Captioning supports inclusive teaching and learning for all types of students, and providing captions and transcripts can help students in the following ways:
-
Enabling adjustment to new ways of working, for those primarily used to in-person lectures now being delivered online
-
Providing a backup for technological difficulty (poor connection when streaming, skipping, lag time), or where students are working in disruptive/noisy environments where captions can aid their understanding
-
Making searching within a recording easier, simplifying revision of key concepts
-
For those who are deaf or have hearing impairments, it provides an essential and legally required access to learning
-
It helps to support those with processing or memory difficulties, by reinforcing verbal content
-
Supporting students whose first language is not English
Do I need to provide captions?
The University Council has released a statement confirming that the University aspires to voluntarily comply with the requirements of the Accessibility Regulations 2018 as a model of best practice.
The University is also bound by a statutory anticipatory duty (under the Equality Act 2010) to ensure that all teaching and learning materials are accessible. We should therefore be providing captioning wherever possible, and in some cases we are required to provide them.
Currently, captioning must be activated by the lecturer manually for each recording. The General Board's Education Committee (GBEC) has considered the benefits and drawbacks of automatic captioning, particularly with regard to accuracy, and has agreed that it is not currently appropriate to enable automatic captions by default. The University will continue to monitor developments in automatic captioning and will review this decision in due course.
What are the expectations for captioning?
Accuracy in automatically-generated captions can vary and depends on a number of factors, such as: the ability of the artificial intelligence to recognise technical or specific terminology, or particular accents and speaking patterns; speaker placement in the room or the quality of audio equipment used; other noise in the environment; or adaptation of sound quality by the recording tool at time of the recording, for example compensating for poor connectivity. For these reasons, it can be common for captions or transcripts to contain misidentified text; this may impact on understanding or create erroneous text, and may also result in transcripts or captions that may imply, or contain, inappropriate or offensive language. Panopto gives reasonable accuracy with the spoken voice (70-75%), though very technical or foreign language content may be misidentified.
Wherever captions are used, GBEC has agreed the following expectations of staff and students:
- Use of captions or transcripts is deemed complementary to the recorded session, and these should not be viewed/used in isolation of the accompanying recording.
- Staff are neither required nor expected to regularly check and edit captions or transcripts, and will not be held responsible for inaccuracies or errors in text. Where particularly misrepresentative or inappropriate text is brought to their attention they are encouraged to take action, but this is not required.
- Students should be aware that captions or transcripts generated cannot be relied upon for accuracy, and that staff will not be checking or editing as standard.
How do I caption my videos?
While exact steps vary depending on the platform you are using, most platforms have an embedded captioning function. In addition, the University provides Panopto as a recording and delivery platform, and Panopto can caption any video uploaded to it, whether recorded in Panopto or not. Recordings made in other platforms (e.g., Zoon, Google Meet, Teams, etc.) should always be uploaded to Panopto and not to Moodle directly; this will enable you to take advantage of captioning and other features.
This document provides details on captioning in primary platforms, as well as estimates of accuracy for the most common platforms. More information on how to produce captions in Panopto, and upload files from other platforms, is also available on the support hub.
If you have been instructed to provide captions for a student who requires these in order to effectively access their learning as a result of a disability, and find accuracy in your chosen platform to be an issue, please contact the Disability Resource Centre. The University has limited access to more advanced captioning agents where required.
Where can I get help?
For support on captioning, see the document in the section above, or visit the Disability Resource Centre’s accessibility pages.
For more information on support for disabled students in teaching and learning please visit the DRC website.
For support on captioning within Panopto, visit the support hub in the first instance, or contact the Lecture Capture team.
For more information on inclusive teaching and learning practice, visit the Cambridge Centre for Teaching and Learning.