Keywords

1 Introduction

Digitalization has changed the way we understand education. The professional practice of teaching any subject in any level of education has been reshaped to provide a more comfortable learning experience for some students. Online learning can be advantageous and offer more opportunities for students with disabilities, as some physical barriers are excluded and as some accessibility servings are more easily applied to a digital classroom (Arachchi et al., 2017; Barron et al., 2004; Kent, 2015, 2016; Salomoni et al., 2007). Nevertheless, the digital progress of educational environments has overlooked the need of some students, and online classrooms are often also not fully accessible or not designed with accessibility in mind (Arachchi et al., 2017; Roberts et al., 2011; Nganji, 2012).

This is the starting point of our study, which aims to review the accessibility of online learning environments, more specifically of educational videos, which are very commonly used in second-language learning. The focus of the study will be in analyzing the accessibility of videos aimed to adult non-native learners of Catalan, due to the familiarity of the authors with the context and since there is content created by public entities that claim to strive for universal access to their resources (see Sect. 1.2).

The accessibility of online educational platforms and content has been researched and some efforts to develop enhanced online learning environments have been performed, especially by researchers in Computer Science. Salomoni et al. (2007) took into consideration both the sensory and motor needs of the user and the technical characteristics of the device they used and developed an e-learning platform that is able to adapt its own materials to the user that is browsing it and their needs. Batanero et al. (2017) tested an adaptation of a Moodle platform following some accessibility principles and they created a system design of the learning platform that could be personalized with additional disability configurations. Luephattanasuk et al. (2011) proposed a method to present questions in an accessible format using WCAG and developed a prototype tool to check whether the accessibility of the question was compatible with NVDA screen reader software. They tested questions of different types (true or false, multiple choice, fill-in gaps) and their prototype showed to work and assess the completion with WCAG successfully, although the tool should be tested with its blind end users. The latest study presented rather than the learning platform. This is an overlooked aspect of e-learning that we will put the focus on in our study. The study aims to show some aspects of online educational videos that can easily be assessed in terms of accessibility and to evaluate some educational materials that the Catalan Government designs for initial Catalan learners.

In this chapter, we will provide an overview of accessibility in online education and delve into the profile of second-language Catalan learners. Following that, we will explain the methodology used in our analysis, present the results, and engage in a discussion of these findings.

1.1 Accessible Online Education

Ensuring universal access to education should not only be a government’s aspiration, but their duty. Article 26 of the Universal Declaration of Human Rights (UN General Assembly, 1948) states that “everyone has the right to education, [which shall be free and compulsory] in the elementary and fundamental stages. […] Higher education shall be equally accessible to all on the basis of merit.” For education to be accessible, we should aim for equity (fairness) rather than equality. And for education to be equitable, opportunities and resources need to be designed and distributed with the needs of everyone in mind, so everyone can reach the same outcomes and benefits. Going back to Article 26 of the UDHRL: not everyone will have the merits to access higher education if they have not undergone inclusive elementary education, if their learning needs have not been met, or if any physical or cognitive difficulty that they may have experienced has been overlooked. In the context of Catalonia, the Department of Education of the Catalan Government agreed on and published an accessibility plan (Departament d’Educació, 2019). This plan was unfortunately only applied to the communications from the department and not to public education practices. Additionally, money recently destined to accessibility in the public educational centers (ACN, 2023) is mainly aimed to topple architectural barriers. Media accessibility and the need to include students with sensory and cognitive disabilities in public strategies and policies should not be left unnoticed.

Media accessibility barriers escalate in the online learning environment. The barriers that traditional education presents for some students add onto the mostly sensory and cognitive obstacles that digital environments and hardware can also present: information communicated exclusively through a visual channel, auditory content that is not written down, hard to navigate digital environments, hardware not adapted to all visual needs, etc. In addition, teachers also often seem to struggle with online teaching environments, which require a certain level of digital literacy and intuition to navigate, and which differ from their traditional teaching skills and know-how (Sánchez-Cruzado et al., 2021). It is important to bear in mind that online education can engender accessibility barriers in terms of (1) content, (2) platform, and (3) hardware (W3 Consortium, 2023b). In order to provide teachers and learning designers with the resources to make online content and platforms as accessible as possible, a few organizations and institutions have collected recommendations and published guidelines, standards, and policies (e.g., W3C, ADAFootnote 1) and some projects that are working for these instructions to be put into practice in online education appeared recently (e.g., IDE@ and IMPACT projectsFootnote 2).

In terms of standards, there is an extensive ISO norm on “Individualized adaptability and accessibility in e-learning, education and training” (ISO, 2023b), which aims to facilitate the matching of digital learning resources to learners’ accessibility needs and preferences.

The W3 consortium has also published some recommendations and collected resources on accessible online learning, referring to their Web Content Accessibility Guidelines and always following the consortium’s approach to web accessibility. These four aspects should be assessed for any online content to evaluate its accessibility and detect the areas that need to be improved:

(1) “Perceivable—Information and user interface components must be presentable to users in ways they can perceive.

(2) Operable—User interface components and navigation must be operable.

(3) Understandable—Information and the operation of the user interface must be understandable.

(4) Robust—Content must be robust enough that it can be interpreted reliably by a wide variety of user agents, including assistive technologies.”

Understanding WCAG 2.0 .(W3 Consortium, 2023a, in Understanding the Four Principles of Accessibility)

Finally, the Disabilities, Opportunities, Internetworking, and Technology (DO-IT) Center at the University of Washington shared a list of recommendations for teaching an accessible online course (Burgstahler, 2021) and presented a section on inclusive pedagogy specially dedicated to neurodiverse students:

(1) Some of the recommendations on course materials encourage to:

– “Use clear, consistent layouts, navigation, and organization schemes.

– Provide concise text descriptions of content presented within images.

– Use large, bold, sans serif fonts on uncluttered pages with plain backgrounds.

– Caption videos and transcribe audio content.”

(2) And some of the recommendations on inclusive pedagogy encourage to:

– “Recommend videos and written materials to students where they can gain technical skills needed for course participation.

– Provide multiple ways for students to learn (e.g., use a combination of text, video, audio, and/or image; speak aloud all content presented on slides in synchronous presentations and then record them for later viewing).

– Address a wide range of language skills as you write content (e.g., use plain English, spell out acronyms, define terms, avoid or define jargon).”

20 Tips for Teaching an Accessible Online Course. (Burgstahler, 2021)

Having presented the above, professionals could find it helpful to have more specific and extensive guidelines on designing and teaching online courses in an accessible way; but it is also true, as we will present in the present study, that sometimes, the already-existing recommendations are often overlooked and not followed in online education.

1.2 Accessibility in Catalan as a Second-Language Learning

According to the last Catalan language policy report (Departament de Cultura, 2020), 36.4% of the residents in Catalonia were born in other regions of Spain or foreign countries. Furthermore, the foreign population is more prominently represented within the working-age demographic, particularly among individuals aged 20–54, with a noticeable concentration between 25 and 39 years old. On the other hand, residents from other regions of Spain tend to be over 65 years old.

According to the language policy report, which references the data from a survey conducted by the Department of Culture and the Statistical Institute of Catalonia (Departament de Cultura and Institut d’Estadística de Catalunya, 2018), 49% of the 1,380,720 people born in a foreign country that lived in Catalonia did not speak Catalan, and 62.5% did not know how to write in this language. Among residents from other parts of Spain (a total of 1,301,357) 38.9% did not speak Catalan and 72.1% did not know how to write in Catalan.

For individuals who could speak Catalan, they learned it through their daily life, regardless of whether they were foreigners or came from other regions of the country. Among foreigners, 37.3% reported learning Catalan primarily in their daily interactions, closely followed by language schools at 31.2%. Meanwhile, among residents from other parts of the country, a significant majority of 54.2% learned Catalan in their everyday lives, with the second most common setting being at work, where they learned from colleagues, accounting for 21.6%. Additionally, among foreigners, 61.7% indicated that they wished to improve or learn Catalan, while 27.7% of residents coming from other parts of Spain said the same. The age group that expressed the greatest interest in learning Catalan was those aged 30–44, with 43.6% showing interest. Following closely, the second most interested age group was those aged 45–64, with 40.3% displaying interest in learning the language.

Other Catalan-speaking regions of Spain register similar percentages of immigrants than Catalonia. Nevertheless, Catalonia has been more actively promoting the need to learn Catalan, although newcomers show a general higher interest in learning Spanish (Bori & Petanović, 2017).

Data strongly suggests a demand for self-paced learning resources tailored to adults, and a noteworthy percentage of residents who were born in other countries have expressed an interest in learning Catalan. Additionally, the data reveals a trend among those who already know Catalan, with a preference for acquiring oral language skills over written language proficiency. Furthermore, non-regulated, informal learning methods appear to be more prevalent in the Catalan learning context. The inclination toward non-regulated learning may be because the primary age group among migrants falls within the working-age bracket, which can potentially restrict their opportunities for structured and formal learning.

A recent study (Massaguer Comes et al., 2023) shows that at the Consortium for Language Standardization (CPNL) (the public agency that offers Catalan courses in Catalonia) it is common for classrooms to have students with very different profiles in terms of their native languages, educational levels, levels of motivation, and contact with Catalan outside the classroom, but tend to overlook that students can also differ in their sensorial and cognitive abilities. While diversity can enrich learning and provide an opportunity to learn from each other, the coexistence of diverse academic profiles and the lack of optimal tools and adaptations can also make it challenging to conduct sessions effectively (Massaguer Comes et al., 2023).

Catalan learners in the CPNL in many cases do not use their knowledge nor socialize in Catalan outside the classroom (Massaguer Comes et al., 2023). The main causes range from having a low educational level or not aiming for positions that require knowledge of Catalan to lacking access to Catalan-speaking social networks and not identifying or perceiving themselves as potential members of these groups in the future (Massaguer Comes et al., 2023). Again, the list of contributing factors does not mention sensorial or cognitive diversity, which should have been considered, as these impediments would probably be very relevant on whether a student succeeds not only in the acquisition of a certain level of language knowledge in the classroom, but also in the ability to use this knowledge in their everyday life.

2 Method

This section will provide an overview of the methodology employed in our analysis, the characteristics of the materials we examined, and the findings derived from this analysis.

2.1 Sample Description and Selection Criteria

In this study, we analyze the accessibility of two video series aimed at Catalan learners. The first series was created by the Secretariat of Language Policy and CPNL, while the second series was created by the Secretariat of Language Policy. These two series were selected as they were produced by public entities entrusted with the normalization and education of Catalan.

The video series “Ep! Escolta i Parla” (‘Hey! Listen and Learn’), hereafter referred to as “Series 1,” is addressed to students who lack literacy skills or are unable to use written language to learn. It comprises 21 videos, each spanning 1–4 min, that are divided in three difficulty modules. According to the series description, upon completion of the course, students typically attain a proficiency level approximately equivalent to A2 on the Common European Framework of Reference for Languages.

The video series “Màgia per a conversar” (‘Magic for Conversation’), hereafter referred to as “Series 2,” has the objective of providing supplementary resources for students engaged in language tandems while learning Catalan. Comprising a collection of 25 videos, each spanning 1–3 min, the series features magic performances designed to inspire discussion topics.

We chose a sample of five videos from each video series, resulting in an approximate duration of 10 min of material for each individual series. In total, we analyzed around 20 min of content. The specific videos were chosen with an effort to ensure similar lengths and total duration, enabling a comparable analysis. In Table 1, a compilation of the analyzed chapters and their respective durations is presented.

Table 1 Overview of analyzed videos
Full size table

3 Analysis and Results

In order to gauge the accessibility of the video series, we examined the video samples to determine if the creators had made them accessible to diverse student profiles. Since these materials are intended to enhance and promote the learning of Catalan and were developed by public entities, ensuring accessibility to a diverse range of students would contribute to better achieving this goal.

We performed a corpus analysis, following Arias-Badia and Matamala’s (2023) methodology, to determine the level of sentence complexity (sentence length and average verbs per sentence) and the distribution of parts-of-speech (PoS), that is, the usage of adjectives, nouns, verbs, and adverbs. This analysis was performed with the online tool SketchEngine and had the aim of helping determine if the videos followed easy-to-understand recommendations such as short and simple sentences and using a verbal style (European Accessible Information Working Group, 2011; Commission, 2012). An easy-to-understand language would mean that the contents are accessible to a wider variety of student profiles. We also manually looked at the speech rate by calculating the words per minute (wpm) and syllable per second (sps) rate.

For accessibility for students with visual impairments, we focused on the information given in the video images to see if this information was also present in the audio and if it was necessary for the comprehension of the videos (AENOR, 2005).

In terms of accessibility for students with hearing difficulties, we checked whether the videos included subtitles and subtitles for the deaf and hard of hearing (SDH) that communicated audio information or audio cues necessary for the comprehension of the video that would otherwise not be transmitted through the visual channel (AENOR, 2012).

Corpus analysis

In Series 1, most videos feature very concise sentences, typically consisting of 6–8 words, and usually containing only a single verb. This is a result of the videos being designed as dialogues involving two or more characters. The one exception is the first video, “Consells pràctics,” which is a monologue. Nevertheless, even in this monologue, the average sentence length remains relatively short at 15 words, and the sentences tend to contain an average of 2.8 verbs per sentence. This could suggest a certain degree of sentence complexity, but it is important to note that most of these sentences are coordinated with connectors such as “and” or “but.”

In contrast, Series 2 features monologues where a single person addresses the audience. In these videos, the average sentence length falls in the range of 10–15 words, with each sentence typically including 2–3 verbs. The sentence structure in Series 2 is slightly more complex compared to Series 1, incorporating both coordinated and subordinated structures (Table 2).

Table 2 Mean length and mean verb per sentence
Full size table

When it comes to the distribution of parts-of-speech (PoS), specifically, the percentage of content words that are verbs, nouns, adverbs, and adjectives, there is a difference observed between the two series. Typically, an average sentence should contain more nouns than verbs. However, upon closer examination of the results, it becomes evident that not only do the videos adopt a verbal style, but also the quantity of verbs used is unusually high in some of the videos.

Throughout the five analyzed videos, Series 2 maintains a relatively consistent use of verbs, ranging between 35 and 37% of the lexical words. Conversely, in Series 1, there is a wider range, with most videos falling between 32 and 39% in terms of verb usage. One video in Series 1 stands out as an outlier, with a 50% use of verbs.

When it comes to the use of nouns, Series 1 exhibits again a broader range, with noun usage percentages spanning from 31 to 45% across its videos. In contrast, Series 2 maintains a narrower range, showing noun usage percentages between 33 and 39%.

A notable difference becomes apparent when comparing verb and noun usage. Series 2 consistently employs both verbs and nouns in nearly equal proportions, with only a slight 2–4% difference in usage between them. However, Series 1 tends to favor either verbs or nouns more prominently. In Series 1, the difference between the percentage of verbs and nouns often falls within a range of 9–11%, indicating a preference for one over the other. However, there are two outliers in Series 1. One video exhibits a 19% higher usage of verbs compared to nouns, while another video in Series 1 shows nearly equal verb and noun usage, with just a 1% difference. It is worth noting that using more verbs makes the language easier to understand. In this case, both series show a higher-than-average verb usage (Figs. 1 and 2).

Fig. 1
A grouped bar graph plots percentage versus verb, noun, adverb, and adjective usage across 5 different activities. Each series depicts varying percentages. Verbs and nouns have the highest bars. The activities are listed in a foreign language text.

Series 1 PoS

Full size image
Fig. 2
A grouped bar graph plots percentage versus verb, noun, adverb, and adjective usage across five series labeled E 1 to E 5. Each series depicts varying percentages. Verbs and nouns have the highest bars.

Series 2 PoS

Full size image

Words per minute (WPM) and syllables per second (SPS)

Speech rate can be measured using different metrics. Although speech rate is often measured in syllables per second (Roach, 2007), studies in Catalan speech rate use the measure words per second as well (Canals et al., 2006; Albaladejo Mur, 2020). Due to this, we have used both measures in this study.

Fig. 3
A bar and line graph plots the W P M and S P S for 5 videos. The bars represent W P M and the line represents S P S. The bars range between 90 and 150. The line follows an ascending trend. The data are estimated.

Series 1 speech rate

Full size image
Table 3 Series 1 speech rate
Full size table
Fig. 4
A bar and line graph plots the W P M and S P S for 5 videos. The bars represent W P M and the line represents S P S. The bars range between 130 and 181. The line follows a descending trend. The data are estimated.

Series 2 speech rate

Full size image
Table 4 Series 2 speech rate
Full size table

  • Series 1: The average speed rate of the videos is 132.32 wpm, with the slowest being 91.94 wpm and the fastest 149.01 wpm. If the speed is calculated by syllables per seconds (SPS), the average speed rate across the five videos is 3.30 sps, with the slowest being 2.24 sps and the fastest 3.78 sps (Fig. 3 and Table 3).

  • Series 2: The average speed rate of the videos is 157.15 wpm, with the slowest being 135.27 wpm and the fastest 180.58 wpm. The average sps in this series is 4.44 sps, with the slowest video having an average of 3.83 sps and the fastest, 4.97 sps (Fig. 4 and Table 4).

Comparing the wpm and sps in each video reveals distinct delivery strategies employed in each series. Series 1 is characterized by a slower speech rate, achieved by pronouncing each syllable at a slower pace. In contrast, Series 2 achieves a slower speed not by pronouncing words more slowly, but by incorporating longer pauses between sentences. This contrast becomes evident when we examine two specific videos: “Em preparo per a una entrevista de feina U8” from Series 1 and “E4. L’elefant” from Series 2. The first video has an average speed of 143.75 wpm and 3.78 sps, while the second video exhibits a lower wpm average at 135.27 wpm but a higher sps average, measured at 3.83 sps.

Another example are the videos “Això és un amic U9” and “El programa de ràdio U7” from Series 1, which have an average speed rate of 145.56 wpm and 3.56 sps and 149.01 wpm and 3.70 sps, respectively. In contrast, the video “E5. Veritats i mentides” from Series 2 has a lower wpm average, at 141.05 wpm, but a higher sps rate, 4.40 sps.

Finally, it is worth mentioning that both series are available on YouTube, which allows the users to slow down the video if needed.

Audio description (AD), subtitles, and subtitles for the deaf and hard of hearing (SDH)

In case there are students with hearing difficulties or visual impairments in the course, including AD, subtitles and SDH in any audiovisual content are essential (Table 5).

Table 5 Overview of AD, subtitles, and SDH features
Full size table

Audio description

  • Series 1: There is no AD option offered, but the organization that has created the content provides transcripts of clips that include descriptions of some relevant actions. Other types of information are lost without the visual channel.

  • Series 2: There is no AD option offered, and there is information lost without the visual channel. However, the topic of the videos being focused on magic is not as relevant as the series’ main goal, which is providing conversation topics for language pairs. The users can still answer to the verbal prompts even if they do not see the magic trick.

Subtitles

  • Series 1: The videos do not include any subtitles, but the organization that has created the content offers transcripts of the clips that include some actions and the dialogue between characters.

  • Series 2: The videos include manually generated subtitles of good quality in Catalan. Sometimes the subtitles do not meet optimal accessibility conditions and appear and disappear too quickly, but YouTube has the option to slow down the video, allowing you to also view the subtitles at a more relaxed pace.

Subtitles for the deaf and hard of hearing (SDH)

  • Series 1: The videos do not include any subtitles, nor any audio information through the visual channel. This information is also not available in the transcription of the actions and the dialogue that is available to the students. If the audio channel is suppressed, no relevant information about the plot is omitted, but some audio effects are missed.

  • Series 2: The subtitles available in the videos in Series 2 do not provide any additional audio information besides the speech of the speaker. Nevertheless, not much information is lost, as the videos focus on a magician that speaks over a musical background.

In summary, both series showed some good accessibility practices, such as using an easy-to-understand language, but lacked some relevant others, like AD or SDH, which could potentially exclude users with sensory disabilities (Table 6).

Table 6 Overview of the analyzed parameters’ results
Full size table

4 Discussion

The objective of this study was to gauge the accessibility of these video series. To do so, we focused on linguistic aspects, speech rate, and the presence of AD, subtitling, and SDH.

In terms of linguistic aspects, the results show that both series adhere to practices of easy-to-understand language, which is any simplified language variety that aims to make information easier to understand (ISO, 2023a) and is thus accessible to different types of users. For example, in terms of sentence length, Series 1’s average ranges from 6 to 8 words, while Series 2 ranges from 10 to 15. To put these results into perspective, the average length found in easy-to-understand texts is 12–17 words (Arias-Badia & Matamala, 2023). Furthermore, in AD, which is characterized for having short and concise sentences (Taylor, 2015), we see averages like 12 in film AD (Arias-Badia & Matamala, 2023), 13 in opera AD (Hermosa-Ramírez, 2021), and 8 in short films AD (Matamala, 2018).

If we look at the average number of verbs per sentence, Series 1 has between 1 and 2 verbs and Series 2 between 2 and 3. Series 1 shows results closer to film AD again (Arias-Badia & Matamala, 2023). As mentioned, in Series 1, the sentences with more than one verb are coordinated with connectors such as “and” or “but,” while in Series 2, there are also some subordinated structures. However, this is not an indication of a higher sentence complexity, given that in Easy-to-Read texts (the easiest version of easy-to-understand) copulative or adversative coordinated sentences are acceptable, as well as subordinated causal, final, temporal, conditional, and comparatives sentences (García Muñoz, 2012).

Finally, the PoS analysis shows unexpected results. While other studies identify nouns as the dominant category (Arias-Badia & Matamala, 2023; Hermosa-Ramírez, 2021), in Series 1, two of the three videos have verbs as the dominant category, while in Series 2, it is three out of five videos. These results are quite positive, since they align with the European Commission’s recommendations (2012), which recommend using more verbs for easier comprehension.

While Series 2 shows slightly longer sentences with more verbs, it is important to remember that this series is addressed to more advanced students than Series 1. Considering this, both Series 1 and 2 are accessible in terms of linguistic aspects.

Regarding speech rate, the average in Catalan oral language is 120–150 wpm (Canals et al., 2006), while the recommendation for oral discourse is 150 wpm (Albaladejo Mur, 2020). Since the material we have analyzed is addressed to Catalan second-language learners, ideally the results should be within these parameters or lower. Series 1 follows these recommendations, with all the videos falling within the 120–150 wpm average or lower. In contrast, in Series 2 three of the videos have a speed rate higher than 150, while the remaining two keep to the average established by Canals et al. (2006).

However, when looking at another speech rate measure, syllables per second, the results are different. Series 2’s results are closer to the average established by Cremades Cortiella (2016) and Cicres Bosch (2008), at 4.92 sps and 6.45 sps, respectively. The range for Series 2 is 3.83 sps–4.97 sps, while Series 1 exhibits a slower pace, ranging from 2.24 sps to 3.78 sps. In essence, Series 1 proves to be quite accessible in terms of speed for both measures, whereas determining the accessibility of Series 2 is more challenging because; while it exceeds the average in wpm, its sps falls within the average range. Series 2 is addressed to more advanced learners, which might account for the faster speed, but it might also pose some difficulties to some students with varied needs. Nonetheless, it is important to note that the viewing platform of both series enables users to adjust the video speed, mitigating any potential drawbacks associated with a faster pace.

One of the main accessibility issues of both series is the lack of AD. Although Series 1 provides a transcription that narrates some of the actions of the characters, the resource is not perceivable and requires a certain level of autonomous research from the user. AD would not only help students with visual impairments to engage with the video the same way as their peers, but would also help auditory learners to stay focused and point out relevant visual information (Krejtz et al., 2012). Another big accessibility issue is the lack of SDH in both video series. Although they present good-quality subtitles, these do not present a text description of auditory information other than the dialogue.

5 Conclusions

Through our small study and theoretical exploration of accessible e-learning audiovisual resources, we have been able to firstly discuss the importance of overcoming the possible barriers than any online educational content or platform may present by taking accessibility into account from the early stages of their design and providing students with accessibility services and secondly to demonstrate through our study that this is nowadays not a common practice, even when the resources are created by public institutions that have been taking steps toward their communications being universally accessible (Generalitat de Catalunya, 2022, 2023).

In the specific context of our study, both Series 1 and Series 2 demonstrate accessibility to students with diverse needs. Series 1's language is easy-to-understand, with a slower speed and video transcriptions containing image descriptions, making it inclusive for all students. Despite lacking AD or SDH, Series 1 is designed in a way that minimizes information loss. In contrast, Series 2, while sharing an easy-to-understand language and a slower pace, provides subtitles but lacks AD or SDH, resulting in information loss for students without access to the visual content. This underscores the critical need for educational materials to either be fully accessible from their design or to provide the necessary accessibility features.

We aim for this chapter to foster multidisciplinary research with one common goal: to make online educational environments as accessible as possible. Online resources should make learning easier and accommodate everyone rather than being an additional obstacle for some people to access education.