UX researcher, Web Accessibility Evaluator
UX Research
Usability Testing
Test Moderation
Accessibility Evaluation
Shuhua Chen
Yutsen Cheng
Javelynn Haryono
Natalie Sroisudaratna
Haley Birch
Oct 2022 - Nov 2022
In March 2021, YouTube launched a new feature, “Shorts,” to compete with TikTok and keep up with the demand for short-form videos during the pandemic. The interface of playing “Shorts'' videos is different from regular long videos on YouTube, which adds to the complexity of the YouTube experience. Since there is little research on this new feature, we aimed to evaluate the usability of ‘Shorts’ to fill this gap. The goal of the project is to conduct a thorough usability evaluation of YouTube, with the aim of providing recommendations to improve accessibility and better meet the needs of individuals who are blind or visually impaired.
YouTube has since risen to be the second most visited website in the world, with more than 2.3 billion monthly users. The top three reasons for using YouTube are “need for help”, “to be entertained”, and “to learn something new.” It has also grown in use for individuals with disabilities. Despite being the second most visited website in the world, video content on YouTube is often inaccessible to individuals with disabilities, specifically the blind and visually impaired.
In a recent study, Hafiar et al. aimed to understand how visually impaired teenagers used the internet in everyday activities. They found that teenagers with disabilities used the internet similarly to teenagers who were sighted. The authors hypothesized that the reason why YouTube became so highly utilized was the potential for audio interaction. Senjam and Primo reported a similar result, finding that visually impaired individuals had similar levels of mainstream app use when compared to sighted individuals. Commonly used apps by their participants included YouTube for entertainment and accessing COVID-19 information. However, it should be noted that the visually impaired may face difficulties while using YouTube.
Image by rawpixel.com on FreepikMany videos on YouTube are often found to have missing alternative text descriptions and inline audio descriptions or extended audio descriptions. Solutions that address these problems include services like “YouDescribe” which enables volunteers to add audio descriptions to YouTube videos. However, inexperienced volunteers did not always produce high-quality and informative descriptions and they struggled to keep pace with newly uploaded videos. Furthermore, since YouTube does not require content creators to provide alternative text descriptions or inline audio descriptions, individuals who are blind and visually impaired struggle to consistently find content they desire. Accessibility needs to be emphasized as the creation and distribution of user-generated videos continue to grow.
Image by vectorjuice on FreepikThe competitive forces, intensity, and attractiveness of the outdoor recreation application industry can be summarized using the Porter Model framework as follows:
● In Liu et al. study of YouTube use, reported that users who were blind used trial and error approach to determine if a video was accessible to them which was time-consuming.
● Kadakia et al. found that content creators on YouTube lacked care or knowledge to include accessibility keyword terms resulting in barriers for users who were visually impaired.
● Senjam and Primo study of YouTube navigation with screen readers, found that participants who were blind reported several problem issues with navigation.
● In a recent 2022 study, several participants with visual impairment had trouble getting used to new interfaces with updates on YouTube.
Ten participants were recruited for the study, five of whom were blind and the other five were sighted. The blind participants were recruited through personal networks, while the sighted participants were recruited via the DePaul University CDM/COMM Participant Pool. All participants were required to be at least 18 years of age. Blind participants were permitted to use any assistive technology during the study.
Prior to conducting the usability test, participants who were blind were sent a consent form via email and the meeting date was confirmed. During the test, we provided participants with a brief description of the study and the usability test that would be conducted, and requested verbal consent from them. We also informed them that a video of their screen would be recorded. Following this, we asked participants preliminary questions about their experience with YouTube. The participants were then given five scenarios and tasks to complete for the study, as outlined below.
It is the holiday season again, and you are excited to gather with your friends for Friendsgiving in two weeks. You have been asked to help prepare this year’s Turkey dish, but since you do not cook often, you are unfamiliar with where to start. You decide to visit the YouTube website and search for a Thanksgiving turkey recipe tutorial.
You have found the Thanksgiving turkey recipe tutorial. You want to write down the grocery shopping list, but you cannot remember all the ingredients. Where would you go to look for the ingredients list?
Your friend was talking about a new feature on the YouTube website called “Shorts” that she believes you would enjoy. You are interested in using this new feature. Where would you go to find this feature?
You like Chris Pratt mimicking a British accent in The Graham Norton Show, and you want to watch it again. Since you do not want to rewatch the entire show, you are looking for a “Shorts” video about this clip.
Now that you have found the “Shorts” video of Chris Pratt mimicking a British accent, you can start watching this video. After finishing the video, you want to find out what others thought about it in the comments section.
We consolidated our data by reviewing the screen recordings and notes documented during the session. Our metrics encompassed the achievement rate for each task, the time taken to accomplish each task, and the participant's satisfaction rating. We uploaded our notes to a Miro board, containing participants' quotes and our research team's observations. Subsequently, we employed inductive coding and utilized the affinity diagram to identify shared themes. We computed the completion time for the successful task and juxtaposed the success rate between participants who are visually impaired and sighted participants.
After analyzing the results from the tests with participants who are blind and sighted participants, we identified the following findings.
Blind participants took longer to complete tasks related to the Shorts feature (tasks 3 to 5), with one out of five participants failing each task
All participants who are blind mentioned the requested tasks are considered easy. However, as they are unfamiliar with the feature or webpage, they would need to spend more time navigating the website sequentially on their first try to understand where to find specific content or features to complete the task. In comparison, sighted participants could quickly scan the web page to complete the tasks even though they may not be familiar with the layout or feature.
All participants were asked to rate each task in terms of their difficulty level for completion, from 1 being very easy to 5 being very difficult
The introduction of two video content formats on YouTube, including Shorts, has complicated the user experience for both sighted and visually impaired participants. Shorts, which is a new immersive feature that delivers mobile-first, bite-sized video content, presents a steep learning curve, particularly for those who are visually impaired, as it differs significantly from regular YouTube videos. The feature aims to engage users by auto-playing videos on a loop unless manually paused, which has caused confusion for some users. During task 4, participants, both sighted and visually impaired, were unsure whether to use the global search bar on the YouTube homepage to search for the Shorts video, with some mistakenly clicking on the Shorts feature first.
All blind participants faced challenges accessing video player controls for both Shorts and regular videos. The automatic play feature of Shorts videos made it difficult for them to hear the screen reader. Participants expressed frustration and difficulties in using shortcut keys, such as pausing the video, with P5 mentioning the need to pop out the video to stop it. Additionally, P5 had trouble fast-forwarding and rewinding regular videos. In contrast, all sighted participants could navigate the controls with ease using the mouse.
Participants who are blind encountered difficulties when navigating the YouTube website. They had trouble identifying errors and troubleshooting when they got lost on the page. The shortcut keys worked differently on the Shorts and regular videos page, which confused participants. They expected a similar shortcut key experience when browsing content on the same website. Some participants accidentally played new Shorts videos while trying to explore the screen freely on the Shorts video page. Furthermore, participants had difficulty locating the Shorts feature using screen reader shortcuts.
The study found that inaccessible videos can be frustrating for participants who are blind and less appealing for sighted participants. Four blind participants failed to find the ingredients list in Task 2 because the content creator did not provide it in the description or transcripts. One sighted participant suggested that the creator should have included the list in the description, while another participant shared an app called YouDescribe that generates audio descriptions and helps individuals who are blind understand video content better.
Based on our findings, we came up with five main recommendations to make the YouTube website more accessible.
One of the challenges for participants who are visually impaired was the website became more complicated when there were two types of video players.
Most blind participants had difficulty pausing videos due to malfunctioning shortcut keys in Shorts videos and forgetfulness in the regular videos. This finding supports Seo and Jung's study's first category of accessibility issues.
Three of the participants who are blind could not identify which videos were Shorts. Two other participants who are blind stated that identifying these videos needed time and full attention.
The majority of participants who were visually impaired experienced difficulty in navigating the website, particularly when confronted with unexpected errors. The challenges related to navigation and the location of new features provide further corroboration for the results of a study conducted by Senjam and Primo in 2022.
It was observed that both participants with visual impairment and sighted participants resorted to the description section of regular videos to seek primary information. Participants with visual impairment found it challenging to determine the accessibility of the video without appropriate information in the description section.
Blind participants had similar YouTube usage to sighted individuals, but they had lower computer usage to access the website. This difference in familiarity may have biased our data.
What we did well?
We successfully learned and implemented the use of screen readers during the usability test, showcasing our adaptability, self-teaching abilities, and great team collaboration.
What could be improve?
Collecting demographic information, such as age and location, could have provided additional insights into usage patterns. Technical issues during interviews, including sound and connectivity problems, posed challenges and caused delays. Conducting in-person testing would be advantageous in mitigating any issues related to Zoom functionality. Due to the small sample size, establishing statistical significance for some findings is challenging. Increasing the sample size in future research can address this limitation.