Use of social networking sites, such as Facebook, is rapidly expanding, but people with intellectual disability are at risk for exclusion because sites like Facebook are not designed for cognitive access. The purpose of the present study was to describe the development and initial testing of a cognitively accessible prototype interface for Facebook, called Endeavor Connect, that was designed to support independent Facebook use by people with intellectual disability. The performance of young adults with intellectual disability when completing five common Facebook tasks was compared when using the Endeavor Connect and Facebook interfaces. Results suggest that, when using Endeavor Connect, young adults with intellectual disability completed more tasks independently with fewer errors and required fewer prompts. Implications for research and practice are discussed.
The use of social networking sites, such as Facebook, is rapidly expanding throughout the world. As of June 2013, 1.15 billion people worldwide were monthly active users of Facebook, and 699 million people were daily active users (Facebook, 2013). Although there are debates about the social implications of the increased use of social networking, a growing body of research suggests that engagement with Facebook can potentially enhance social capital (Brandtzæg, 2012; Valenzuela, Park, & Kee, 2009). Social capital can be defined as “features of social organizations such as networks, norms, and social trust that facilitate coordination and cooperation for mutual benefit” (Putnam, 1995, p. 67). Social capital enables individuals, through their social networks, to connect and potentially cultivate information and resources that enable them to gain economic, educational, and vocational advantages (i.e., bridging social capital) and to gain emotional and physical support (i.e., bonding social capital). Researchers have found that Facebook can impact bonding social capital by facilitating direct communication through the posting of comments and messaging, typically with individuals the users also know offline, and bridging social capital (where ties tend to be weaker) through browsing profiles and liking photos and posts (Wilson, Gosling, & Graham, 2012).
Researchers have also found that people with intellectual disability all too often have limited social networks that severely restrict their opportunity to build social capital. Clement and Bigby (2009) noted that “typically, people with intellectual disability have small, highly restricted social networks characterized by interactions with other people with intellectual [disability], family members, and paid workers” (p. 264). Building social networks and community connections, resources, and supports, however, remains one of the most frequently identified goals of people with intellectual disability and their families (Kampert & Goreczny, 2007; Shogren, 2012). With the growing role of technology and social networking sites like Facebook in shaping access to social capital, people with intellectual disability are at even greater risk for exclusion because of difficulties with technology access and use. Despite the ubiquity of technology in the day-to-day life of most people in society, evidence over the past decade suggests that people with intellectual and developmental disabilities too often have limited access to technology and that technology is underutilized by this population for a variety of reasons (Carey, Friedman, & Bryen, 2005; Palmer, Wehmeyer, Davies & Stock, 2012; Tanis, Palmer, Wehmeyer, Davies, & Stock, 2012; Wehmeyer, Palmer, Smith, Davies, and Stock, 2008), primarily because the majority of technologies are not designed to be cognitively accessible.
Advocates in the field of applied cognitive technologies have recently developed and endorsed a statement titled “The Rights of People With Cognitive Disabilities to Technology and Information Access” (Braddock, Hoehl, Tanis, Ablowitz, & Haffer, 2013), suggesting the criticality of ensuring that people with cognitive disabilities have “access to comprehensible information and usable communication technologies” (p. 98) that use principles of universal design and follow best practices in the development of innovative applied cognitive technologies. As a field, applied cognitive technologies refers to “research and development to provide technology supports that enable people with cognitive disabilities to successfully function in typical environments, to increase participation in tasks and activities in typical environments, and to promote social inclusion, self-determination, and an enhanced quality of life” (Wehmeyer & Shogren, 2013, p. 92). Applied cognitive technologies focus on electronic and information technologies that are used by all people, and their application to enabling people with cognitive disabilities to access typical living, work, education, and community environments.
Researchers working in the field of applied cognitive technologies have applied best practices related to universal design to the design of systems to support access to public transportation and community access (Davies, Stock, Holloway, & Wehmeyer, 2010; Mechling, Gast, & Langone, 2002), cell phones (Stock, Davies, Wehmeyer, & Palmer, 2008), employment (Furniss et al., 2001; Taber, Alberto, & Fredrick, 1998), the Internet and web browsers (Friedman & Bryen, 2007; Lancioni, Van den Hof, Furniss, O'Reilly, & Cunha, 1999; Wehmeyer et al., 2011), portable reading systems (Davies, Stock, King, & Wehmeyer, 2008), education (Ayers & Langone, 2008; Bouck, Taber-Doughty, Flanagan, Szwed, & Bassett, 2010), and money management tools (Davies, Stock, & Wehmeyer, 2003). Yet, such principles have not been widely applied or applied to the development of tools to support access by people with intellectual disability to social networking sites, such as Facebook. Facilitating access to Facebook for people with intellectual disability is important given its widespread use and the growing body of data suggesting its impact on social capital. The purpose of the present study was to describe (a) the iterative design and development and (b) initial pilot testing of the usability of a cognitively accessible interface for Facebook, called Endeavor Connect, developed to support independent Facebook use by people with intellectual disability.
Twelve adults with intellectual disability were recruited from the Colorado Springs School District 11 transition program, as well as Mosaic and Carmel Community Living, two local adult service provider organizations. The average age of study participants was 29.5 years old, with a range from 20 to 45 years old. The sample included five females and seven males. The average IQ score for the group was 53.5, with a range from 38 to 66. None of these participants had a Facebook account and, thus, were not Facebook users.
The usability testing took place over a 2-week period at AbleLink's Cognitive Innovation Lab in Colorado Springs using a Gateway All-in-One touchscreen computer with a 21.5 inch display. The 12 participants received training and were observed when engaging in five Facebook-specific tasks when in the mainstream Facebook program and again when using the Endeavor Connect interface. Differences in the: a) number of scripted Facebook tasks successfully completed, b) the amount of assistance required to successfully complete each task, and c) the number of errors made in completing each task were examined across Endeavor Connect and Facebook. The five common Facebook tasks included: a) reading/comprehending an incoming post, 2) making an outgoing post, 3) posting a picture, 4) navigating to a designated friend's Facebook page, and 5) navigating back to the “Home” page. Each task was analyzed to determine the steps necessary to complete the activity under both conditions.
The order in which each participant interacted with Facebook and Endeavor Connect was randomized to attempt to control for learning effects. Each participant received training on how to complete each of the five tasks in Facebook or Endeavor Connect immediately preceding their interaction with the respective system. Training involved a demonstration of each task followed by one hands-on walk-through of the common Facebook tasks to allow participants to engage in each condition with as much prompting support as needed.
After training was completed, each participant interacted with Facebook and Endeavor Connect, respectively, and data were collected on the three dependent measures using a standard data collection form. Data collectors were trained on use of the data collection instrument and participated in mock sessions with project staff to help identify and address how to score various actions. To limit potential frustration and feeling of failure on the part of research participants with intellectual disability, the number of prompts required and errors made for each task was capped at three each. Therefore, if a study participant required more than three prompts or made more than three errors, the task was scored as not completed and research staff demonstrated how to complete the task before moving on to the next task.
After completing the task sequences in Facebook and Endeavor Connect, a brief interview was conducted with each participant to identify their perceptions of ease of use and preferences for Facebook or Endeavor Connect.
Prior to and during the design and development of Endeavor Connect, informal interviews were conducted with people with intellectual disability, their families, support providers, and administrators to determine their perceptions of needs related to using Facebook and, as the Endeavor Connect prototype was designed and developed, the effectiveness of specific supports to promote access to and use of Facebook. These informal interviews shaped the design and development process.
The first phase of the design and development activities involved identifying initial system requirements for Endeavor Connect. These activities included (a) identifying best practices in the design of applied cognitive technologies that support independent computer use by people with intellectual disability; and (b) reviewing current technical developments in Facebook features, use statistics, and privacy features/policies to identify specific features needed in Endeavor Connect to interface with Facebook.
Identifying best practices
Previous research (Davies et al., 2008; Davies et al., 2010; Davies, Stock, & Wehmeyer, 2001, 2004; Stock, Davies, Wehmeyer, & Lachapelle, 2011; Stock et al., 2008) conducted by the research team over the past 15 years on the development of applied cognitive technologies was reviewed to identify best practices in the design of applied cognitive technologies to guide the development of Endeavor Connect. Several best practices were identified, including:
Combining use of pictures and system-generated audio prompts for navigation to provide the user with greater independence and self-direction.
Using familiar voices for audio prompts (i.e., a parent, a sibling, friend, or support provider) to promote enjoyable and compelling use of the software.
Using clear, uncluttered interface designs, such as oversized picture-buttons, to maximize the effectiveness of touchscreens used for desktop, tablets, and smart phone devices and to support people with physical disabilities.
Using interface designs that utilize consistency and repetition to promote independent use.
Using “error minimization” techniques, such as removing buttons from the screen when their use is inappropriate, unnecessary, or distracting to limit the possibility for errors.
Including a complete set of user-controlled customization options to provide support providers with the ability to adapt the available features and interface to meet the unique needs of different end users with intellectual disability.
Building in assessment and management systems to support tracking of use/progress.
Identifying Facebook-specific features needed in Endeavor Connect
As described previously, Facebook was one of the first social networking tools developed and currently has over a billion users worldwide. Since September 2006, anyone over the age of 13 with a valid e-mail address can become a Facebook user. Users can engage in a variety of activities, such as adding friends; posting messages, photos, and videos; playing games; signing up for Facebook apps; and updating personal profiles to notify friends about themselves and their lives. Facebook users can also join groups created by companies, government entities, nonprofits, social causes, employers, news outlets, schools, and many other organizations. Facebook includes both personal privacy and security features (e.g., user login, ability to permit or withhold public access to items posted to a user's Timeline or Home page, limiting availability of content to approved individuals). Requirements development work included identification of potential strategies and feature options that may be important in future iterations of the Endeavor Connect system to enhance the mainstream Facebook privacy and security features. For example, a potential option that can be selected by the end user may include the ability for things like incoming friend requests or even day-to-day outgoing posts to be automatically screened via a third-party email notification to a chosen parent or caregiver prior to being presented to the Endeavor Connect user for consideration. These types of important feature options, though identified during this project, were reserved for future research and development.
Based on a review of the Facebook Platform (Facebook Developers, 2010) and interviews with key stakeholders, specific technical objectives to enable Endeavor Connect to interface with Facebook were also identified. Although the scope of the project did not allow for all of these requirements to be developed and evaluated, they included:
Developing the interface to have clean, uncluttered screens and multimedia techniques to promote independent usability.
Supporting the most widely used Facebook features such as posting to the Home page or Timeline, video posting, managing friend requests, and accessing games and apps as user-controlled options with an interface “look and feel” similar to that of the mainstream Facebook interface.
Developing a set of user-controlled safety and security options beyond those already provided by Facebook (such as third-party approval of friend requests) to provide additional supports for security and protection against bullying and other potential social networking hazards.
Developing a configuration interface to allow support providers to activate or deactivate security options and access to various Facebook elements such as apps, games, friend requests, and so on.
Developing cross-platform versions of the system to enable operation on a wide variety of hardware platforms (Windows, iPhone, iPad, Android tablets and smart phones, and other mobile devices).
Developing accessible online multimedia training materials for both support providers and end users with intellectual disability to learn how to implement and utilize the Endeavor Connect system to safely access Facebook.
Consequently, Endeavor Connect was designed to provide a method of accessing existing Facebook accounts, but without needing to use the mainstream Facebook user interface. The prototype interface was designed such that account preferences and security features set up in Facebook will carry over to Endeavor Connect. The Endeavor Connect interface was developed to mimic the mainstream Facebook Home page interface, except that many of the options, features, menus, and other clutter that require high levels of cognition were removed to provide an appealing but simplified appearance. Endeavor Connect was designed to support the most frequently used features of Facebook: accessing Home page posts, creating and posting on the Timeline or Home page, posting pictures, navigating to friend's pages, and navigating back to one's Home page.
When first launching Endeavor Connect, the system provides a login prompt. To avoid the barrier of having to log in each time, an option to retain the user's login information is provided so that on subsequent startups the system goes directly to the accessible version of the end user's Home page in Endeavor Connect. Figure 1 provides a screen shot showing the appearance of the Home page in Endeavor Connect for an end user after logging in. The posts on the Home page in Endeavor Connect, as shown in Figure 1, are pulled directly from, and are therefore identical to, the posts on the end user's screens in the mainstream version of Facebook. In other words, Endeavor Connect displays all things posted by friends in Facebook, regardless of whether the end user is in Facebook or using the Endeavor Connect interface. Similarly, anything posted by an end user in Endeavor Connect is also posted to approved friends' mainstream Facebook pages.
At the top of the screen in Figure 1 are two optional toggle buttons that can be used to increase the volume and text size, as needed. Each post—either by the end user or by a friend of the end user—is displayed chronologically along with one or two buttons to the right of each post. The green speaker button, which is present for every post, can be clicked to initiate a built-in text-to-speech feature that reads the post aloud. Special programming was done so that the text-to-speech system first identified the person making the post before proceeding to read it. For example, the post that is at the lower part of the screen in Figure 1 would be read as: “Chet Kincaid said ‘My favorite color is blue.’” This feature was designed to allow non- or low-readers to better comprehend incoming text posts. The Like button (i.e., the blue “thumbs up” button) is available for any post on the end user's Home page or Timeline and allows the end user to “like” posts on their Home page or Timeline. If “Like” is selected for any post in Endeavor Connect, it is also transferred to the friend's mainstream Facebook Home page or Timeline. Users can also click on any images posted to see them at nearly full screen and print them, if desired.
In terms of making posts in Endeavor Connect, two options were developed. Tapping the button with the keyboard letter “T” near the top of the screen (see Figure 1) opens a simple dialog box, shown in Figure 2, with a system-generated audio prompt that states “Please type the message that you would like to post.” After entering text, the “Post” button is selected to complete the post, or the “Cancel” button can be used to abort the text post. The second method for posting in Endeavor Connect allows for audio-recording of messages, eliminating the need for typing. By clicking on the microphone button (shown at the top of Figure 1), a display opens and a system-generated audio message is played, prompting the user to “Please record your Facebook message after the beep.” This is immediately followed by the familiar beeping sound. This display is shown on the left side of Figure 3. After finishing the message, the user clicks the “Done” button and the display shifts to that shown on the right side of Figure 3, with the system-generated audio prompt to “Click the blue Post button to post this message to Facebook, or click Cancel to start over.” This allows users to easily re-record their audio post in the event of misspeaking. Audio posts appear in the user's Home page in Endeavor Connect with the green speaker button, and the recorded message plays when the green speaker button is clicked, as shown in the audio post at the bottom of the screen in Figure 4. In Facebook, audio posts appear as an audio link that, when clicked, plays the audio message using the friend's default media player.
Another commonly used feature in Facebook is posting pictures, and by clicking the camera icon at the top of the screen (see Figure 1) users can initiate a sequence to post digital images to Endeavor Connect and, thus, Facebook. After clicking the camera icon, the box shown on the left side of Figure 5 opens. This box displays all images that have been placed in a designated folder (the folder can be set up by the end user or a support provider in the Endeavor Connect options menu). A system-generated audio message prompts the user to “Please select the picture that you would like to post to Facebook.” Upon selecting the desired image, it is displayed in the window on the right of the screen and the system prompts the user to “Select the Post button to post this picture to Facebook, or press Cancel to start over.” Users can click on any number of pictures while searching for the desired image before selecting the Post button. Posted images appear on mainstream Facebook pages, just like all other images posted to Facebook, but appear somewhat larger when displayed in the prototype Endeavor Connect version (see Figure 4 for an example of how a picture appears on the Endeavor Connect Home page). Of note, there is also a green speaker button to the right of the posted picture that, when pressed, will read the name of the individual who posted the picture.
The final features to be noted are along the right side of the screen in Figure 1. The top button (house icon) can be used to return to the user's Home screen, which displays some of the personal information they have provided and mimics the Timeline feature in Facebook. The second button (two silhouettes with talking bubbles) returns the user to their Home page, which displays the user's personal profile. The third button (three human silhouettes) brings up the end user's list of friends, including their pictures. This is shown in Figure 6. Clicking on a friend's pictures takes the user to their friend's Home page, which has the same format as the end user's Home page (see Figure 1).
Design and Analysis
The purpose of the pilot test reported in this article was to examine the usability of the accessible interface of Endeavor Connect, as compared to the mainstream Facebook interface. Specifically, differences in the a) number of scripted Facebook tasks successfully completed, b) amount of assistance required when successfully completing each task, and c) number of errors made in completing the common Facebook tasks were examined when individuals with intellectual disability used Endeavor Connect and Facebook. Data collection provided information on the number of steps correctly completed, the amount of assistance required to complete the tasks, the number or errors made, and the success rates for task completion using Facebook and Endeavor Connect. Notes were also taken on each participant's interaction with each system. The results were analyzed descriptively, to explore the degree to which participants completed more tasks successfully in Endeavor Connect, and empirically, using Sandler's A-statistic, to examine mean differences in the three dependent measures. Interview data was also summarized to provide information on participants' perceptions of their experiences with the two platforms.
Eleven participants were able to complete all five Facebook tasks with three or fewer prompts or errors per task when using Endeavor Connect, while only four participants were able to do so while using the mainstream Facebook interface. When comparing the average tasks completed independently, Endeavor Connect led to significantly more independent completion, (average across participants of 4.6 vs. 3.8 out of 5 tasks completed; p = .006). It appeared when observing participants that the overall complexity and screen clutter, in addition to difficulties with reading or writing, made it difficult for many participants to use Facebook by itself. When using the mainstream Facebook interface, participants would often click on incorrect screen elements or stop using the system and ask for help on how to proceed. The amount of information shown on a mainstream Facebook page at times required multiple verbal and gestural prompts from research staff for completion of a single task.
Relatedly, in terms of the number of errors, some participants tended to click on elements on the screen even when they appeared to be uncertain of how to complete a task. The most common error was selecting the incorrect interface element (e.g., selecting “Post” instead of “Friends”). Several participants repeatedly used the steps that had been successful in a previous task, even if they were not relevant to the present task. Other participants, however, tended to stop interacting with Facebook when they were uncertain and asked research staff for help. This led to variability in the number of errors across participants, but, overall, participants had fewer errors per task when working with Endeavor Connect (average .05 vs. .30; p = .013). There were some specific errors that were the most common within Endeavor Connect, however. For example, several participants selected the “Cancel” button instead of the “Post” button when posting audio messages or digital pictures; this appeared to be a result of the system-generated audio that prompted users to “Select the Post button to post this picture to Facebook, or press Cancel to start over.” Some individuals appeared to cue into the last part of that message. Therefore, alternative settings will be developed for the final version that allow the system to be set up to simply say “Select the Post button to post this picture to Facebook” or to use further verbal identifiers in the system-generated prompt, e.g. “Select the blue Post button to post this picture to Facebook, or press the red Cancel button to start over.”
Finally, in terms of the number of prompts, more participants asked for help or simply waited for a prompt instead of engaging with Facebook or Endeavor Connect, as shown in the higher rates of average prompts to average errors per tasks, although participants still needed significantly fewer prompts when using Endeavor Connect (average .22 vs. .58; p = .024). The most challenging task across systems for participants was posting pictures, and many participants needed step-by-step prompts when using the mainstream Facebook interface. When interacting with Facebook alone, many participants also asked for help when text was involved, particularly typing a post. Several participants appeared uncomfortable or nervous when asked to type a post in Facebook and quickly reached the maximum number of prompts or errors before the task was discontinued. Fewer issues were encountered when participants were able to audio record their voices and hear the playback from posts in Endeavor Connect.
Overall, when asked which system they preferred, participants generally reported that Endeavor Connect was easier to work with (“This one [Endeavor Connect] is probably easier for me”). They also reported liking to record and hear their voices (“I like recording my voice better, it's easier” and “I like recording my voice for people”) and to have Endeavor Connect read the text aloud (“It told me what it said”). Perhaps most importantly, participants identified ways that Endeavor Connect could support their social networking (“I'd like to talk to my dad and my friends on here” and “I could use this to talk to my friends”). During the study, one participant spontaneously recorded a post saying “I would like to go to the movies with a friend this weekend.” When reminded of the purpose of the study and that her use was just “pretend,” the participant stated “I know, but that is what I would have said.”
The design, development, and pilot testing of the Endeavor Connect interface suggests that a specially designed, cognitively accessible interface to Facebook provides an effective means to support greater access to the social networking site by individuals with intellectual disability. Given the ubiquity of Facebook in society and research suggesting its potential impact on the building of social capital in the general population (Wilson et al., 2012), creating meaningful ways for people with intellectual disability to access Facebook is necessary to promote meaningful inclusion in society, a fundamental right of people with disabilities, and to create opportunities for economic, educational, and vocational gains.
By using best practices in the design of applied cognitive technologies, including principles of universal design and the use of error minimization techniques, personalization and customization, and reduced screen clutter, as well as insights into the most common uses of Facebook and the preferences of people with intellectual disability and their systems of support for using Facebook, the technical team was able to develop an effective interface that retrieves information from Facebook (and delivers it back) and displays it in a format accessible to people with intellectual disability. The data from the pilot test suggested that people with intellectual disability are more able to be successful when navigating Endeavor Connect (vs. Facebook) and make fewer errors and require fewer prompts. Further, people with intellectual disability report enjoying the features in Endeavor Connect, particularly features not readily available in Facebook, such as recording messages to post.
Further research is needed that moves beyond pilot testing and more systematically examines the degree to which people with intellectual disability are able to independently utilize the features of an accessible social networking program, like Endeavor Connect, over time. Such testing should also address the limitations inherent in the pilot testing, such as the small number of participants and the use of the same participant group to examine the usability of Facebook and Endeavor Connect. Additionally, the pilot testing conducted in this study was brief, and examination of the degree to which people with intellectual disability are able to sustain independent use of Endeavor Connect over time is needed. The impact of Endeavor Connect on relationships, online and offline, and bridging and bonding social capital should also be examined to evaluate general concerns that the use of social networking may under some measures actually increase social isolation in terms of face-to-face human contact. Further, work is also needed to examine effective strategies to balance independence and potential risks in the use of Facebook. Facebook itself has security features available, such as limiting who can see certain types of posts, and additional features were identified for Endeavor Connect, such as the ability to set up third-party access and review of posts, friend requests, and other critical features that need to be developed and evaluated in future research. The most appropriate and respectful ways to support people with intellectual disability to balance the risks inherent to sharing information publically on the Internet (e.g., cyberbullying, exploitation) need to be further examined. There is a general concern in society for abuse of photos, predatory behavior, and cyberbullying within the realm of social media use, but it appears that, despite these concerns, people will continue to grow their use of social media. The potential for additional technological precautions and protections that may be enabled by the Endeavor Connect approach is a very important area to consider for further research.
Technology-mediated social networking through sites like Facebook is clearly becoming a part of day-to-day life for the majority of people in society. Promoting access to such opportunities is necessary to ensure the full participation of people with disabilities in all aspects of society, congruent with their preferences, interests, and desires. This study demonstrated that it is possible to create a cognitively accessible interface design, using proven cognitive design methodologies, that parallels the overall appearance and functionality of Facebook but is optimized to meet the needs of people with intellectual disability, including graphic representations, computer-generated audio prompts, and error minimization techniques. Further work is needed to examine the usability of Endeavor Connect, strategies to promote access to Facebook to people with intellectual disability, and the creation of a range of supports and support tools that facilitate the use of technology by people with intellectual disability that are embedded in mainstream applications.
This article is based upon work supported by the U.S. Department of Education under Project Number H133S110021. Any opinions, findings, conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the views or policies of the Department of Education. AbleLink Technologies would like to thank those individuals who volunteered to participate in the pilot study and other testing activities conducted in Phase I of this project. Without their help, the project could not have been accomplished. Additionally, appreciation goes out to the staff at Mosaic, Carmel Community Living, and the Colorado Springs School District #11 Transition Program who provided opinions and feedback and otherwise facilitated various tasks in the project.
Daniel K. Davies, Steven E. Stock, Larry R. King, and R. Brian Brown, AbleLink Technologies, Inc.; and Michael L. Wehmeyer and Karrie A. Shogren, University of Kansas.