By Radhika Iyengar at the Center for Sustainable Development
The image on the right is a Standard (Grade) 1 Telugu literacy textbook used in the state of Telangana, India. By the end of Grade 1, children are expected to fluently read and comprehend this text. However, a 2017 literacy assessment conducted by Columbia’s Center for Sustainable Development in the Medak District of Telangana found that this expectation was far from reality. The table below shows the results of the assessment for 10 schools.
We see that what the children are expected to read is very different from what they can actually read. One hundred percent of the children should be able to read 100 percent of the words. Yet the assessment showed that at the end of Grade 1, the children could read just 25 percent of the words correctly. They could read 36 percent of the letters correctly.
The theory that a textbook like the one pictured above could support student learning in Grade 1 is based on the effectiveness of such methods in wealthy nations, where students entering Grade 1 come prepared with experiences of having been read to since birth by literate family members, and exposed to text through different kinds of media. The fundamental difference, however, is that many students in Telangana do not enter school with the same level of preparation. Many students come from low-income homes with parents who lack literacy skills, and are therefore being asked to decipher text for the first time. Their learning needs are therefore very different.
Cognitive neuroscience research tells us that reading does not come naturally to the brain. For early readers, the skill of identifying individual letters is a visual exercise, separate from language. Once this acquired skill of identifying letters becomes automatized through long hours of practice, the brain is then able to link the part of the brain that processed vision to the part of the brain that processes language. This learning process connects different parts of the brain that were not connected earlier. How can we form these networks so that our brains get activated to become fluent readers? What does cognitive neuroscience tell us about how to make these connections faster to then translate into reading? Much information passes through our brains, however we remember very little.
Based on cognitive neuroscience research on how children read, a reading program is being designed and implemented in Mahabubnagar District of Telangana State, India, in conjunction with the Center for Sustainable Development and the District Government of Telangana. The first year results have shown remarkable progress.
Accuracy in the syllabic scripts is a significant problem. Particularly in the southern scripts, children keep making mistakes throughout primary school. The data show that the treatment group overall read 70 percent of words correctly compared to only 20 percent in the control group. Percentages were even higher for letters. As expected, children in higher grades were reading more accurately than the children in lower grades.
The main principles of this method are very simple yet profound, and are grounded in cognitive science. The human brain is like a funnel. Around 7 items can trickle into the brain in about 12 seconds. Information in short-term memory can be held for only a very brief amount of time. Secondly, the amount of information that can be retained in short term memory is not much, as it has very limited capacity.
To use this concept in teaching literacy, children should be given smaller chunks of information which can process through their working memory into their long-term memory. They should be taught letter-by-letter first through simple procedures so that their working memory can efficiently retain small bits at a time. With consistent practice, the children start recognizing letters and their associated sounds until the reaction time to decoding letters and words reduces to milliseconds.
Along with linguists from Hyderabad University, textbook developers in Medak district and educationalists from the Center for Sustainable Development team operationalized these principles into developing a workbook on Telugu literacy. The book included letters in simple, consistent fonts and in big and small sizes. To not overload the brain, each day was dedicated to learning a single letter and its sound. Each letter was followed by blended letters to form words followed by sentences. Unlike the Grade 1 textbook, which was overcrowded with large, colorful, distracting pictures and small texts with little practice, this workbook included letters in big fonts, hardly any pictures and provided much text to read for ample practice. Through this approach, the entire alphabet should be learned within the first 100 days of Grade 1.
Last week the district collector and magistrate of Mahbubnagar, Telangana State announced a scale-up of this approach to 70 schools with more than 2000 children from Grades 1 to 3 participating in this 100 days literacy program.
Acknowledgement: The author would like to thank Dr Helen Abadzi who made cognitive neuroscience principles adaptable to local language literacy programs across the world. Thanks to Srinivas Akula who leads the Telangana Literacy efforts.