May 2, 2019

Update: Eye tracking Study supported by Facebook

Eye tracking of woman watching video with SLS
Eye tracking of woman watching video with SLS

Dear Friends,

Here is an update on the eye tracking project that we are working on. 

Project Summary: Same Language Subtitling (SLS) is the idea of subtitling audio-visual content in the “same” language and script as the audio. See https://www.youtube.com/watch?v=tOsWToI2PIw.   

Several studies have found that regular viewing of existing film songs on TV, with SLS, results in reading skill improvement. In 2-3 years of frequent exposure, an early-literate who cannot yet read, transitions subconsciously to functional reading ability because the brain cannot but try and match sound and text that is in perfect synchronization. Bill Clinton called SLS “a small change that has a staggering impact on people’s lives.” See https://www.youtube.com/watch?v=juZOlmf9APk. 

The research questions we would like to explore with early-readers are: How much reading along occurs with SLS on film songs? Does that depend on reading ability and/or prior knowledge of the song/lyrics? 

We are hopeful of answering the above questions by doing a thorough eye tracking study. 

Project Goals: In 2 different low-literacy regions/languages (e.g., Hindi and Telugu), we will identify 100 early-literate and out-of-school people per state (half male and half female). Early-literates are defined as people who self-report as “literate” but who cannot read a simple Grade 2 level text, with a minimum level of fluency. 

Show each of the 200 early-literates, two different content genres: 1) 4 film songs in their language of literacy, with and without SLS, and 2) a 5-minute film clip of dialog, with and without SLS. Two of the songs will be well-known to the subject and two, not. All the sessions will be eye-tracked using a table-top eye-tracker like Gazepoint. 

To determine the quantum of reading due to SLS, our analysis will compare for both content genres, the number of eye-fixations and time spent in the subtitle band, with and without SLS. 

For 20 subjects representing the diversity of our sample, we will create a video capturing their eye-tracking, like we did in the Rajasthan study: https://www.youtube.com/watch?v=1QuJKet4SPw&feature=youtu.be. 

The first phase of this project was recently conducted in Rajasthan where we conducted the eye tracking study on 150+ weak readers. We are now in the process of analyzing the data. A detailed report will soon be available.

Here are some small video clips of the eye tracking study: 

Reading assessment test: https://youtu.be/Cx17F0wd9dI 
Girl watching video with SLS: https://youtu.be/dqYbsYHO1T8
Man watching video with SLS https://youtu.be/vAJ2xiQUKPc
Feb 5, 2019

Facebook Research Award for Eye-Tracking Study

Viewing content with same-language subtitles
Viewing content with same-language subtitles

Facebook Research recently announced the winners of their Global Literacy & Accessibility Challenge, and one of the four winners is PlanetRead founder Brij Kothari's proposal, "Eye-tracking of Same Language Subtitling on Film Songs and Dialog." We are very grateful for this award, which will support PlanetRead's development of eye-tracking research on same-language subtitles for televised films and songs. 

As Facebook Research notes, the awards, given to PlanetRead's proposal and three other studies, will support "new and innovative research to understand and address global literacy and accessibility issues."          

Half a billion people in India are “early literates” whose quality of basic education and digital access is severely compromised. PlanetRead has been working to solve this problem with a well-researched solution that can potentially transition a majority of early literates to functional reading: same-language subtitling (SLS), which involves subtitling audio-visual content word for word, in the same language and script as the audio.

Studies have found that regular viewing of existing film songs on television with SLS results in reading skill improvement. Within two to three years of frequent exposure, without conscious effort, an early literate who can only make out letters can transition to a functional reading ability; the brain cannot help trying to match sounds with text that is perfectly synchronized.

Most of the existing research on subtitling relates to second-language learning and or media access, but only a handful of studies have explored subtitling’s links with basic reading and literacy. Thanks to the Facebook Research Award, PlanetRead will be able to add to studies measuring the effectiveness of SLS.

PlanetRead will explore the following research questions: How much reading along occurs with SLS on film songs and film dialog? Does reading along depend on the person's existing reading ability and/or prior knowledge of the song lyrics or dialogue? We propose to use eye-tracking software to measure the viewing of film songs and dialogue by 200 early literates. To determine the quantum of reading caused by SLS, the same content will be shown with and without SLS.

According to Facebook Research, findings from all four award recipients' studies will be presented in California later this year and "will be shared publicly with the research community." We look forward to sharing our findings with our supporters as well. 

Links:

Nov 5, 2018

PlanetRead Featured in Harvard Economics Review

Studies showed that children read along with SLS
Studies showed that children read along with SLS

PlanetRead founder Brij Kothari was recently featured in the Harvard Economics Review blog with a post about two studies measuring the efficacy of same-language subtitles (SLS) to improve reading ability in Indian schools.

"The Reading Revolution Will Be Televised" describes one study measuring whether children pay attention to SLS when subtitles appear on animated stories ("AniBooks" created by PlanetRead's sister company BookBox) and a second study investigating whether the addition of SLS leads to improved reading ability.

As the post describes, thanks to the 2009 Right to Education Act, India has been extremely successful in getting school-age children into classrooms, but far less successful in teaching them to read in those classrooms. Kothari estimates that half of the children in grades 1-5 are falling behind in reading and are likely to remain functionally illiterate—unable to read at a second-grade level or better—unless action is taken to help them. 

PlanetRead's first study, supported by USAID and Tata Trust, followed children at a government school in Rajasthan. Using eye-tracking, PlanetRead researchers found that even those children with the weakest reading skills were automatically drawn to reading along with SLS. The study demonstrates that adding SLS to children's cartoons on television could offer inescapable reading practice for the vast majority of India's children who watch children's television programs at home.

 The second study investigated whether reading along with SLS would result in improved reading skills. PlanetRead researchers worked with children at low-income schools in Delhi, with half the schools getting regular classroom exposure to AniBooks with SLS and half serving as a control group. As Kothari explains, "The AniBook intervention alone contributed an additional 70% of an entire school year’s achievement in reading score."

Read "The Reading Revolution Will Be Televised" for all the details of PlanetRead's ongoing research, and please consider supporting our work to advance the cause of same-language subtitling for literacy!

Links:

 
WARNING: Javascript is currently disabled or is not available in your browser. GlobalGiving makes extensive use of Javascript and will not function properly with Javascript disabled. Please enable Javascript and refresh this page.