Canada’s public broadcaster and McMaster University’s CanChild research center recently teamed up to study optimizing captions for kids with disabilities. And they have found that while comprehension doesn’t change much depending on the style, children do have their favorites.
The main takeaways from the Optimization of Closed Captioning for Children with Disabilities report for broadcasters and producers is that kids prefer karaoke-style captions (where the words are highlighted as characters speak). Traditional bottom-oriented captions and dynamic captions that display in different spots on the screen were also tested. But the style doesn’t seem to have a significant impact on comprehension or reading time.
CanChild and CBC worked with 67 kids ages six to 10 with different reading levels and abilities. The main goal of the study, which ran from October 2023 to July 2024, was to determine which caption style helped comprehension the most, and which one kids liked best.
It was concluded that CBC Kids could certainly continue using traditional captions, but some experimentation with other styles could create a more entertaining and engaging experience. As a result, CBC Kids has pledged to test out karaoke-style captions on its website, where it will gather viewership metrics and feedback to further gauge effectiveness and popularity.
This study is just the latest industry effort to rethink captioning for kids. Ireland’s JAM Media is developing a software tool called Zaki Signs that combines mocap, AI and the Unreal Engine to create real-time, ASL interpretation featuring an animated character who signs along with the show. The goal of the tech is to better support younger kids who can’t read, and create a more engaging experience than just putting words on the bottom or having a human interpreter signing in the corner.
In 2022, a Kids Industries survey about captions found that most US kids watch content with the subtitles or captions on—not necessarily because they need them, but because it adds to their enjoyment of the content.