Uncovering the Needs for a Hybridized Interaction Design Model for Sign Language Learning Through Experts’ Feedback

Nielsen’s and Molich’s design guidelines have been famously adapted in studies related to the design of interactive products, including for learning environment, and for hearing-impaired learners. In current situation, the learning materials for hearing-impaired learners are lacking in terms of positive interactions that promote two-way communication, partly because they do not apply appropriate design principles appropriately for the users. This affects their stimulation in learning activities. Hence, this paper aims at uncovering the current design principles applied in learning materials for learning sign language for the hearing-impaired learners. Besides an elicitation of literature, a semi-structured interview with the experts in hearing-impaired curricula has been carried out. It reveals that a study needs to urgently be carried out in proposing a heuristics design model that is specifically able to evoke the positive learning experience.


Introduction
The number of people with disabilities (PWDs) has been increasing rapidly around the world. The statistics from [1] reveal that 15% of the world's population have some forms of disabilities; in which one-third of them are children below 15 years old [2]. It is estimated that over 1 million children with disabilities are born every year worldwide [3]. Many of them are not well-treated by their parents because it involves a long-term medical engagement, which they are not able to afford for, especially in developing countries [3]. They have different types of disabilities including physical, learning, hearing, and visual. Among those various types of disabilities, hearing-impaired including deaf is considered the serious one. [1] reports that 285 million people in the world are hearing-impaired and deaf, in which 90% of them live in developing countries. Meanwhile, in Malaysia, the Malaysian Social Welfare Department reports that the officially registered disabled people as at December 2018 are 472, 228 people [4]. From that figure, 35,273 were hearing-impaired, increased 2,423 people from 32,850 in December 2008. Referring to the facts in Table 1, it could be deduced that from the year 2008 until 2018, the registered hearing-impaired people in Malaysia drastically increased to be double. Therefore, a study that focuses them as the main subject should be brought forward to ensure they are able to contribute to the nation building with the agenda of digital society 5.0 similarly with general people.  [5]. Social innovation in digital society 5.0 is looking forward at a society that is able to break down the existing sense of stagnation, a society that respects others, a society that is able to go beyond the limit, and a society that its' every member is able to lead an active and enjoyable life. In digital society 5.0, new values are created through innovation [6]. The value of new innovation will bridge the gaps in various aspects like region, age, gender, as well as language to meet desired needs [6]. In such kind of aspiration, the hearing-impaired and deaf people is not an exemption. Therefore, this study is part of the digital society 5.0 agenda, particularly in enabling the hearing-impaired and deaf people have similar accessibilities with the general people. Language should not be the barrier for them to express their feelings, emotions, and thoughts. This could be achieved through innovation particularly in content development of Assistive Technology.

Advancement of Assistive Technology (AT) towards digital society 5.0
Initiatives in assisting the People with Disabilities (PWDs) including hearing-impaired and deaf learners in using Assistive Technologies (AT) have been growing in most developing countries. AT can be categorized into hardware (i.e. assistive listening devices (ALDs) and software (i.e. text captioning on a screen or monitor). The advancement of AT has triggered meaningful impacts on the various aspects of PWDs life, as seen in smartphones and tablets [7]. In line with that, the field of education has received large impacts by AT, including in formal and informal education. As an example, primary schools that offer special education equip their computer labs and classrooms with assistive listening devices, frequency modulation, and personal amplification. All these AT are utilized as part of the teaching devices to support and improve the disabled learners' learning activities in their educational setting. However, most of such technologies focus on the technical aspects, which is difficult for hearing-impaired and deaf learners (in primary schools) to operate on their own. This requires an assistant to play roles appropriately. In contrast, AT that focuses on digital content that stimulates hearing-impaired learners to have active learning experience similarly with general students

Sign Language
Sign language is a gesture-based body language that uses a manual correspondence of non-verbal communication to transmit meaning. It involves the body movement including movement of hand and facial expression to express information from one person to another [10]. It is also called as oral language, which is specifically developed for people with hearing-impaired and deaf problem. In this study, sign language is the main content of Hybrid Interaction Design Model. Prior to developing the intended model, an initial stage of investigation related to the needs and the availability of Hybrid Interaction Design Model has to be conducted. To achieve that, an elicitation of literatures and usercentered design approach has been utilized throughout the study as elaborated in the next section.

Materials and Method
There are two phases involved in this study which are (i) phase oneelicitation of literature and (ii) phase twouser centered design approach. In the first phase, 10 existing sign language design models and 10 existing applications of Nielsen's Design Guidelines have been reviewed through a systematic literature review approach. The credible online database which are Scopus, ScienceDirect, IEEE, WoS, Emerald, and EBSCO Host have been utilized for the searching process. Meanwhile in the second phase user-centered design approach has been adapted to complete the study. Qualitative is the appropriate approach as the main target users of this study is hearing-impaired and deaf learners. In phase two, semi-structured interview has been conducted with experts in one Special Education Primary School (Hearing and Deaf Learners) located at Northern of Malaysia. Fig. 1 illustrates the summary of research activities involved in this study.

Result Analysis and Discussion
This section analyzes and discusses the results gathered throughout this study. They are discusses based on the research objective and research activities as highlighted in the previous section, which are divided into three (3) subsection; (i) analysis on the existing sign language design models, (ii) analysis on the applications of Nielsen's Design Guidelines, and (ii) analysis on the semi-structured interview.

Analysis on the existing Sign Languages design models
Sign Language is a core subject, compulsory for all hearing-impaired and deaf learners since they are in pre-school. In the era of digital society 5.0 sign language becomes more important for them as it is the only communication language among them. Learning Sign Language requires the hearingimpaired and deaf learners to have similar cognitive ability like general learners. However, their constraint to fully utilize their hearing organ making them difficult to understand and remember the contents [11]. These difficulties disrupts their stimulation and interaction process [12]. Accordingly, various models have been carried out to assist hearing-impaired and deaf learners in learning (listed in Table 2). Unfortunately, the interaction components in most of the existing Sign Language Models is treated with minimum focus and the features are insufficiently explored to make it sensible to the hearing-impaired and deaf learners. Also, it was found that most of the existing Sign Language design models fail to hybrid the Nielsen and Molich's Design models, which is highly potential in evoking the cognitive ability and positive interaction between the hearing-impaired and deaf learners with their teachers. Hence, it ought to be noted that this is the research gap that should be the focal point of this study.

Analysis on the existing applications of Nielsen's Design Guidelines
There are many heuristics for design model developed by researchers. Most of the sets of heuristics design model aim to enhance the interface object particularly to fulfil the user needs [23]. This is to ensure the positive and satisfying learning experience among the target users. There are eight design principles for heuristics design model which are (i) simple and natural dialogue, (ii) speak the users' language, (iii) minimize user's memory load, (iv) be consistent, (v) provide feedback, (vi) provide clearly marked exits, (vii) provide shortcuts, and (viii) deal with errors in a positive and helpful manner [24]. However, literatures show that most of the design model particularly in producing digital learning content utilizing the heuristics design principles at the evaluation phase [25] [26] [27] [28]. This means several iteration process is required at the evaluation phase, which is inappropriate for the design model particularly involving special-need users like hearing-impaired and deaf learners. Although iteration is still required at the evaluation phase, it has to begin at the initial phase of the study. Therefore, this study fills the gap by adapting the heuristics design principles at the initial stage of the study. All of this design principles will be adapted in the proposed model hybridizing with Nielsen and Molich's Design Guidelines. Meanwhile, there are nine general principles of interaction design, which has been developed and refined by Nielsen's and Molich's since 1990s. These 10 general principles are (i) visibility of the application status, (ii) match between application and actual world, (iii) user control and freedom, (iv) consistency and standard, (v) error prevention, (vi) recognition rather than recall, (vii) flexibility and efficiency of use, (viii) help users recognize, diagnose, and recover from errors, and (xv) help and documentation [29]. They are also called heuristics as they cover the broad perspective that appropriate with the development of digital content application and are customizable appropriately for the target users [30]. Table 3 demonstrates the digital learning content that adopts Nielsen's and Molich Design Guidelines, carried out in various contexts of study. Referring to Table 3 it indicates that hybridizing Nielsen and Molich's Design Guidelines into the design and development of sign language digital content is still insufficiently explored. This indicates that, Interaction Design Model for Sign Language hybridizing Nielsen and Molich's Design Guidelines is needed in ensuring it enhances the cognitive ability as well as stimulate the positive interactions of hearing-impaired and deaf learners during their learning activities. To strengthen the results gathered from the elicitation of the literature, a semi-structured interview has been conducted. It is discussed in the next sub-section.

Analysis on semi-structured interview
The interview session has been conducted involving experts, who are the Coordinators of Special Education Department. The interview was held in one of the primary schools that offers curricula for hearing-impaired and deaf learners. A set of semi-structured interview questions were asked ( Table 4). The types of questions were developed based on [41], which are classified into (i) knowledge, (ii) experience, (iii) opinion (iv) sensory, and (v) feeling. Based on their answers, it was found that currently there is no standard Sign Language model specifically designed or proposed by Ministry of Education to stimulate the hearing-impaired and deaf learners' learning interest in Sign Language particularly one that hybridizes Nielsen and Molich's Design Guidelines [Q1]. Also, it is reported that hearing-impaired and deaf learners have to learn by using the similar learning materials designed for general students [Q2]. As they have the limited hearing function, it is too difficult for them to adapt the general students' learning styles [Q2]. This makes many of them feel that Sign Language is difficult to learn. In regards to that, the pedagogical approach depends on the teachers' existing knowledge and creativity to ensure the hearing-impaired and deaf learners are able to grasp the delivered knowledge [Q3]. According to those experts, the best method for teachers to teach Sign Language to hearing-impaired and deaf learners is visualizing the scanned-images on screen [Q3]. However, this technique is still hard for the hearing-impaired and deaf learners to stimulate their learning optimally, to have positive interactions with teachers or even with their peers since scanned-images do not show step-by-step gestures [Q4] [Q9]. Another technique in current practice is one-way interaction, in which teachers have to speak more because of the limited learning materials particularly in the form of Interaction Design Model for Sign Language [Q3]. Again, this technique is hard to inspire the hearing-impaired and deaf learners to have a two-way interaction. O top of that, the experts also expressed that in Special Education Schools, Assistive Technology (AT) devices have been provided by Ministry of Higher Education to assist teachers in preparing the learning materials for hearing-impaired and deaf learners such as alerting devices [Q5]. However, the number of the AT devices is too limited. Besides, most of the time, the hearing-impaired and deaf learners still need teachers to assist them in using the AT devices. In fact, the situation becomes worse when the provided AT breaks down or needs maintenance. Learning Sign Language requires the students to understand, practice, and remember. However, as the hearing-impaired and deaf learners face limited hearing function, it is difficult for teachers to engage them in positive interactions in order to understand their problems as well as to measure their performance [42] [Q6]. This reflects that a comprehensive Interaction Design Model for Sign Language is necessary, to stimulate the interest of the hearing-impaired and deaf learners in learning Sign Language. The difficulties they face in their learning activities indicate that the need for the model is urgent [Q7]. In fact, it is expected that the model is able to ensure hearing-impaired and deaf learners are facilitated, enjoyed, and motivated to stay focus in the learning activities similarly with general students. Thus, the Interaction Design Model for Sign Language should conceptually address the main learning needs that are able to attract and sustain their attention in their learning activities [Q8]. The intended model should also be able to offer positive learning experience in supporting the hearing-impaired and deaf learners interactively and attractively with enjoyment and encouragement concept [43] [Q10]. In ensuring that the students are able to absorb the learning content, they should actively involve in class activities [42]. So, this could be supported by offering unique interaction experience specifically