Accuracy Matters
While the growth of video-sharing websites (such as YouTube and Vimeo) has resulted in an abundance of media at our fingertips, the videos found in these repositories often do not contain captioning, or fully accurate captioning. Is 95% accurate enough? Let's consider the following example represented in the image below.
The image above shows a snapshot of a video wherein Governor Mitt Romney is giving a speech. YouTube's "auto-captioning" attempted to caption the video. This auto-captioning service mistakenly identified the phrase "of strength" as "airstrikes." While an occasional typographical error would not cause great consternation, one inaccurate word can dramatically change the nature of the message.
In the previous example,Mr. Romney began the speech by calling for a strong military. It's not necessarily unreasonable that he may be calling for increased airstrikes at the end of the video. A hearing-impaired individual relying on these inaccurate captions would not be able to determine his intent.
This captioned video, while almost entirely accurate, would not be considered to have met the American with Disabilities Act (ADA) compliance standards. In addition to causing confusion, auto-generated captions like the above example often lack punctuation, and contain multiple spelling errors, making it challenging for learners to follow.
How Accurate Then?
While the US government has been reluctant to provide a specific measure of what constitutes an appropriate level of accuracy, we should strive for 100% accuracy in captions. The US government has indicated that vendors which provide captioning for live legislative broadcasts should be capable of providing 98.6% accuracy. Therefore, we will use 98.6% as our metric for ensuring captions satisfy expectations related to the Americans with Disabilities Act.
Another important element to keep in mind regarding captions is punctuation. To satisfy ADA expectations, punctuation is necessary. Many of the auto-generated captions (such as YouTube's technology) create captions which may appear accurate at a glance. However, such captions do not include punctuation, and thus are not ADA-compliant (regardless of how accurate they may seem.) It's not difficult to understand how the absence of punctuation can inhibit learning a nutritious breakfast will always help start the day Jason Frank has too many legos ... I think you get the point. :)
Sounds in Videos
Any sounds that are essential to the learning should appear noted in the captions as well. For example, in the following video clip, it is essential for learners to know that a dolphin creates a clicking noise when using sonar. The clicking noise is indicated in parenthesis within the captions. You'll only need to watch about the first 20 seconds of the video below to see the notation.
Captions vs. Transcripts
Captions are necessary if the timing of the audio component needs to be coupled with the video component in order for viewers to understand the material. For example, in the video below, it is necessary for learners to view the captions along with the illustrations to understand the material.
In the video below (used to promote another workshop), captions are not necessary as the audio does not need to be timed with the video in order for the information to be understood. A transcript could suffice (as shown in the example linked beneath the video.) Transcripts can be provided in a Word document, .pdf file or typed directly into a web page.