Assessment for Learning (A4L) is an idea that must be applauded, mandated, and funded, ideally in that conceptual order. Mandated, especially, by development countries, development agencies, and civil society.
Perhaps it is now commonplace to say “we’ve gotten going on access to education, now it is time to add learning”. And it is important reminding ourselves of the absolute numbers of children who are in school, but learning little -- around 250 million -- as calculated by UNESCO’s Global Monitoring Report (GMR).
Children in low-income countries learn very little
But we also need to remember that the average child in low-income countries seems to be learning (as proxied by international assessments) at the level of the 5th or lower percentiles of children in high-income countries. This means that the average child in low-income countries scores less than about 95% of the children in high-income countries.
To put it another way, children in grade 6 or so in low-income countries seem to have about as much reading and arithmetic skill as children in grade 2 or 3 in high-income countries. This is tragic, not only for the wasted treasure and effort it represents, but because it undervalues the lives and efforts of children and their families.
It threatens to undermine the respect that families have for education and what it could do for them. The new Sustainable Development Goal (SDG) for education recognizes this, calling for meaningful learning, whereas the Millennium Development Goal for education had only an indirect reference to quality and learning.
We can gain much by doing better learning assessments
But rather than put things negatively, in terms of what we lose if we don’t assess for learning, maybe it is best to put things positively. First, we should note that both the “A” and the “L” in the A4L initiative are key. A lot of the emphasis naturally gets put on the “A,” and, at the outset, perhaps rightly so.
Improving assessment (e.g. improving in the domains where we mostly know how to assess, but also creating—almost from scratch—measurement in new domains) requires a lot of technical development, agreements and international consensus. It also grabs attention as part of the SDG process.
And the “A” is chronologically first: you can’t do an assessment for learning unless you assess. But, as Joshua Muskin has pointed out in a forthcoming Brookings Institution working paper on using learning assessment data to improve teaching and learning, the “L” is critical. It is what will actually create hope and progress.
Assessing tells us a lot
To drive this point home, I thought I’d share two graphics from the NGO Room to Read (full disclosure: I sit on their Board). These graphs show progress in average words read per minute in grades 1 and 2, using reading fluency as a simple indicator. Students are from schools where Room to Read has trained teachers and librarians in the best practices for literacy instruction (blue lines) and in comparison schools (gray lines) without these interventions.
Reading fluency is assessed at beginning and end of grade 1, and at end of grade 2. A few points strike me as very worthy of note, and as a sort of case-in-point for the whole A4L approach.
First, thanks to assessment, we can document what happens (and it can happen fairly quickly) with improved literacy instruction based on teacher support. This creates hope that good instruction can drive change and feed virtuous cycles.
Second, the instructional model used is dependent on data for driving learning improvement, not just for reporting and accountability (though that too)—and it works.
But thirdly, and perhaps most importantly, we can see, in microcosm and almost in real time, thanks to the assessment data, the gap that can widen when instruction is not well-supported.
With poor instruction, a small gap opens up between the beginning and the end of grade 1, and that gap continues to widen in grade 2.
This is a microcosm of what is happening with equity of learning in many low-income countries: schools for the well-off do better right away (and the well-off start off with some advantages, because they often have access to early childhood development). Then gaps begin to emerge and then grow bigger and bigger.
Education must be equitable – and assessments can help
Education, therefore, does not make the contribution to equity that it could make. It is also a microcosm of the gap that opens up between low-income and high-income countries, where, in the latter, a solid foundation is laid.
These are data from projects. We need to get similar data (and for various subjects and levels, of course) at a national, systemic level. These data would allow authorities to understand which gaps in the learning process are common across schools. That knowledge would then help to improve teacher support.
The ideas in the A4L concept note can help get us there: the theory of change is sound, the principles are effective, and the activities proposed are consistent with the theory of change and the principles. It is a coherent package that can work.
There are other initiatives being developed by the UNESCO Institute for Statistics to support the monitoring of the SDG for education. These (and other) initiatives have to be synergistic if we are to take advantage of the current historical juncture, and do it with efficiency.
Working out roles and synergies will be critical, and it can be done. At that point we can all act together, with strong mandates and reasonable levels of funding. Read more: