30 lakh primary-schoolers to be tested in India’s largest ever learning assessment exercise

The National Council for Educational Research and Training is getting ready to conduct its largest ever learning assessment test for primary school children in September-October. With a sample size of over 30 lakh students drawn from Classes 3, 5 and 8 from schools across the country, it will dwarf all other large-scale assessments in India, public or private, and even some international ones.

The council, which advises the Central and state governments on school education matters, has been conducting a National Achievement Survey once every three years since 2001. But the upcoming exercise – born of the general alarm about the declining quality of education in the country – is vastly different. “We are still calling it the National Achievement Survey,” said Y Sreekanth, who heads the council’s educational survey division. But unlike the old version that tested children of one class in each round, the new one will test three classes at once. Its organisation will be decentralised – this was imperative, given the scale of the operation – and it will also be linked to Aadhaar, the 12-digit biometrics-based unique identity number. Most importantly, the results will be out within a year unlike the three years it previously took. Authorities hope this, along with the involvement of teacher-training centres in the exercise, will make early intervention possible.

Anil Swarup, school education secretary in the Human Resource Development Ministry, said the “on-field assessment” will be conducted in September and October and “the feedback assessment” will be completed by December-end. “Changes will be introduced in the last quarter of the academic year based on the assessment and feedback,” he was quoted as saying by PTI.

Added Sreekanth, “The effort is to complete the process in one year and provide feedback within the same academic session.”

In numbers

The original National Achievement Survey tested 1.5 lakh to 2.5 lakh children in each round. “We derived samples with the state as the unit and held tests in all 36 states and Union territories,” explained Sreekanth. “But from now, the district will be the unit and the data much more granular. The objective is to find out how [students] are performing and provide intervention.”

The 30-lakh sample size will include 4,500 primary-schoolers from each of India’s 700-odd districts.

To put that into perspective, the Annual Status of Education Report, of the non-governmental organisation Pratham, tested over 5.6 lakh children between the ages of three and 16 for its last survey. And in 2015, the international Programme for International Students Assessments tested 5 lakh 15-year-olds across 72 countries.

Learning outcomes

Like in previous National Achievement Surveys, students of Classes 3 and 5 will be tested on mathematics, environmental science and language, and those of Class 8 on science, social science, mathematics and language. But the National Council for Educational Research and Training has tweaked the design a bit. “We will see in each class [how] students have performed against the new learning outcomes,” said Sreekanth, referring to a document released by the council in January that spells out what skills students of a particular class are expected to master.

“There are items of different kinds in the test – some of a basic level, some intermediate and some advanced,” he added. “We give all three to see how children are performing and this can be very useful for teacher-training both pre-service and in-service [training for a teaching qualification and training while employed]. Also, we have seen that children do better in subjects like mathematics if they can relate to the items and have visual cues.”

The test data will then be analysed against three parameters – location (rural or urban), gender, and social categories – and reports on the same will be released to the public.

In addition, there will be questionnaires to gather information from and about teachers as well as about school facilities and households.

Quick results

The results of previous surveys took three years to come. But with radical logistical changes and the involvement of various institutions, Sreekanth expects the new exercise to speed up the results process, even with the massive increase in the sample size.

In a first, the National Council for Educational Research and Training has roped in Block Resource Centres, set up by the government to conduct in-service training of teachers, and District Institutes of Education and Training, which train teachers pre-service, for this exercise.

The tests will continue to be multiple-choice and written on regular answer scripts but those will no longer be shipped back to the council for processing. “Teachers will mark out the answers on OMR [optical mark reading] sheets [used for computerised paper-checking],” said Sreekanth. Next, the sheets will be scanned at the district level and loaded on to a purpose-built software so that they can be shared with the council. “But an autogenerated report will be available for the schools quickly, almost as soon as the data is uploaded,” he said. And help will be at hand from the teacher-training centres for schools that need course correction mid-session.

This process was pilot-tested in March with 15 schools in each of 17 states, including Karnataka, Manipur and Himachal Pradesh.

The test is also being linked – in another first – with the District Information System for Education, a database of information about schools that is maintained by the National Institute of Educational Planning and Administration, and through that to Aadhaar. “This will allow us to monitor students’ performances over time and we do not have to collect background information every time we test them,” said Sreekanth. “The district information system will handle data collection on social categories, facilities available in schools and teachers.”

Fewer tests?

When the government released the draft for the new learning outcomes in January and announced its intention to include these in the Central Rules of the Right to Education Act, educationists feared this would lead to a “testing regime”.

But Sreekanth pointed out that the expansion of the National Achievement Survey would eventually render the State Level Assessment Surveys obsolete. In his opinion, they are of little help anyway. “The data generated was not properly utilised, we were only getting numbers – how many are meeting the benchmark,” he said. “What we need instead is descriptive, qualitative feedback. We need to know where exactly a student is poor and supplement the efforts of the schools.”

He added that the National Council for Educational Research and Training, in general, discourages repeated testing. “Some schools were conducting the SLAS [State Level Assessment Surveys] multiple times and others were conducting the Summative Assessment 2 [second term exams] through standardised tests,” he said.

In addition to the new National Achievement Survey, the council has designed assessment tests for all primary classes (1 to 8) that schools can administer on their own.