, they will simply not receive honey. Also, at this stage, a golden star is given to children for one-third of the words that are pronounced correctly.

It should be noted that speech recognition accuracy was also checked in this game, and it was above 90%. When checking the speech recognition accuracy of the designed game, even the game with people with distortion, omission, and substitution disorders was examined. The designed game was able to recognize these cases as well successfully.

In the fourth stage, 69 sentences are taught to children in the form of animation. In this way, the desired sentence is first shown to the children in animation, and the children are asked to guess the correct sentence (Fig. 6). Like the third stage, speech recognition technology is used in this stage.

Figure 6
figure 6

Teaching sentences in the form of animation.

Full size image

In the first two stages of the Ava game, children become familiar with the consonants and syllables of the Farsi language, which helps their phonological awareness. In the next two stages, children practice the correct pronunciation of words and sentences. During these four stages, they receive positive auditory and visual feedback and progress gradually.

The game setting was created as a lush forest environment. A wagon was used to design the main menu, displaying different stages. A bee was chosen as the game's avatar, and in various stages of the game, honey and stars would be given to children as prizes if they completed the stages correctly. To avoid impacting the child's self-confidence, no points are deducted in the event of a mistake, and the child is simply asked to repeat the desired action. Also, since the users of this game are 3–6 years old children, the help system was designed in two ways: (1) written help for parents, which explains to them how to play the game by children, and (2) spoken help for children, which explains to them how to play the game at each stage.

Stage 3: evaluation of the designed game

Following the game's design, its usability was assessed in two distinct ways. (heuristic evaluation with Nielsen's principles and the PSSUQ questionnaire).

Table 2 provides demographic information about the participants in these two evaluations. In both evaluations, 50% of the evaluators were women, and 50% were men. The evaluation was done by 22 SLPs using the PSSUQ questionnaire. Still, the heuristic evaluation with Nielsen's principles was done by six experts in medical informatics, health information technology, and health information management who were familiar with heuristic evaluation.

Table 2 The demographic information of the participants who evaluated the Ava game.
Full size table

The results of the heuristic evaluation of the designed digital game with Nielsen's principles are given in Table 3.

Table 3 Determined usability problems per violated heuristics and severity.
Full size table

This evaluation led to the identification of 38 usability problems by six evaluators, which left 23 unique usability problems after removing duplicates. The "error prevention" principle had the highest severity, with a mean severity of 3.83. Also, the "user control and freedom" principle had the highest number of identified problems (5 problems). No problem was detected for the two principles of "match between system and the real world" and "flexibility and efficiency of use." About 70% of the identified problems were cosmetic and minor, and only 13% of the identified problems were catastrophic.

Table 4 shows the usability evaluation results of the designed game with the PSSUQ questionnaire. 22 SLPs did this evaluation with a mean age of 30.68 ± 7.65 years. Also, 15 (68.2%) of the evaluators had a bachelor's degree, 5 (22.7%) had a master's degree, and 2 (9.1%) had a PhD. The SLPs had a mean clinical experience of 5.45 years (SD = 5.36).

Table 4 The results of evaluating the usability of the designed game with the PSSUQ questionnaire.
Full size table

Since SSD is a common speech disorder with a high referral frequency to Speech-Language pathology clinics in Iran, and every SLP who starts clinical practice usually visits SSD cases so frequently, it can be said their clinical experience is equal to the experience of treating SSD patients and sufficient to do this study. The highest score obtained for the game was related to interface quality, with a mean score of 6.25 out of 7. The lowest mean score was related to information quality, with a mean score of 5.84 out of 7. The overall score obtained was 6.14 out of 7. Also, the mean score of SLPs' satisfaction with the game was 6.09 out of 7, which showed the therapists' high level of satisfaction with the designed game.

Table 5 compares the mean score of each sub-dimension of System Usefulness, Interface Quality, and Information Quality and the score of question 16 between two groups of men and women. According to these findings, the average score of none of the variables mentioned showed a difference between the two groups of men and women.

Table 5 Comparison of the mean score of the PSSUQ questionnaire dimensions between two groups of men and women.
Full size table

Table 6 compares the mean score of each sub-dimension of System Usefulness, Interface Quality, and Information Quality and the score of question 16 in different educational levels. Based on these results, the mean scores of the mentioned variables in various educational levels did not differ statistically significantly from each other.

Table 6 Comparison of the mean score of the PSSUQ questionnaire dimensions between different educational levels.
Full size table

The "Ava" game was evaluated using two observation methods and a survey. Ten children with SSD were surveyed, with a mean age of 4.6 years (SD = 1.17). Fifty percent of the children were female.

The results of observing children during the game showed that children liked the game's graphics, and the sound effects used in the game attracted their attention. Also, the user interface of the game, which was colorful, encouraged children to continue playing. In the sentences training phase, they liked the animations very much, and the desire to see the next animation encouraged more children to pronounce the sentence correctly. Also, the children did not get bored or cry during the game. Also, the game entertained the children during the treatment session. The feedback that was considered in the game was motivating.

The problems observed during the game playing by the children included the following: 3-year-old children could not go through the game stages alone, and their parents had to help them more. During the audio recording, the children did not know they must click on the microphone, and the broadcaster needed to give the instructions at the beginning of this step. Also, downloading the animations considered in the sentences training phase took more time because of the low internet speed in Iran. The results of the survey with children are given in Table 7. 100% of children stated that the game was designed beautifully, and 90% of children also liked the game.

Table 7 The results of the survey with children.
Full size table

Discussion

In this study, a smartphone-based digital game was designed for 3–6-year-old children with SSDs. This game included four stages of teaching the consonants, syllables, words, and sentences. In the game's design, an attempt was made to provide children with a comprehensive tool for speech therapy by considering all the consonants of the Farsi language (23 consonants). To determine the appropriate words, a survey was conducted among SLPs in Iran, and an attempt was made to use valid words suitable for teaching 3–6-year-oldchildren. Also, to determine the sentences, SLPs expressed their opinions in the focus group sessions.

A heuristic evaluation of this game revealed 23 usability issues, with approximately 70% categorized as minor and cosmetic problems. The SLPs also reported being satisfied with this tool and rated its usability at 6.09 out of 7. The evaluation of the game by children showed that children liked the user interface of the designed game, but three-year-old children had to play the game with the help of their parents.

A review of speech-language therapy games showed that some games have been designed for articulation and speech sound disorders8,22,23,24,25. Many of these games have also been designed in English, which is useless for non-English speaking children. The game designed in this study can be used for many Farsi-speaking children. In fact, due to the nature of speech disorders, non-English speaking countries are forced to create these games based on the official language of their country. Hence, the design of the mentioned game can be a model for the designers of speech therapy games in other countries.

Only a few games designed in speech-language therapy teach the four levels of consonants, syllables, words, and sentences. Some designed games only focus on teaching some consonants and not all of them26,27,28, and others cover two to three levels29,30,31. Unlike other games, the Ava game designed in this study is not limited to a few consonants, and all Farsi consonants are taught in this tool. Also, this game teaches children consonants, syllables, words, and sentences at four levels, so it can be suggested as an auxiliary tool by SLPs to use at home.

Nielsen's usability heuristics identified 23 unique problems in the game designed in the present study. The most significant identified problems were related to "user control and freedom". The reasons for this usability problem include the following: The Ava game does not allow canceling an action in progress, and the user cannot easily undo his/her actions. These findings are consistent with the results of an exploratory evaluation of a clinical decision support system by Hwayoung Cho32.

This study identified the "user control and freedom" principle as the problem needing correction32. In another study33, this principle was the most significant and common usability problem. To address this issue, it is advised that game designers provide the possibility of exiting different stages of the game at any time and to backtrack from actions in progress.

Among the identified problems, “error prevention" had the most severity, with a mean severity of 3.83. Two problems were identified regarding this principle, and the experts believed that correcting this problem is necessary and has a high priority. The experts gave this problem a high severity level because the length of each word and sentence was not specified in the teaching words and sentences phase. They believed that failure to observe this principle might increase children's mistakes in guessing words and sentences.

In the study of Farzandipour and colleagues34, who evaluated the heuristic of admission and medical records subsystems, this heuristic was identified as the “highest rate of problems.” Modifying this heuristic by designers can reduce the occurrence of possible errors. In the present study, where the users are 3–6-year-old children, paying attention to this principle can increase children's use of this game.

The results of the PSSUQ questionnaire also showed that the highest mean scores were related to the principle of interface quality. In this study, an attempt was made to use the right colors in the user interface design to provide a more attractive environment for children to play. Attempts were also made to design a simple user interface and display only the necessary information to not confuse children. Also, due to the young age of the Ava game users and their inability to read, we tried to communicate with the children and guide them during the game by listening and speaking. In this game, we also tried to pay careful attention to the user experience (UX) related to how the user interacts with the game.

During the graphic design phase, emphasis was placed on the precise and proper execution of the game logic, as it plays a crucial role in determining the user's interaction with the software. It also creates a sense of friendship and attracts the user's attention. The user may get bored or tired after using the software for a long time and doing speech therapy exercises. In this game, we tried to keep the audience in the environment longer and prevent fatigue and boredom by creating meaningful movements through an integrated and coordinated design.

One of the strengths of the present study is the survey of many SLPs and the holding of numerous focus group meetings in the phase of determining words and sentences. Also, in this study, we tried to identify the usability problems and determine the satisfaction of SLPs by evaluating the designed game using Nielsen's usability heuristics and the PSSUQ standard questionnaire.

Also, this study, like other studies35, has some limitations and threats to validity. One of the weaknesses of the designed game is that it is only available online, and the researchers had no choice but to use the speech recognition feature other than the online form. Another limitation of this study is that we did not use this game in a real clinical environment. For future studies, it is recommended that the effectiveness of the designed game in treating sound speech disorders is assessed.

Conclusion

This study introduces a serious smartphone-based digital game for children with speech sound disorders. This game was designed to teach consonants, syllables, words, and sentences in 4 stages of an interesting game under the supervision of parents. In other words, SLPs can use this game as an auxiliary tool for speech therapy at home under the supervision of parents. During the initial two stages, this game solely focuses on teaching the differentiation between phonemes (phonological awareness).

In the stage where words and sentences are taught, it has speech recognition technology that gives the necessary feedback to children during exercises. The proposed game was evaluated by technical experts and SLPs, both confirming the usability of the game. According to the satisfaction of SLPs with the designed game, it is recommended that this game be implemented and used in clinical environments. Due to its appealing and engaging design, this game has the potential to enhance adherence to treatment sessions at home.