Blavatnik School of Government, University of Oxford
Among the exciting and interesting new research and findings presented at the 2019 RISE Annual Conference, a reminder kept popping up: how devastatingly low current learning levels are in so many developing countries. There is an increasingly powerful case to be made that the first step in making an education system coherent for learning is to focus on universal, early, conceptual, and procedural mastery of basic literacy and numeracy.
Karthik Muralidharan, for example, presented new results from a large-scale evaluation of MindSpark, a software that adapts lessons to children’s skill levels. But it was the baseline data, both from an earlier pilot in Delhi and from the scale-up to 6,500 children in Rajasthan, that he said he considers “among the most important to understand developing country education.” Both studies showed that the average child was already 2.5 grade levels behind the curricular standards for math by Grade 6 and that by Grade 8 the gap had extended to 4 grade levels behind. The typical child in Grade 8 had really only mastered the fourth-grade curriculum.
And these low levels of learning can also lead to little or no progress in learning. The learning progress of the control groups in the two studies (as an estimate of the “business as usual” learning trajectories) showed that children who began in the bottom third of the distributions in the control groups made no learning progress at all during the period of the study, which for the scaled-up program was two years. According to Muralidharan, these children were so far behind the curriculum that they were not able to learn anything. As he put it,
“If you’ve fallen behind then you’re basically left behind forever, and hence the centrality of foundational literacy and numeracy as the single most important thing we try to do in trying to build education systems.”
Similarly, an experiment run in The Gambia showed that, despite the time they spent in school, a large portion of control-group children could complete almost no tasks in the learning assessments (such as recognizing numbers, simple addition, two-digit addition, etc., and a similar set of simple tasks for reading like recognizing letters) at endline. In the session’s Q&A, Ben Piper, an education expert at RTI, remarked how “heartbreakingly bad” the base case outcomes were, almost as if there were no school at all.
Multiple presentations also suggested that low learning contributes to dropout, showing that low learning is not just a “learning” problem, but also a “schooling attainment” problem. A presentation by Rafael de Hoyos on a scholarship program in Mexico found that scholarships had zero impact on secondary school graduation rates or on science or math test scores. The most plausible explanation the authors found was that many eligible students did not seem to have the minimum learning level needed to successfully face the requirements of upper secondary school. Their recommendation was that such scholarship programs should be accompanied by remedial education (or, presumably, by improved instruction in earlier grades so remediation isn’t necessary).
In a presentation using Young Lives data from five countries (Ethiopia, India, Pakistan, Peru, and Vietnam), Abhijeet Singh showed that test scores at age 12 (roughly Grade 6) are “massively predictive” of years of schooling completed at age 22, indicating that low learners are more likely to drop out. He suggested that building foundational skills in primary school may lead to higher schooling attainment. And Muralidharan, in his presentation, stated that the low learning observed at baseline and in the control groups in his studies can help explain dropout: “It’s because going to school day after day and not understanding anything is a miserable experience."
The link between low learning and dropout is important for efforts aimed at getting more kids in school and for longer. If children are dropping out because they aren’t learning, then pushing for them to stay in school longer (such as through secondary completion) may do little to improve learning outcomes. Efforts that improve learning, on the other hand, could be “win-win” and improve both learning and attainment, as children who otherwise would have dropped out are enticed to stay in.
At RISE, we’ve been thinking for a while about findings like these and what they imply for making education systems coherent for learning.
Taken together, such findings strongly suggest that education systems need to prioritize something we’ve dubbed universal, early, conceptual and procedural mastery of basic skills. (If that phrase isn’t a mouthful, I don’t know what is, but each word is in there for a reason. We’ll work on a decent acronym).
Taking the phrase apart, here’s what we mean:
While these may seem obvious, they often aren’t top priorities.
Universalizing secondary school completion, teaching 21st Century skills, providing vocational training, and reducing the “digital divide” appear frequently in lists of education priorities. Yet none of these will produce much fruit if they aren’t preceded by UECPMBS (and yes, we need a better acronym).
RISE blog posts and podcasts reflect the views of the authors and do not necessarily represent the views of the organisation or our funders.