One report card is not enough

One report card is not enough

Efforts to measure learning outcomes are hobbled by disagreement on the instruments. But there is no one best way

The 12th Five Year Plan acknowledges this learning deficit and targets improvement of education quality as its priority. (AP)
The 12th Five Year Plan acknowledges this learning deficit and targets improvement of education quality as its priority. (AP)

If education were only about schools, physical infrastructure and ensuring universal enrolment, then India has succeeded spectacularly. Every large habitation has a school, most schools have buildings and teachers assigned, students have study materials. The first step, schooling, has been taken. It was hoped or assumed that once the schooling inputs were in place, education and learning outcomes would follow automatically. However, that learning outcomes are woefully deficient has been established by numerous Indian and international assessments. The next step is to turn schooling into real education.

The 12th Five Year Plan acknowledges this learning deficit and targets improvement of education quality as its priority. It seeks to introduce “a system-wide focus on learning outcomes that are assessed through classroom-based Continuous Comprehensive Evaluation (CCE) independently measured, monitored and reported at the block/ district/ state levels”. It also advocates the articulation of “clear learning goals at the end of each class” that are “understood by parents and teachers”, as well as its “bottom-up community-driven monitoring”. We propose a three-fold scheme of measurement of learning outcomes to turn that agenda into action.

A major impediment to moving learning outcomes to primacy over inputs in the politics, policy, and practice of education is disagreement on the assessment and measurement of student learning. What doesn’t get measured doesn’t get done. But part of the debate about assessment is a fear that measuring learning will create too much “high stakes” pressure. This fear is odd because there already is lots of measurement, creating information that is collected, analysed and reported. The current state and district elementary education “report cards”, to tell us “where we stand”, have 824 separate pieces of information for each district. We can know the number of classrooms with a ramp, the number of classrooms needing minor repair, the number of students that are blind. But the current report cards on schooling are completely silent on education.

Suppose we do decide to measure progress on learning, a fundamental principle of economics is that achieving optimal results requires as many instruments as targets. If you want to kill two birds you almost always need two stones, as two birds with one stone is dumb luck, not a plan. There has been too much fruitless debate about the one best way to measure and report on learning — fruitless because there is no one best way. There are different purposes of assessing learning and hence different approaches are necessary, even if all agree on the conceptual base.


There are three audiences for information on learning: educators, who need information to guide their day-to-day and year-to-year instructional activities; government officials, who supervise the system and who need information to understand how the overall system is progressing; and parents and citizens, who need to know how their child, their school, their locality, their state and their country are doing.

Measuring and rendering learning outcomes in a manner that resonates cognitively with the three separate audiences of educators, supervisors/ officials and parents/ citizens is a formidable challenge.

A few principles should underpin any such effort.

One, the National Curriculum Framework (NCF) and CCE are already adopted and are in place and provide an excellent platform to initiate this process. However, the inherent complexity — several competencies for each subject in each grade and the multiple learning milestones for each child in each academic year — while perhaps necessary to guide educators, headmasters and teachers on day-to-day instructional activity, cannot be the sole way in which supervisors and parents are informed. So, we should assign instruments to targets, and different audiences should be informed in different ways.

Two, information on learning outcomes should serve as decision-support and not merely become an end in itself. Accordingly, supervisors, already over-burdened with large numbers of schools in their jurisdiction, should be able to utilise this information to allocate their scarce inspection and review times to assist schools and teachers who need the greatest support with overcoming learning gaps.

Three, information should be rendered in the simplest and most cognitively striking manner. Measures that seek to capture everything run the risk of communicating nothing. Four, any assessment framework, for teachers, supervisors and parents, should be designed with the typical — not extraordinary — stakeholder in mind.

We propose not a perfect approach but a workable one, with three different ways of reporting information. This is a three-pronged strategy, where each prong is not a competitor but rather a complement to the other. Empowering teachers, supervisors and parents with information on learning levels is central to this strategy. The classroom teacher needs to track the learning trajectory of each student, no less, so as to enable them to achieve their grade-specific competencies. However, to achieve this, while in the long run the full NCF might be used, given the existing lags in learning, teachers are likely to fall short. Therefore, to start the use of CCE for a practical and field-tested set of basic competencies, situated within the NCF, can be the first step towards bringing ambitious yet achievable learning outcome objectives into schools.

Though school supervisors require much less information, with decreasing granularity up the hierarchy, they need it to improve the quality of monitoring. Rather than continuous measurement of all learning objectives, supervisors can use periodic rounds of dipstick evaluations. Done with rigour in assessing chosen basic competencies as proxies for the larger agenda, these can become benchmarks for supervisors in monitoring learning levels within and across schools.

Finally, parents need this information to be aware of their child’s learning level. Simple and cognitively striking report cards, which render the child’s learning trajectory in each subject against competency benchmarks as well as the classroom median, can keep parents informed and create demand-side pressures for accountability.

An information cascade that appropriately integrates the varying requirements of all three stakeholders can be a powerful force to improve student learning outcomes. The illusion of one best way of assessing all student learning should not block movement towards multiple ways of good measures, adapted to their purpose.

Pritchett is professor of practice of economic development at the Harvard Kennedy School, US. Natarajan is an IAS officer, batch of 1999, and a graduate student at HKS