1 Introduction

Deaf users face many barriers related to written language on the Web. For example, the majority of deaf people have difficulties to reading long and complex text [6]. Many studies of deaf education consider sign language as the first and highest proficiency language, and a written language as a second language. Thus, many deaf people read word by word, trying to understand their meanings by associating each word with known signs. With respect to design, there are also several difficulties for designers in making the content accessible in all situations, such as to produce clear and simple texts, as suggested by W3C [11].

Our previous studies investigated the Web accessibility requirements for people with hearing loss [3], as well as the use of computational resources on bilingual deaf literacy [2]. In this paper, we focus on design alternatives to encourage deaf users to learn written language while they use the Web. The article presents thirty-six recommendations to deal with the difficulties and requirements previously elicited. The recommendations were structured according to common usage situations in which the deaf users reported difficulties. These situations include specific barriers to the use of synchronous communication tools (e.g., to send a message in an instant message tool) and use of asynchronous communication (e.g., to post a message in a blog).

We performed a qualitative study with 15 deaf students and 2 teachers with the objective of analyzing these recommendations. The participants had various educational levels (from primary school to university graduate) and their ages ranged from 12 to 30 years old. All the students had high levels of hearing loss, were fluent in Libras (Brazilian Sign Language), and consider Portuguese (Brazilian official written language) as a second language. An interpreter and a teacher with many years of experience in teaching sign and written languages also participated in the study.

This paper presents and discusses recommendations, prototypes and results of the evaluation practices with the users. An analysis of the results points out limitations of the approach, as well as challenges from theoretical and technological perspectives. We intend to contribute with a better understanding of Web accessibility for deaf users and with design alternatives to stimulate written language learning.

The work is structured as follows: Sect. 2 presents the theoretical background and related work; Sect. 3 describes the recommendations, prototypes and practices with deaf users; Sect. 4 presents the interpretation of the results, discussions and limitations; and, Sect. 5 concludes the paper.

2 Theoretical Background and Related Work

This section begins by presenting concepts related to deaf culture and the importance of sign language as an support for written language learning (Sect. 2.1). It subsequently addresses Web accessibility and Web tools for deaf users (Sect. 2.2).

2.1 Deaf Culture and Bilingual Literacy

The literature reports many challenges related to the deaf social integration of the deaf, along with the full acceptance of their rights and duties to participate effectively in society, as well as fully exercise their citizenship. Even when immersed in a largely oral culture, deaf people have their own community, culture and language. Several studies (e.g., Chomsky [1]) point out that sign language has a crucial role in the intellectual development of the deaf.

Our study is aligned with the idea that a written language can be learned independently of the oral language that it represents [10]. Thus, when we introduce deaf learners to a written language, first it is necessary to situate the learners in a context making use of their first language, i.e., sign language [9]. In line with other studies (e.g., Sánchez [8]), we assume the necessity of bilingual contextualization and the promotion of written language literacy instrumented by sign language.

One of the challenges in deaf students’ education is how to create opportunities that go beyond meaningful literacy experiences in the written language, by constituting a bilingual situation, in which the sign language is dominant in the situations of enunciation [4]. The written language, as a second language, supports the expansion of their social relations and appropriation of cultural elements of hearing people.

Nowadays, learning to read is vital for all individuals in society, and even more vital for the deaf [5]. However, there is still a lack of methods and pedagogical practices to teach written language in bilingual classes [13]. This limitation of bilingual educational practices is not restricted to the context of Brazil (or developing countries). The study of Marschark and Harris [7], for instance, highlights that deaf students of secondary schools in the United States were able to read, on average at the same level as 8 to 9 years old students. Despite advances in the field, other studies (e.g., Wauters et al. [12]) also point out challenges in deaf literacy education. In this context, the use of computational resources is a promising alternative for supporting and improving deaf literacy practices [3].

2.2 Technology and Web Accessibility for the Deaf

The use of techniques that explore virtual and Web environments may stimulate deaf students in bilingual literacy process. According to studies of Lorenzet [6], the use of interactive resources with Libras in virtual environments brings benefits for Portuguese language learning. These resources make the learning process pleasant and easier for the deaf learners. Her study also points out that the difficulties related to vocabulary and meanings can be minimized by including a sign language dictionary application. The association of written terms with sign language explanations made for the learning process more dynamic.

Various barriers that deaf users face when accessing the Web are related to the “Internet language”. They include, for example, the use of foreign language terms, technical concepts, complex text and Internet slang [2]. A deaf user with low literacy typically reads one word at a time; she/he tries to understand the meaning by associating each word with a known sign. This results interferes with the comprehension of long and complex texts. The W3C standards for accessibility [11] emphasize this aspect by suggesting that text should be written in a simple and clear way. However, it is problematic to put this recommendation into practice.

Nowadays, there is a vast literature about web accessibility for the deaf. The literature includes the research of new technologies (e.g., tools for automatic translation from written language to sign language, and tools for text simplification), as well as design aspects of how to make use of the new technologies. The focus of our work is on how to use the Web as an instrument for written language learning. We aim to produce an inclusive process, which supports deaf users’ interpretation of complex content and promotes their long term autonomy as web content consumers and producers.

3 Recommendations, Prototypes and Practices

This section starts with presenting 36 recommendations for inclusive design (Subsect.3.1). After that, Subsect.3.2 presents examples of interface prototypes constructed according to these recommendations, and Subsect.3.3 describes practices with deaf users that evaluated the proposed prototypes.

3.1 Web Design Recommendations for Encouraging the Learning of Written Language

The recommendations presented in this section were previously elicited with 29 deaf students [1]. The studies took place in the Deaf Service Center in the city of Macapá in Brazil - CAS (Centro de Atendimento ao Surdo), and the Deaf Mission (in Libras) of the Baptist Church of Macapá. The studies included interviews, semantic and norm analysis followed by participatory solutions proposals as presented in [1]. The studies resulted in 121 high level design recommendations, of which 36 concerns chat and blog writing activities, the focus in this paper. These recommendations were later discussed with a new group of deaf users in the context of this work.

According to the participants, the main difficulty concerning reading is the use/comprehension of long text with unusual terms. The users suggested including visual dictionary applications with avatars. By using these applications, deaf users can associate and learn new meanings by accessing explanations in sign language. In addition, examples of phrases in written language can also be proposed to contextualize the use of the difficult words.

Thus, we recommended the design of dictionary applications with explanations in sign language, visual resources, and examples in written language to explain and contextualize the use of the terms. The written problems should also be supported by explanations in sign language.

Table 1 describes recommendations and solutions for synchronous (e.g., online messages) and asynchronous (e.g., blogs) communication systems. A total of 36 recommendations were elicited with the users. The recommendations include the main grammatical rules and alternatives for supporting the learning of how to avoid written errors.

Table 1. Chat and blog design recommendations
Full size table

3.2 Prototypes Designed According to the Recommendations

Figure 1 presents a low fidelity prototype that illustrates a solution to be used in synchronous communication artifacts. This solution proposes the inclusion of explanations about written errors. As in a grammar check tool, the explanations are linked to errors underlined in red. An important component is that the explanations use clear texts with terms that are well known by deaf users. In this prototype avatars and images are a second option, to be used when the user has difficulties understanding the explanation in written language. This prototype is in line with the recommendations in Table 1 (#1, #8, #10 and #17).

Fig. 1.
figure 1

Example of low fidelity prototype with links to explain grammar errors

Full size image

Figure 2 presents a prototype with an avatar that gives explanations in sign language next to a textual description (recommendation 34 and 36). In this example, there is an explanation of how to use grammar punctuation rules.

Fig. 2.
figure 2

Example of an avatar used to explain the use of grammar punctuation

Full size image

Figure 3 presents a prototype with solutions to support deaf users to produce text in a blog. In this prototype, hints were included for supporting the designers in the modelling of functionalities to help deaf users compose blog posts. For example, the designer can include a dictionary option in the task bar (recommendation 18 and 33).

Fig. 3.
figure 3

Example of explanatory links to improve writing

Full size image

Figure 4 presents an example of the prototype to deal with polysemic words, popular expressions and uncommon expressions. This prototype implements the recommendation to highlight and provide explanations about polysemic words (recommendation 9), popular expressions and uncommon expressions. In the example of Fig. 1, a blog’ post makes use of the word “foot” (pé – in Portuguese) in different ways, for example: pé da serra (foot of the mountain- place), pé de pato (swimming fins – equipment) and pé de moleque (a kind of candy similar to peanut brittle). The words are highlighted by the prototype, and examples and explanations are provided using an avatar. The prototype provides explanations when the users read or produce texts, thereby, stimulating them to learn new ways of using the words and expressions.

Fig. 4.
figure 4

Example of prototype with an avatar exemplifying polysemic words

Full size image

3.3 Evaluating the Prototypes in Practices with Deaf Users

The studies took place in the Deaf Mission (in Libras) of the Baptist Church of Santana, Brazil. A teacher, an interpreter and 15 students participated in this study. Table 2 presents the basic profile of the participants in the study.

Table 2. Participants of the study
Full size table

Practice 1 (recommendations 9, 18 a 33)

Objective: This practice aimed to explore the meanings of the word foot (pé). The practice also included the following themes: figure of speech, informal language and text structure. The text reading activities were performed using the Blog editor of Fig. 4.

Procedure: The practice was performed in four steps as follows. In the first step, we presented a text without support of images and sign language interpreters. The objective was to verify how the users make sense of a text with various meanings of the word foot (pé). Table 3 summarizes central aspects of the first step of practice 1, including the goal, number of participants, number of sessions, number of students who performed the task correctly, and remarks about the students’ performance.

Table 3. Summary of the first step of the first practice
Full size table

During the second step, we presented a text with images and an avatar that explained the various meaning of the word foot (pé). The overall performance of the students on text comprehension was better as compared with the first step, however, some students also demanded support of the interpreters. Table 4 summarizes central aspects of the second step of practice 1, including the goal, number of participants, number of sessions, number of students who performed the task correctly, and remarks about the students’ performance.

Table 4. Summary of the second step of the first practice
Full size table

The third step of practice 1 included the application of an individual questionnaire with open questions: “Did you understand the text?”, “Did you know these meanings previously?” and “Did you like this practice?”. In addition, we asked them to give suggestions of how to improve the prototype.

During the fourth step, each individual evaluation was analyzed and discussed in group; each participant presented his/her suggestions, problems and possible solutions to improve the prototype.

Practice 2 (recommendations 1, 8, 10 and 17)

Objective: This practice aimed to investigate alternatives of how to perform a synchronous communications between deaf-deaf and deaf-hearer using instant message tools. This practice was performed in pairs, with the objective of making an appointment, including the day, hour and place, as well as to plan an agenda for this appointment.

Procedure: The practice was performed in three steps as follows. In the first step, the teacher and interpreter presented the objectives of the practice and the instant message tool. They also performed an example of how to execute the task (i.e., make an appointment using an instant message tool). Immediately after, they asked the deaf students to perform the task in pairs. The majority of the participants were able to establish a successful communication and make an appointment as requested. Few students had problems related to the technology (i.e., in the use an instant message tool) (Table 5).

Table 5. Summary of the second step of the first practice
Full size table

In the second step, we asked to the students to make an appointment with a hearing colleague. As in the step one, this task included to determine the place, date, and agenda (Table 6).

Table 6. Summary of the second step of the second practice.
Full size table

In the third step of practice 2, we asked each student to evaluate the prototype individually, indicating problems and giving suggestions of what could be improved in the prototype.

4 Interpretation of the Results, Discussion and Limitations

The results confirmed previous studies about the difficulties of deaf users to interpret, to produce, and to communicate on the Web. The results indicate that the proposed recommendation can improve the accessibility of Web applications; however, they also highlight some limitations of the proposed recommendations and prototypes.

In general, we highlight the following results from these practices:

  1. 1.

    Deaf students had difficulties to interpret and use figure of speech, even with the use of exemplifying images. In some cases, they needed the support of a interpreter to complete the task;

  2. 2.

    In some cases, the explanations using written language were not sufficient. During the discussions some students emphasized the limitations of these explanations for the comprehension of complex texts. The students also acknowledged that written explanations are less efficient than explanations using sign language, yet written explanation potentially stimulate the learning of written language;

  3. 3.

    They considered the Portuguese grammar rules too complex to be explained using short warnings. Longer explanations in sign language with examples in written language are needed in some cases;

  4. 4.

    According to the participants, design solutions for deaf accessibility can include various images to support a better understanding of the textual content, nevertheless additional explanation in sign language is necessary for a better understanding of the complex terms;

  5. 5.

    The participants suggested that instant message applications could include “animated emoticons” in order to support the direct correspondence of sign language with the most commonly used words. According to the participants, this feature may improve the communication between deaf users, as well as help hearing people to learn basic signs.

The practices with the students highlighted that, in some cases, the explanations in written language and sign language (using avatars) are not sufficient to learn new concepts. In such cases, interpreters and teachers have to provide additional explanations according to the individual difficulties and the students’ previous knowledge. This confirms the central role of educators in deaf literacy and the role of the technology as supporting artifacts. The results also point out possibilities of future research on generation of dynamic explanation based on the students’ difficulties.

This study is limited in number of participants; consequently, the results are not statistically conclusive. Particularly, the communication between deaf students and researchers is limited by the proficiency of the researchers in sign language or the number of interpreters. Consequently, studies with a lager numbers of students demand undue resources. Despite of the size limitations, this study contributed with preliminary evidence for the value of the use of visual resources (images and avatars) in design solutions for encouraging the learning of written language by deaf students.

5 Conclusion and Future Work

The written language is one of the most difficult barriers to Web accessibility for deaf users. Despite of the advances in the field, deaf literacy is still an educational research challenger. In this paper, we presented a set of web recommendations, prototypes and practices with users aiming to investigate design alternatives to improve Web accessibility. The objective was to use Web resources to stimulate deaf literacy and autonomy on the web.

Results from practices with deaf students pointed out the potential of associating images and avatars in synchronous and asynchronous communication artifacts, while it also identified difficulties and the central role of the teachers on the process. In addition, the results highlighted design implications and future research activities. As the next steps of this research, we propose to improve the prototypes, use them in longer studies (e.g., an entire semester), and evaluate their use in daily activities. The future work also includes research for the automatic generation of personalized explanations according to the individual needs and context.