Understanding motivations for creating accessible media content

Deliverable 4: Understanding motivations for creating accessible media content

Document Technical Details

Document History

Executive Summary

Social networks have permeated every facet of modern society daily life as they allow people to engage with one another. In particular, for people with disabilities, these platforms have, since early on, played an important role in disability advocacy, as they provide a vehicle for meeting new contacts with disabilities, learning about issues and news related to it, and discussing about challenges and solutions for improving social media accessibility. Despite the contributions to accessibility, social networks present substantial accessibility barriers for users with disabilities. The complexity of their interfaces compared to many typical websites comes from the fact that they are primarily composed of content created by users themselves. While most major platforms have an interface somewhat accessible, most user-generated content is still inaccessible.

The mechanisms to be developed in this project will support users in creating accessible content, by facilitating and recreating an authoring flow that is intuitive and accessible and, in addition, engages users in the production of accessible content. Besides, SONAAR also aims to retrieve alternative descriptions that had been already provided by other users in order to increase the amount of accessible user-generated content. Moreover, the solution proposed in this project aims to support accessibility practices in both desktop and mobile devices. The present study aims to uncover what does or can motivate people to author and share accessible content. This report describes the activities conducted to gather information on motivational factors that contribute to users engagement in accessibility practices in their social networks.

The findings obtained so far identified a series of topics that must be addressed by SONAAR in order to guide future design choices to create more user engagement in accessibility practices. Users tend to be more aware of accessibility approaches and their benefits when in contact with people with disabilities or when they are faced with some assistive technologies or adaptive strategies adopted by these people. In general, people are interested in participating in the inclusion process even though most of them are still unaware to the availability and benefits of accessible approaches. In order to engage users in these activities, SONAAR needs to provide examples on how people with disabilities are currently accessing social media and how they benefit from accessing accessible content. Moreover, SONAAR should provide users a easy-to-read and accessible tutorial in which users can find clear information and guidance on how to create accessible media content.

Contents

• Introduction
• Accessibility on social networks
  • Content sharing
  • Image captioning
• Objectives
  • Scope
  • Problem statement
  • Goals of the deliverable
  • Research questions
• Method
  • Procedure
  • Analysis
• Findings
  • General information
  • Social media usage
    • Device access
    • Frequency
    • Type of content
  • Accessible practices
  • Discussion
  • Qualitative data
    • Alternative description
    • Barriers
    • Content
    • Motivations
    • Reasons
    • Social networks platforms
    • Suggestions
• Analysis
  • Research questions
  • Implications for the project
• Conclusions
• References
• Annex 1: Online survey
• Annex 2: Code list

Introduction

The SONAAR project aims to respond to the accessibility challenge present in current social networks due to the large amount of user-generated content. That means that, for these platforms to be truly accessible, they have to go beyond ensuring that their interfaces are accessible: they need to ensure that the content their users produce is accessible as well. Despite the recent efforts in accessibility features, major social networks still present substantial accessibility barriers for users with disabilities and little is known about the barriers that mainstream users encounter when creating media content, which is at the root of the issue.

In this context, SONAAR’s overall goal is to prototype a mechanism capable of increasing the amount of accessible user-generated content available on social networks on mobile and desktop platforms. In order to achieve this goal, SONAAR needs to facilitate the authoring of accessible content and to create more engagement from users in creating accessible content. Additionally, it is crucial that an accessible content authoring process is provided. The deployment of a mechanism that will exploit the possibility of retrieving an alternative text for the same media content that has already been provided by another user is also expected.

In order to find adequate concepts and solutions to guide the development of the project’s following activities, further requirements must be established to respond to the specific needs of current social network users. This implies, for example, information about their usage of social networks and their levels of engagement with accessibility practices in social networks. For that, this report describes the conducted activities in order to gather information on the motivational factors that influence these users to currently publish accessible content and what can motivate them to become more engaged in these practices.

This document is structured as follow: Section 2 summarizes the current state-of-the-art regarding media accessibility on social networks. Section 3 presents the objectives established such as the research questions addressed in this context. Section 4 outlines the methodology adopted. Section 5 details the gathered data and the results obtained. Section 6 analysis the findings and discusses how they contribute within the context of the project. Section 7 described final conclusions and further activities to be conducted in the scope of this work package.

Accessibility on social networks

Social networking services, such as Facebook, have a high adoption rate among blind users. This also happens with Twitter, which evolved from a very simple and text-based interface to one that is now filled with multimedia content. These platforms rely on user generated content that, with the growth in publications containing visual content, pushes the social networks systems to become increasingly inaccessible [1].

Numerous efforts have been made to improve accessibility in visual content on social networks with some of these initiatives coming from the service providers themselves. For instance, in 2016, Twitter included a feature allowing users to compose their own alternative descriptions for their images. However, users claim that this initial feature had its drawbacks as it had to be enabled by the users themselves also being hard to find and understand [14]. Even though this changed in 2020 [15], as at the moment this resource is active and available by default, the impacts of this measure have not yet been discussed. Meanwhile, Facebook made a choice to use automatic descriptions by tagging each image uploaded using image detection and recognition algorithms and enabling the user to edit the alternative description automatically provided [5,6].

Content sharing

One of the greatest challenges today is the engagement of users in providing accessible content. Through an analysis of over a million images posted on Twitter, Gleason et al. [7] observed that only 0.1% contained an alternative text. While it is possible to make Twitter itself accountable for this low number, by not enabling by default the feature that allows the inclusion of the alternative description for many years, they also observed that even the users who enabled this feature to provide the alternative text descriptions did not always write them.

Visually impaired users engage in major photo-related activities as actively as other Facebook users, considered by them as part of the social network experience. However, most of the time the practices that include taking and editing photos as well as providing an alternative text for it, involved getting help from trusted sighted contacts [10,16,17] or implementing workarounds for conducting such activities [1]. In a survey conducted by Mathur et al. [10], friends and family members of visually impaired users reported having written an alternative text for some of the photos they uploaded to social media, but they also conveyed that writing alternative text is time consuming and requires more thought than inaccessible uploading practices. Although these last outcomes may seem contrasting with the really low percentage of images posted on Twitter that contained alternative texts observed by Gleason et al. [7], they also observed that images exposed to visually impaired users contained descriptions slightly more often than the average. Moreover, according to Wu et al. [17] visually impaired users are much more likely to have friends who are also visually impaired.

Image captioning

This low user compliance in providing alternative text descriptions is a common web accessibility problem and therefore some alternative methods are usually employed to fill this gap. Stangl et al. [13] classified some of the existing approaches to generate image descriptions as human-powered approaches, automated image descriptions approaches and hybrid image description technologies. Human-powered are recognized by users for its accuracy and quality of responses. Techniques such as Crowdsourcing may have slow response times for real-time needs and high financial cost. Friendsourcing may improve the quality and trustworthiness of the answers received – as friends may be better able to answer questions because they understand the question asked – and also removes financial costs of the service, but the social costs of exposing one’s problems and vulnerability are a serious concern for these users as they may appear or feel less independent [2]. As for automated approaches, unlike the previous technique, they are fast and cheap, allowing platforms to deploy them at scale [4,8]. While there are significant efforts undertaken over the past years on image understanding and automated captioning, the accuracy of these captions may not be yet sufficient. Caption and phrasing models have an important impact in skepticism as blind and visually impaired people are likely to rely more on automatically generated captions than on their intuition, making decisions based on misinformation [8]. Besides that, limitations still exist on the adequacy of these systems used in the open world, especially when captioning the wide variety of images posted to social media [12]. Thus, in order to fill the gap in image captions, hybrid image description technologies propose a combination of automatic techniques and human intervention to explore a trade-off between both techniques, as explored in [8,9,11,12].

Objectives

Scope

SONAAR aims to facilitate the user-generation of accessible content on social network services by developing a solution that supports the authoring and consumption of media content on social platforms in both desktop and mobile devices.

This deliverable regards the third work package (WP3) of the SONAAR project, focused on engaging users in the production of accessible content. The activities detailed in this document aim to better understand users’ motivational factors in authoring accessible media content in social networks and how can we motivate users to engage in authoring and sharing accessible content. Specifically, we want to understand why some of these users are motivated and, as consequence, engaged in accessibility practices. These motivational factors can be from social context as well as possible technological aspects.

The data gathered will be translated into documentation that can be presented to end-users when posting or sharing media content. This set of information has to be able to motivate end-users in creating accessible content as well as address the challenges they face through this process, as described in a previous deliverable (D1: Accessibility barriers to publishing content on social networks).

Moreover, we also want to investigate eventual features and information that should be considered on the prototypes being developed in the scope of SONAAR project in order to drawn more attention from users to the subject and also educate them in creating accessible media content. For that, considerations provided in this report will also be useful for the design and development choices in further activities.

Problem statement

Despite the existing features already provided by major social networks, users are often not engaged in accessible practices when publishing their content. This leads to a significant amount of not accessible content being published which could be addressed by the users themselves. The methodology established to carry out the tasks defined in WP3 is focused on strategies to educate authors of media-based content for social networks on the best way to create accessible content, as well as to investigate motivational factors that could potentially contribute to a higher level of engagement in accessible authoring practices.

Goals of the deliverable

This deliverable aims to report the research conducted to investigate motivations of social network users in authoring and sharing accessible media content. This report also discusses the implication of the data collected and the findings on the project, in particular, in particular which aspects should be taken into account in the next stages of the project to engage users in accessible practices.

Research questions

The amount of media content available today in social networks is relatively high and most of this content is not properly described. This means that most of social network users are not engaging in such practices. As reported in a previous deliverable (D1: Accessibility barriers to publishing content on social networks), these users encounter a number of barriers in this process, so a first step to engage these people would be to address these barriers. Secondly, it is necessary to draw up strategies to get users to engage more in this process. For that, we wanted to investigate the motivating factors among people who are driven to author and share accessible content as well as what can possibly be done to engage other users in these activities. For this purpose, two research questions were established:

• RQ1: What are the motivations for users of social networks to create accessible media content? With this question, we try to identify the motivational factors that lead users to be currently involved in accessible practices.

• RQ2: How can we motivate users of social networks to create accessible media content? This research question aims to investigate extra factors that may contribute to engaging users who are not used to sharing and creating accessible content as well as create extra motivation in users who are already involved in this process.

Method

Procedure

In order to address the research questions established, this study was structured into three different phases. Ethical approval to run the study was granted by the Faculdade de Ciˆencias da Universidade de Lisboa Ethics committee.

Preliminary interviews to inform survey design. In order to gather relevant input to better identify key aspects to be addressed by participants in subsequent phases, we conducted an initial interview with 7 social network users. For this phase, we focused on obtaining a representative sample of the different social network user profiles addressed in this study, i.e., sighted and blind users, with different levels of accessibility awareness. For that, the research team recruited a number of participants from their contacts, meeting the established criteria. More information about the interviewees is summarized in Table 4.1.

Interviews were remotely conducted, taking between 30 to 40 minutes. The questions asked were pre-established for the survey study while participants were encouraged to propose suggestions and improvements to enhance the overall understanding of the survey. Questions were built around the main topics of this research, such as their usage of social networks, general barriers encountered when sharing and authoring accessible media content, awareness to the possibility of including alternative texts in digital media, and general motivation and considered importance of sharing and authoring accessible media content. Based on their contributions and insights, some adjustments were made to the initial survey in order to construct the following survey and interview study.

Survey study. We conducted an online survey to gather data as well as to recruit a representative sample of participants to the subsequent phase. The survey was constructed using Microsoft Forms in four different languages (Portuguese, English, French and Spanish) in order to reach a diverse sample of participants (Annex 1: Online survey). We disseminated it through the research team social media contacts, including their institutions and fellow organizations (including disability-related ones), gathering a total of 258 valid answers. From that, 163 answers in Portuguese, 52 in English, 30 in French and 13 in Spanish. Participants were aged from 17 to 73 years old with 64 (25%) of them self-reporting having some kind of disability. Further details will be provided during data analysis.

The survey took about 15 minutes to complete and featured a brief technology use and demographics section, followed by some short open questions about authoring practices in social networks. Before finishing the questionnaire, participants were invited to share additional thoughts as well as to share their emails if interested in participating in the next phase.

Interview study. Subsequently, we contacted the 93 survey respondents who shared their emails in the survey, providing more information about this next phase. We followed up and interviewed the 20 that answered back. More details about the participants are summarised in Table 4.2. We asked participants who do not author accessible content to do so prior to the final interview as well as to take notes about the experience to share it later. We focused the interview on the participants’ experiences, seeking to identify barriers and underlying motivations for accessible content authoring. In the instances where there weren’t intrinsic motivations to pursue accessibility, the interview shifted to understand what could be done, and how to prompt compliance.

Analysis

In order to analyse the gathered data, all conducted interviews were transcribed and two researchers reviewed independently a subset of them through a mixed inductive and deductive coding approach. The coders compared their initial set of codes and their categorizations in order to develop a unified list of codes. Then, the two coders reviewed a new subset of transcriptions taking into account the consolidated and revised codes in order to reach an agreement. Cohen’s Kappa (K) [3] was calculated on agreement between the two coders and results indicated that a good level of agreement was reached (K=0.70). Following that, the coding of all the data collected was performed. A total of 150 distinct codes were identified and the complete list of codes generated and used in this analysis is present in Annex 2: Code list.

Findings

We got valid answers from 258 participants. In this section we present the data gathered through the survey containing demographic questions as well as their social media usage and accessible practices followed by a brief discussion on these findings. Next, we detail the findings obtained from the follow-up interviews conducted and insights into the themes that have emerged from them.

General information

In order to learn more about our sample, participants were asked for basic characteristics such as age, gender and technologies usage. It is important to point out that, for this survey, questions were not mandatory, in order to let the participant choose which information they would like to share with us. This sample comprised 258 participants, of whom 133 were female, 120 were male, 1 non-binary, 3 responded that they preferred not to answer and one did not provide an answer. Their ages ranged from 17 to 73 years old, with a median age of 55 and a mean of 47 years (IQR = 23,75). Regarding the operating systems they use, most of them reported using Windows as a desktop or laptop system and Android for their mobile devices (Table 5.1).

Furthermore, 64 participants self-reported having a disability. Among those, 46 reported having a visual impairment, 9 of them a hearing impairment, 5 users were motor impaired, 4 declared having problems with cognitive performance as other reporting have autism, colorblindness or a chronic condition as illustrated in Table 5.2.

We also asked these participants about the use of assistive technologies and 41 of them reported using a screen reader, 17 used a braille display and 6 used a screen magnifier. Other technologies or strategies were also mentioned by them such as read aloud functions, larger fonts or Siri virtual assistant (Table 5.3).

In order to get insights on the respondents familiarity with the needs of people with disabilities, the survey asked if they had family, friends or acquaintances who have a disability and if they were aware of any of their contacts on social networks having a disability. As reported in Table 5.4, most of the participants do not have family members, friends or acquaintances with a disability. However, the situation is the opposite when analyzing their contact list in social networks.

Participants have more contact with motor and visually impaired people, followed by people with hearing and cognitive impairments (Table 5.5). Some of them also declared having friends or family members with Osteogenesis and Mowat–Wilson and Down syndrome.

Social media usage

In order to better understand the usage of social media by users with and without disabilities, several questions were asked regarding their authoring and sharing practices. In addition, to further analyse the difference in these practices between the groups of users with and without disabilities, the following data will be presented for each group. A more comprehensive discussion of the data obtained will be presented at the end of this chapter. For comparative purposes, the data presented in this section will be represented in percentage, concerning the total of respondents in each group profile.

Device access

We also aimed to investigate the usage of different devices and interfaces by participants when performing their activities in social networks.

Mobile devices. The values illustrated in Figure 5.1 represent participants’ frequency in accessing, posting and sharing activities through a smartphone or a tablet. Participants with disabilities were more frequently engaged in posting and sharing activities than users without disabilities, with only 22% of them reporting posting and sharing content several times per day through a mobile device. However, regarding accessing their social networks, 88% of participants without disabilities reported using a mobile device several times per day against 81% of participants with disabilities. All data collected are also represented in Table 5.6.

Desktop or laptop devices. Besides the difference between the 52% of participants with disabilities and the 43% of respondents without a disability that declared accessing their social networks several times per day using a desktop or laptop device, when it comes to posting and sharing activities several times per day this gap is even more significant, as shown in Figure 5.2 and Table 5.7.

How users access social networks. We also asked respondents how they access social networks on these devices. While the percentage of participants who uses a mobile app doesn’t differ significantly between these two groups, participants with disabilities tended to access social networks through the platform website on desktop or laptop computer, or on mobile devices (45%) more than participants without disabilities (33%) as seen in Figure 5.3 and Table 5.8.

Frequency

We also asked the participants which social networks they use and how often. Their answers are categorized next by type of activity: access, posting and sharing.

Social networks frequency of access. Concerning the social network used by them, Facebook was considered the most used social network among participants without disabilities (48%) while Twitter was chosen by half of participants self-reporting having a disability (50%). However, only one of them reported having Instagram as their primary accessed social network, being Instagram more popular than Twitter among participants without a disability (Figure 5.4 and Table 5.9).

Social networks frequency of posting. Participants without disabilities reported to post most of their content on Instagram followed by Facebook and Twitter. Regarding participants with disabilities, they declared to prefer to post their content on Facebook followed by Twitter and a very low percentage of them declared using Instagram as their main social network to post their content. This values are represented in Figure 5.5 and Table 5.10.

Social networks frequency of sharing. The same pattern was observed, since most of participants self-reporting having a disability declared to use Facebook primarily to share content from others followed by Twitter. However none of them mentioned using Instagram in this activity. Regarding participants without disabilities, they reported to share more content in Facebook, followed by Twitter and Instagram, as presented in Figure 5.6 and Table 5.11.

Type of content

Finally, we also seek to investigate how often participants post and share different types of content on their social networks.

Content posting. We also asked users to share withus how often they post some specific type of content on their social networks, i.e., text, audio, video, image, meme, GIF. From the participants’ responses it is possible to identify that participants with disabilities reported to post content much more frequently than users without disabilities. Visual content also represents a difference between two groups, since around 40% of participants without disabilities are not used to post memes and GIFs while this percentage for participants with disabilities goes up to 70%. The percentage of participants with disabilities that declared posting images is also considerably lower than that of participants without disabilities. This values are represented in Figure 5.7 and Table 5.12.

Content sharing. We also asked participants to share with us how often they share some specific type of content on their social networks, i.e., text, audio, video, image, meme, GIF. The same trend is also observed in sharing practices. Participants with disabilities reported sharing much more textual content than participants without disabilities, and there is also a slight difference regarding media content. However, in this case, we observed a change in pattern as, in the case of sharing activities, 3% of participants with disabilities declared to share Memes and GIFs several times a day, as presented in Figure 5.8 and Table 5.13.

Accessible practices

We also asked our respondents about their accessibility works. practices when using social net- The first question on this section was if they provided an alternative textual description for any of their last three posts containing any type of media content (i.e.: audio, video, image, meme or a GIF). The answers obtained in this question were completely contrasting, as illustrated in Figure 5.9 and in Table 5.14. Most participants without disabilities declared not to have added an alternative description, contrary to the majority of participants with disabilities who declared to adopted such practice.

In the sequence, we asked the reason why participants did not provide an alternative text for their last three media content posts. The majority of them did not know that it was possible, followed by those that did not know where to write an alternative text description and those that find this activity too time-consuming. Also, 20% of them provided another reason such as forgetting or not thinking that it was necessary for their social networks’ friends and followers. Open answers to this question will be more thoroughly analyzed later. These values are presented in Figure 5.10 and Table 5.15.

In contrast, we also asked participants that did provide an alternative description for their last three posted media content how they did it between a set of 3 predefined answers with an alternative open option. Most of them (51%) used the functionality provided by the social network, while 46% chose to integrate the description in the text of the post, as presented in Figure 5.11 and Table 5.16. Also, two participants used an automatic service for alternative text generation service while one of them declared having added the description in a threaded post via the open answer. It is also important to highlight that participants could choose more than one option.

In the sequence, Figure 5.12 and Table 5.17 summarize participants opinion on why social network users in general do not provide alternative text descriptions for media content they post or share. They also had a predefined set of 5 answers and another one with an open answer. Participants could choose more than one option. Most of the participants consider that other people don’t know that this is possible, followed by participants that consider that other people don’t think it has any impact. Another 7% of them provided other answers concerning the awareness of this task. These answers will be more detailed and analyzed later.

When asked about additional steps to enhance media content accessibility in general, the percentage gap between users self-reporting as having a disability and users without a disability is even more significant, as shown in Figure 5.13 and Table 5.18. We also asked them to describe these additional steps and most of the participants reported having practices such as video captions, plain language, and alternative text descriptions. These answers will be further analyzed in detail.

We also asked them about accessibility concerns from their acquaintances when posting or sharing content on social media. Participants could choose between a set of predefined answers containing besides Yes or No, I don’t know, and prefer not to answer. Half of the participants self-reporting as having a disability declared having family or friends posting or sharing accessible content, while most non-disabled participants had no idea about accessibility practices of their acquaintances (Figure 5.14 and Table 5.19).

Finally, we asked participants if they were interested in having access to a tool that would support them in accessible media content authoring. While there was a more substantial number of participants self-declaring as having a disability interested in a support tool for accessible media content authoring, in general, most respondents would be interested in such a resource as shown in Figure 5.15 and Table 5.20. Following, we asked these groups of people about the features that they would include in it. Most of the answers were concerning the possibility of giving appropriate suggestions for alternative text description or a more user-friendly way of doing it.

Finally, participants were also invited to share additional thoughts or comments about the subject. We collected 25 open answers that will be further analyzed.

Discussion

Data collected from the responses provided in the questionnaire indicate that participants with disabilities do not engage in authoring and sharing practices of media content such as images, GIFs, or memes at the same frequency that participants without disabilities. The opposite situation is found when analyzing the frequency of activities related to textual content, being more frequent among participants with disabilities. It was also possible to observe that participants with disabilities tend to use Facebook and Twitter as often as participants without disabilities. However, Instagram was not so popular among them, in particular in accessing, i.e., consuming media content.

Concerning accessibility practices, participants with disabilities are often more involved in additional steps to enhance media content accessibility in social media as well as their acquaintances.

Finally, the main reason claimed by participants to not publish accessible content is not knowing it is possible or where to write an alternative text description. However, when it comes to why other users in general do not provide alternative text descriptions, besides not knowing it is possible, the second most popular option was concerning people not thinking it has a real impact.

Qualitative data

In the follow-up interviews conducted with 20 of these participants, we investigated some of the questions presented in the research in more detail. As previously mentioned, 150 codes were identified, organized in a two-level hierarchical structure as presented in Annex 2: Code list. This discussion will focus on the level 1 codes with the highest incidence as a baseline to our analysis and correlation with other identified codes.

Alternative description

An accessibility strategy for providing an alternative way of accessing some content, participants that were invited to learn and share some accessible content used this resource to promote the accessibility of posted content:

"What I actually used was a description of the images, I think it is a resource that I see that is used and that is what I relied on, because I see a lot."

More specifically, participants also discussed different strategies of providing an alternative description such as an audio description, automatic description generated by platforms, integrating it in the content post or an input field.

Barriers

This code is of extreme relevance since it concerns one of our research questions. Participants with and without a disability identified a series of barriers faced by them while sharing or consuming media content on social networks. First, a major challenge is concerning the unawareness of people in general on the use of technologies by people with disabilities and, as consequence, on what they should do to improve the accessibility of their content:

"There is very little information about these tools, I didn’t even knew it myself."

Also, in order to share or understand these media, participants with disabilities identified the lack of different kinds of accessibility resources in general such as alternative descriptions for images provided by other users, the additional context in automatic descriptions generated by platforms, or a feature to assist themselves in describing their posted media content:

"Twitter allows us [...] to provide descriptions of the images we publish. The problem [...] is that there is nothing automatic, so myself, for example, if I want to do it [...], someone has to tell me beforehand what it was."

Another factor reported by participants with disabilities is that their social network feeds are in general not accessible specially image-only posts and some other types of media such as GIFs or Stickers:

"The GIFs, I can’t picture myself, what it’s for or what it does. There are more and more little things like this appearing. It’s not always easy to make them accessible."

Another major issue raised by participants was that the accessibility features provided by platforms are often not user-friendly and hard to find:

"If they told me “Oh yes, you can make these things accessible, you just have to take a lot of very complicated steps to get it done”. Obviously, this will make it very difficult for people who are interested in doing it. Accessibility features are difficult to find and not very discoverable, they’re hidden. And in most cases, they’re turned off by default."

Some of the participants invited to post accessible content to this research and share their experience with us also reported that, even if they were aware of the importance of this activity and manage to find a way to provide an alternative description, they don’t know what to write when trying to represent an image through text:

"My biggest difficulty was this, not knowing if there was any standard or not, what would be necessary to include or not, if I should put more information, less information, if I should be more specific, give more details."

Besides, participants also reported having issues with API support on accessibility features in third-party clients as they can’t access URL feed. With the increase of features of color filters in images or color being used as the only means of communicating, color contrast was also cited as a barrier concerning specially colorblindness users. Loading speed and scroll speed were considered setbacks of accessibility practices as it adds more information to be loaded by apps and it may lead to redundancy of content.

"People could describe it just like I did or record a short audio that could be heard by those people [. . . ] it would be great for them, but I think it represents a setback for not visually impaired people."

Content

Regarding types of content, image is a major concern for social network users, especially for participants with a visual impairment when it comes to major platforms and accessibility features. Participants without a disability mentioned the difficulty of finding information and platform support on how to provide accessibility for videos such as adding video captions or automatic translations. Another issue raised by visually impaired participants concerns images containing text and videos without an audio description such as these with only soundtrack in the background:

"And they send a video, and they send a picture, and then I don’t know what’s in it. It’s often videos that are without sound or images with text in them. And so, they send the picture, and you don’t really know what’s in it."

Still, in this context, the story is a new format that has become quite popular on major platforms that supports a mix of images, videos, text, and even stickers. However, it raises another challenge for users as, once more, they don’t know how to provide accessibility for this type of content:

"Create a story with subtitles, describing what the person is talking about. I never realized that this could be a form of accessibility for a deaf person, for example."

The diversity of media types that these platforms provide often fails to match the accessibility features provided. Content such as stickers and GIFs are a concern for visually impaired participants, however, emoji was reported as being quite accessible for them:

"There are messages that contain emojis, but by chance, I can read the emojis [...], only now there are the stickers that I do not read."

Another issue raised by participants on types of content was concerning links that, when posted on social networks, often generate a preview of the website and often not containing any information about it:

"In other words, whoever simply put the mouse over the image with the preview of it couldn’t see anything at all."

Finally, participants were also interested in combining the automatic location provided by these platforms with alternative descriptions, the possibility of providing an alternative description through a pre-recorded audio, and having more memes described and explained through social networks.

Motivations

Participants were mainly engaged in accessible practices in order to promote inclusion in general or in particular for having a disability themselves or having acquaintances with disabilities. For that, by improving the accessibility of its content, it makes their information reach more people through their social media:

"I think that the more information we have, the more we can reach people, and the more that people see and that people have access to diversity, maybe empathy improves, maybe people’s quality of life improves because they have access to different information, then they can understand about different things and they can develop critical thinking, develop ideas and develop their subjectivity and develop their own individuality."

Besides, most participants also mentioned producing accessible content as a consequence of it being required by platforms or by law through initiatives such as guidelines compliance demanded by the governments in numerous countries, but also because they consider that it’s the right thing to do in general:

"What motivates me? I think it was a combination of we’re required by law to do it. And of course, it’s also just the right thing to do."

We also asked participants what they consider as other motivations or possible benefits of improving the accessibility of media content on social networks. Improving search engine results (SEO) and evolving AI solutions concerning visual and auditory competencies could also be an outcome considered as a motivation for many participants. Platform support is also considered as a motivational factor as it assists users in being aware of this possibility and removes some of the aspects previously mentioned as a barrier to accessible content authoring:

"If I knew how to do it, I could feel more motivated, if I had these instructions."

Participants also enhance alternative descriptions provided by major platforms that are automatically generated and often not very clear or informative in order to provide additional context and improve access:

"Facebook and Instagram have decided that people don’t know [...] and I think that’s actually in a way worse, because then things come out, like “picture contains a tree or may contain a table”. And it turns out that a tree is a person, and that table is a dog. [...] They take it out of people’s hands. And when it comes to being accessible, you can’t take it out of somebody’s hands [. . . ] Let them decide that they want to say, you know, a person or maybe give the person’s name or, you know, a silhouette, whatever they, how they want to describe it, whatever they were thinking of, why they were posting, that image should be what they’re posting."

Finally, providing alternative descriptions also improves users’ creative ability, enables them to develop skills when trying to portray their images through text or for language learners such as acquiring a second language. This information is likewise useful to provide an alternative access option in cases such as sound or internet network constraints, for example.

Reasons

Many participants who are aware of the role of accessibility simply don’t know how to make their content accessible or how major platforms support these practices. Combined with finding barriers such as those previously described during their attempts, many of them eventually end up not adopting this as a usual practice:

"I don’t know what it is to share something in an accessible way, I at least don’t do it and it doesn’t seem that the people I know do it, I think it’s because first of all for not knowing."

In some cases, participants have to deal with the lack of consistency of these features having to constantly relearn how to use them due to the constant interface updates:

"I find it difficult to find the field, to put the alt text in because every time I go into Twitter and send a picture, it’s changed."

Social network media posts are also claimed to be a moment that users want to share quickly and providing an alternative description for this content represents an extra and time-consuming step:

"There are days that I publish more than 20 stories a day [...] so, to do this process manually, of typing what is the caption..."

Some participants also stated that they just forget to do it or that it is unnecessary because they are used to sharing their content only with close friends and family and none of them has accessibility needs:

"Personally, I don’t see why, I don’t see any reason to do it because I don’t have anyone who would need this additional explanation either."

Some participants with disabilities consider that some people just don’t care enough to meet their needs. Partly because not everybody knows a person with a disability, so people don’t reflect on this issue. Partly also because most users are interested in a broader audience, without thinking about inclusion:

"Other people don’t care, no matter how much we tell them."

Furthermore, many of these reasons previously mentioned are reinforced by the fact that users may feel that these accessible practices are optional, as they are not a requirement for them when interacting with media in major platforms:

"I know that now we can include it [alternative description], but they don’t have to, they can send it just like that. Well, I think they won’t do it."

Social networks platforms

As previous data showed, Facebook, Instagram and Twitter are the most mentioned platforms by participants.

Facebook was often associated with automatic alternative descriptions. However, while most of non-disabled users are not completely aware that this feature is already available, visually impaired participants stated that these descriptions are not rich enough to be largely used:

"In the case of Facebook [...] I see something interesting, I pick up my phone, I take a picture and I post it directly from the app [...] I don’t think they ever gave me the option to put a description in this picture."

Largely adopted by visually impaired participants, Twitter has been pointed out by previous studies for hiding its accessibility features, which was also remarked by them:

"Twitter’s way of making a field for people to put in the alt text is the best way to do it, but I can’t tell you how many people can find it."

Since most visually impaired participants do not interact frequently with Instagram, this platform was mainly highlighted by sighted people. Participants that were invited to engage in accessible practices for this study provided an alternative description for their images writing their texts in the content post but providing an accessible experience for videos is an obstacle for them. Furthermore, these participants were really confused why blind people would be interested in using a highly visual social network such as Instagram:

"I also share a bit of this stigma, which is that if Instagram is highly visual, a blind person will use it? So, this sharing is like having the Olympic games and the Paralympic games mixed, it would be great to see this inclusion, but I don’t know also if there is this motivation on the part of the blind people to use a tool that is so visual."

LinkedIn, WhatsApp and Youtube users also mentioned the lack of accessibility in media content such as images, videos and audios pointing out the dynamism present in WhatsApp and how other people do not think about providing an alternative way of interpreting its content:

"On WhatsApp, for example, I have an acquaintance who is deaf, [...] we posted [...] an audio file and so [...] she said ”What the heck is that? What’s going on?”."

One blind participant decided to take a chance on Tiktok, also a highly visual platform, and shared with us his experience so far as having an inaccessible content in general but also an inaccessible interface:

"TikTok, I am testing it without getting anything from it. [. . . ] you can be lucky enough to get the short video clip, which it has speech on, which gives you a little bit, but then you can fit 300 short videos without even [...] understand what it’s about. So, it’s totally inaccessible. And in addition to that, the app itself isn’t accessible as well [...] TikTok I would say totally inaccessible, and therefore, I’m not using it. I don’t know if I would use it if I could, but I’m not using it."

Other platforms were also mentioned as being used, such as Messenger, Parler, Flickr, Nextdoor, Reddit, Slack, Snapchat, Soundcloud, Telegram, but no further relevance was shown by participants during the in-depth discussions on accessible practices. Finally, the lack of support of accessibility features when using these platforms in general through APIs and third-party clients may also be hindering the accessibility of media content.

Suggestions

Many interviewees came to the conclusion that one effective way of improving accessibility in social network media content is through initiatives that raise awareness concerning many aspects of this context. Participants believe that social exclusion is also reflected in social media and, for that, users do not consider this issue in general:

"It is much more comfortable for you to keep yourself in the non-possibility of providing accessible content because in fact, and I speak as an example, structurally, this is common."

Besides, social networks are also being called up to emphasize accessibility features as currently they don’t draw enough attention of their users to this matter:

"They have to be more visible in the authoring processes [. . . ] even if they weren’t hidden, they could also be promoted, which they aren’t either."

Participants also suggested that platforms, non-disabled users and governments have the responsibility to ensure this process. People need to have a greater commitment to the whole process of inclusion, platforms need to make accessibility a part of authoring flow and governments have a dual role, the legal context of digital media regulation as well as in the context of public policies for social inclusion. On many occasions, participants made the connection with mandatory accessibility measures found in physical structures such as ramps and handrails. Thus, many of them believe that this should be a requirement both by governments to platforms and by platforms to users.

"Imagine it [Facebook] doesn’t have many definitions of accessibility, if they want to be running in Portugal maybe our government would say to Facebook “In Portugal you must at least have these accessibility options”, for example. That happens sometimes, I think that was not bad either."

Finally, from a technological perspective, desirable features could include a content warning prompting users to provide an alternative description or captioning for their content as well as a suggested alternative description for it, giving the user not only an example of it but also the opportunity to correct automatically generated descriptions. Furthermore, participants are also interested in audio descriptions and more personalized accessibility settings such as color contrast or some accessibility guidance and hints.

Analysis

In this section we will analyze the data collected and previously presented. This analysis will focus on the findings that can help us answer the research questions determined for this deliverable and their impact on the project in general as well as on future activities.

Research questions

• RQ1: What are the motivations for users of social networks to create accessible media content??

Users are mainly motivated in doing the right thing and promoting inclusion for people with disabilities. Although some of them are used to share media content only for a specific audience, such as family and close friends, many of them are interested in enabling these people to have access to information as well.

As a side effect, accessible practices may also have an impact in improving search engine results for websites and current AI solutions. Providing alternative access options is also useful when users can’t turn on the sound on their devices or when experiencing internet network constraints. Besides, it can be helpful as a support for language learners or for accessing content in other languages.

• RQ2: How can we motivate users of social networks to create accessible media content?

Considering the willingness of including people with different abilities and cultural contexts showed by interviewees, some strategies may be applied in order to motivate more users to create accessible media content. People tend to be more aware when in contact with a person with disability but also when they are confronted with current accessibility approaches. They show curiosity and interest in knowing more about the subject and they are motivated in understanding how their content is being consumed by people with disabilities. The unawareness of users on how and why people with disabilities use social networks was one major challenge identified in this study, making them part of this process as well as educate them about accessible practices and alternative access may contribute to engaging them more frequently in such practices.

Implications for the project

This report concerns the third Working Package (WP3) that aims to engage users in the production of accessible content, addressing also one of specif goal of SONAAR. As previously stated, users are interested in participating in the inclusion process even though most of them are still unaware of the availability and benefits of accessible approaches. Besides technical aspects, in particular, the barriers presented in the first deliverable (D1: Accessibility barriers to publishing content on social networks) that need to be addressed and may have an impact in users engagement, other aspects should be considered in order to educate users in accessibility practices:

• Examples to create awareness: Our research showed that most users are not aware of how users with disabilities interact with technologies and social networks. SONAAR should present users of social networks with information about the benefits of engaging in accessibility practices. That means that we should inform users about how media contents are being accessed and consumed by people with disabilities and the impact of creating accessible content.
• Tutorials to guide: Besides having a user-friendly interface, SONAAR should provide users with an easy-to-read and accessible tutorial in which users can find clear information and guidance on how to create accessible media content. This tutorial should provide information on different types of content, social networks and devices.

Conclusions

The findings detailed in this report reinforce the objectives established in the SONAAR project. Most social network users are not aware of the existence of accessibility features as well as their benefits. Besides, current machine-generated alternative descriptions do not provide enough information for users with disabilities to fully understand media content, in particular, in a social network context.

In general, users have shown interest in getting involved in accessibility practices. However, they are not being sufficiently motivated and correctly guided by the main platforms at this moment to the point of overcoming all the barriers present in the current flow of interaction. In order to address most of the current problems, users have to be a part of this process as they play an essential role in improving the accessibility of current media content on social networks. These considerations are well aligned with SONAAR activities concerning the dissemination of accessibility practices and their importance for people with disabilities.

Activities detailed in this report are part of WP3, aiming to engage users in the production of accessible content. The findings described provided us with information on which factors are important to motivate users and how we can educate them on how to properly create accessible content. SONAAR needs to create more awareness among users by informing them about how media content is being consumed by people with disabilities as most users are not aware of how people with disabilities interact with technology and social networks.

Providing users with examples of different ways of accessing and interpreting media can contribute to creating this awareness and engagement. In addition, the SONAAR mechanism should also provide accessible and user-friendly tutorials to guide users who are already motivated to engage in accessibility practices and currently have little information on how to achieve this. From that it will be easier to educate them on how to properly create accessible content on their social networks. Finally, further activities in the scope of WP3 include updating authoring services to include the information gathered, as well as validating its effectiveness with users.

References

[1] Cynthia L. Bennett et al. “How Teens with Visual Impairments Take, Edit, and Share Photos on Social Media”. In: Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems - CHI ’18. Vol. 2018-April. New York, New York, USA: ACM Press, 2018, pp. 1–12. isbn: 9781450356206. doi: 10.1145/ 3173574.3173650. url: http://dl.acm.org/citation.cfm?doid=3173574. 3173650.

[2] Erin Brady et al. “Investigating the appropriateness of social network question asking as a resource for blind users”. In: Proceedings of the ACM Conference on Computer Supported Cooperative Work, CSCW (2013), pp. 1225–1236. doi: 10.1145/ 2441776.2441915.

[3] Jacob Cohen. “A Coefficient of Agreement for Nominal Scales”. In: Educational and Psychological Measurement 20.1 (1960), pp. 37–46. issn: 0013-1644. doi: 10. 1177/001316446002000104. url: http://journals.sagepub.com/doi/10.1177/ 001316446002000104.

[4] Carlos Duarte, Carlos M. Duarte, and Luís Carriço. “Combining Semantic Tools for Automatic Evaluation of Alternative Texts”. In: Proceedings of the 16th Web For All 2019 Personalization - Personalizing the Web. May. New York, NY, USA: ACM, 2019, pp. 1–4. isbn: 9781450367165. doi: 10.1145/3315002.3317558. url: http://dl.acm.org/citation.cfm?doid=3315002.3317558.

[5] Facebook. Facebook Reports First Quarter 2020 Results. 2020. url: https://investor.fb.com/investor-news/press-release-details/2020/Facebook-Reports-First-Quarter-2020-Results/default.aspx.

[6] Facebook. “How do I edit the alternative text for a photo on Facebook”. In: (2020). url: https://www.facebook.com/help/214124458607871.

[7]Cole Gleason et al. ““It’s almost like they’re trying to hide it”: How User-Provided Image Descriptions Have Failed to Make Twitter Accessible”. In: The World Wide Web Conference on - WWW ’19. New York, New York, USA: ACM Press, 2019, pp. 549–559. isbn: 9781450366748. doi: 10.1145/3308558.3313605. url: http: //dl.acm.org/citation.cfm?doid=3308558.3313605.

[8] Cole Gleason et al. “Future research directions for accessible social media”. In: ACM SIGACCESS Accessibility and Computing 127 (2020), pp. 1–12. issn: 1558- 2337. doi: 10.1145/3412836.3412839. url: https://dl.acm.org/doi/10.1145/ 3412836.3412839.

[9] Darren Guinness, Edward Cutrell, and Meredith Ringel Morris. “Caption Crawler”. In: Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems - CHI ’18. Vol. 2018-April. New York, New York, USA: ACM Press, 2018, pp. 1–11. isbn: 9781450356206. doi: 10.1145/3173574.3174092. url: http: //dl.acm.org/citation.cfm?doid=3173574.3174092.

[10] Reeti Mathur and Erin Brady. “Mixed-Ability Collaboration for Accessible Photo Sharing”. In: Proceedings of the 20th International ACM SIGACCESS Conference on Computers and Accessibility - ASSETS ’18. New York, New York, USA: ACM Press, 2018, pp. 370–372. isbn: 9781450356503. doi: 10.1145/3234695.3240994. url: http://dl.acm.org/citation.cfm?doid=3234695.3240994.

[11] Meredith Ringel Morris et al. “Rich Representations of Visual Content for Screen Reader Users”. In: Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems - CHI ’18. Vol. 2018-April. New York, New York, USA: ACM Press, 2018, pp. 1–11. isbn: 9781450356206. doi: 10.1145/3173574.3173633.

[12] Elliot Salisbury, Ece Kamar, and Meredith Ringel Morris. “Toward Scalable Social Alt Text: Conversational Crowdsourcing as a Tool for Refining Vision-to-Language Technology for the Blind”. In: Aaai Hcomp 17 Hcomp (2017), pp. 147–156.

[13] Abigale Stangl, Meredith Ringel Morris, and Danna Gurari. “”Person, Shoes, Tree. Is the Person Naked?” What People with Vision Impairments Want in Image Descriptions”. In: Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems. New York, NY, USA: ACM, 2020, pp. 1–13. isbn: 9781450367080. doi: 10.1145/3313831.3376404. url: https://dl.acm.org/doi/10.1145/ 3313831.3376404.

[14] Twitter. Accessible images for everyone. 2016. url: https://blog.twitter.com/ en%7B%5C_%7Dus/a/2016/accessible-images-for-everyone.html.

[15] Twitter. “Twitter Accessibility”. In: (2020). url: https://twitter.com/TwitterA11y/ status/1265689579371323392.

[16] Violeta Voykinska et al. “How Blind People Interact with Visual Content on Social Networking Services”. In: Proceedings of the 19th ACM Conference on Computer- Supported Cooperative Work Social Computing - CSCW ’16. Vol. 1. New York, New York, USA: ACM Press, 2016, pp. 1582–1593. isbn: 9781450335928. doi: 10.1145/2818048.2820013. url: http://dl.acm.org/citation.cfm?doid= 2818048.2820013.

[17] Shaomei Wu and Lada A. Adamic. “Visually impaired users on an online social network”. In: Proceedings of the 32nd annual ACM conference on Human factors in computing systems - CHI ’14. New York, New York, USA: ACM Press, 2014, pp. 3133–3142. isbn: 9781450324731. doi: 10.1145/2556288.2557415. url: http: //dl.acm.org/citation.cfm?doid=2556288.2557415.

Annex 1: Online survey

Demographic questions

1. Age
2. Gender
   • Female
   • Male
   • Prefer not to answer
   • Other: ___
3. Which operating system do you use (Choose all that apply)
   • Windows
   • MacOS
   • Linux
   • iOS
   • Android
   • Other: ___
4. Do you have some type of disability?
   • Yes
   • No
   • Prefer not to answer
5. Which ones? (Choose all that apply)
   • Blindness
   • Low vision or Visual impairment
   • Cognitive
   • Deafness or Hard-of-hearing
   • Motor
   • Prefer not to answer
   • Other: ___
6. Do you use any assistive technology?
   • Yes
   • No
   • Prefer not to answer
7. Which ones? (Choose all that apply)
   • Screen reader
   • Braille display
   • Screen magnifier
   • Prefer not to answer
   • Other: ___
8. Do any of your family members, friends or acquaintances have a disability?
   • Yes
   • No
   • Prefer not to answer
9. Which ones? (Choose all that apply)
   • Blindness
   • Low vision or Visual impairment
   • Cognitive
   • Deafness or Hard-of-hearing
   • Motor
   • Prefer not to answer
   • Other: ___
10. Are you aware of any of your contacts on social networks (friends, followers, etc.) having a disability?
    • Yes
    • No
    • I don’t know
    • Prefer not to answer
11. Which ones? (Choose all that apply)
    • Blindness
    • Low vision or Visual impairment
    • Cognitive
    • Deafness or Hard-of-hearing
    • Motor
    • Prefer not to answer
    • Other: ___

Social network usage

For the next questions, we will use the term Social Networks as a reference to online platforms used by people to communicate or to build social relationships with other people. Some examples of these online services are: Facebook, Twitter, Instagram, WhatsApp, WeChat, among others (https://en.wikipedia.org/wiki/Social_networking_service#Largest_social_networking_services). We will also distinguish some of the activities carried out on these social networks:

• Post your content refers to sharing new content of your own authorship. For example, posting an image that you created on your Twitter or Instagram timeline as well as sending a message on WhatsApp or Messenger.
• Share content from others refers to sharing content authored by another person. For example, sharing on your Twitter or Instagram timeline an image posted by someone else as well as forwarding a message in WhatsApp or Messenger.

Social networks usage - Access

12. How frequently do you use a smartphone or tablet to access social networks?
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
13. How frequently do you use a desktop or laptop computer to access social networks?
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never

Social networks usage - Posting your content

14. How frequently do you use a smartphone or tablet to post your content on social networks?
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
15. How frequently do you use a desktop or laptop computer to post your content on social networks?
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never

Social networks usage - Sharing content from others

16. How frequently do you use a smartphone or tablet to share content from others on social networks?
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
17. How frequently do you use a smartphone or tablet to share content from others on social networks?
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never

Social networks usage - Access

18. How do you access social networks? (Choose all that apply)
    • Website on mobile device or tablet
    • Website on desktop or laptop computer
    • Mobile app
    • Other: ___

Social networks usage - Access

What social networks do you access and how often do you access it? (Provide information about the 3 social networks you access the most)

19. First most accessed social network: ___
20. Frequency of access:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
21. Second most accessed social network: ___
22. Frequency of access:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
23. Third most accessed social network: ___
24. Frequency of access:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never

Social networks usage - Posting your content

In what social networks do you post your content (text, audio, videos, images, memes or GIFs) and how often do you post it? (Provide information about the 3 social networks you post more often - these can be the same or different from the ones in the previous question)

25. First social network in which you most post your content: ___
26. Frequency of posting:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
27. Second social network in which you most post your content: ___
28. Frequency of posting:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
29. Third social network in which you most post your content: ___
30. Frequency of posting:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never

Social networks usage - Sharing content from others

In what social networks do you share content from others (text, audio, videos, images, memes or GIFs) that you authored and how often do you share it? (Provide information about the 3 social networks where you share with more often - these can be the same or different from the ones in the previous questions)

31. First social network in which you most share content from others: ___
32. Frequency of sharing:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
33. Second social network in which you most share content from others: ___
34. Frequency of sharing:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
35. Third social network in which you most share content from others: ___
36. Frequency of sharing:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never

Social networks usage - Posting your content

Please tell us how often you post your content on social networks, according to the following types of content:

37. Text:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
38. Audio:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
39. Video:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
40. Image:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
41. Meme:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
42. GIF:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never

Social networks usage - Sharing content from others

Please tell us how often you share content from others on social networks, according to the following types of content:

43. Text:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
44. Audio:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
45. Video:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
46. Image:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
47. Meme:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never
48. GIF:
    • Several times per day
    • Once a day
    • Several times per week
    • Several times per month
    • Less than once a month
    • Never

Social networks accessibility practices

Several social networks currently offer their users the possibility of providing alternative text descriptions for visual elements in order to improve the understanding of these contents by people with visual disabilities.

49. For this question consider the last three posts you have authored with any kind of media content (i.e.: audio, videos, photos, memes and GIFs). Did you provide an alternative text description for any of these media contents?
    • Yes
    • No
50. Why didn’t you provide an alternative text description for any of these media contents? (Choose all that apply)
    • I didn’t knew it was possible
    • I don’t know how to write a suitable alternative text description
    • I don’t know where to write an alternative text description
    • It’s too time consuming
    • Other: ___
51. In which social network? (Choose all that apply)
    • Facebook
    • Twitter
    • Instagram
    • Other: ___
52. How did you provide an alternative text description for any of these media contents? (Choose all that apply)
    • Writing the description in the text of the post
    • Using the functionality provided by the social network
    • Using an automatic service for alternative text generation
    • Other: ___
53. Why do you think social network users do not provide alternative text descriptions for the media content they post or share? (Choose all that apply)
    • They don’t know it is possible
    • They don’t know how to write a suitable alternative text description
    • They don’t know where to write an alternative text description
    • They consider it is too time consuming
    • They don’t think it has any impact
    • Other: ___
54. Besides improving access and understanding of the content by people with disabilities, what are or would be your motivations for posting accessible media content on social networks?
55. When considering posting media content do you take any additional steps to enhance the content accessibility?
    • Yes
    • No
56. Which ones?
57. In which social networks? (Choose all that apply)
    • Facebook
    • Twitter
    • Instagram
    • Other: ___
58. Are you aware of if your family, friends or acquaintances have any concerns about accessibility when posting or sharing content in social media ?
    • Yes
    • No
    • I don’t know
    • Prefer not to answer
59. Which ones?
60. Would you like to have access to a tool that would support you in the authoring of accessible content?
    • Yes
    • No
61. Which features would you include?
62. Do you have some additional thoughts or comments about the subject that you would like to share?
63. The next phase of this research will consist of a second study where participants will be invited to describe in greater detail the difficulties felt when posting any kind of digital media, taking into account the accessibility of this content. If you are interested in participating in this next stage of the study, please provide us with your email address so we can provide you with further details:

Annex 2: Code list

The following table presents the list of codes resulting from the analysis of transcribed interviews and from the open answers provided in the survey conducted. The first column of this table represents the Level 1 codes, which have a greater scope, while second column represents Level 2 codes associated with it, in a more specific context.

Table 1: Code List

• User
  • Disabled
  • Non Disabled
• Operating system
  • Android
  • iOS
  • MacOS
  • Windows
• Disability
  • Blindness
  • Cognitive
  • Colorblindness
  • Deafness/hard of hearing
  • Low vision
  • Motor
• Assistive technology
  • Braille display
  • Dyslexic plugin
  • Larger font size
  • Screen magnifier
  • Screen reader
  • Text-based web browser
  • Text-to-speech
• Device
  • Desktop
  • Laptop
  • Smartphone
  • Tablet
• Interface
  • Native App
  • Website
• Social Network Platform
  • Facebook
  • Flickr
  • Instagram
  • Linkedin
  • Messenger
  • Nextdoor
  • Parler
  • Reddit
  • Slack
  • Snapchat
  • Soundcloud
  • Telegram
  • Third-party clients
  • Tiktok
  • Twitter
  • Whatsapp
  • Youtube
• Content
  • Audio
  • Emoji
  • Gif
  • Image
  • Link
  • Location
  • Meme
  • Sticker
  • Story
  • Text
  • Video
• Feeling
  • Being obnoxious
  • Exclusion
  • Fascination
  • Frustration
  • Looking stupid
  • Personal
  • Stigma
  • Utopic
  • Warm
• Who
  • Acquaintance
  • Developers
  • Government
  • Other people
  • Platforms
  • Self
• Reason
  • Don't care
  • Don't know how to
  • Forget
  • Interface updates
  • Not a requirement
  • Privacy
  • Time consuming
  • Unnecessary
• Motivation
  • Additional context
  • Alternative access option
  • Creative ability
  • Develop skills
  • Evolving AI solutions
  • Having a disability
  • Inclusion
  • It's the right thing to do
  • Language learners
  • Platform support
  • Reach more people
  • Required
  • SEO
• Barrier
  • Can't access URL feed
  • Don't know what to write
  • Color contrast
  • Hard-to-find ...
  • Image-only posts
  • Lack of ...
  • Loading speed
  • Not accessible
  • Not user-friendly
  • Scroll speed
  • Unawareness
• Accessibility strategy
  • Check reposts
  • Color compliance
  • Limited use of...
  • Mobile version
  • No flashy/rapid GIFs
  • Plain language
  • Video captions
  • Video limited duration
  • Ask for assistance
• Hashtag
  • Camel case
  • Indicative
• Feature
  • API support
  • Browser extension
  • Color filter
  • Content warning
  • Linked accounts
  • Personalized accessibility settings
  • Preview
  • Visualization Group
• Alternative description
  • Audio
  • Automatic
  • Content post
  • Field
• Suggestion
  • Emphasize accessibility features
  • Raise awareness
  • Requirement
  • Responsibility
• Other sites
  • Amazon
  • Netflix
• Games
• Legal