The education secretary has been attacked for failing to halt the gathering of attainment data based on Scotland’s national literacy and numeracy tests, after the international experts who reviewed the curriculum questioned its usefulness - and after it was used recently to create “crude league tables” that “demoralise schools and pupils”.

Defending the Scottish National Standardised Assessments (SNSAs) - and the national data on literacy and numeracy collected on the back of them - Shirley-Anne Somerville said that teachers were using the tests more, and “finding them more useful as the years have gone on”.

Ms Somerville made her comments in an appearance before the Scottish Parliament’s Education, Children and Young People Committee this morning, under questioning from Liberal Democrat education spokesperson Willie Rennie.

Ms Somerville also:

• Told the committee that she “did not intend to take over day-to-day control of the SQA [Scottish Qualifications Authority]” after the Scottish Conservative education spokesperson, Oliver Mundell, suggested it was time for her to “step in”, given the body’s “catalogue of errors” and “poor judgement”.
• Admitted there had been “sub-standard practice in years gone by” at the SQA when it came to its equality duties.
• Said that councils had made “great strides” improving ventilation in schools, amid ongoing concerns over the impact of Covid on staff and pupils, and said that there would be a report on progress published later this month.

Each year the Scottish government asks teachers to report if pupils in P1, P4, P7 and S3 are attaining the expected level in literacy and numeracy for their age and stage, with the school staff expected to use the SNSAs to inform their judgements.

Concerns over literacy and numeracy data from Scotland’s schools

The judgements are typically published in December - although the Covid pandemic has led to some disruption - and recently a national newspaper used the 2019 data to create a league table of P7 pupils’ performance.

However, in its June report, the OECD questioned if the data should be published at all.

It said: “While this data is interesting, reporting it on a national scale and tracking small changes in percentages as evidence of improvement or otherwise may not be giving the system the robust data needed to monitor student achievement.”

The OECD recommended that Scotland “redevelop a sample-based evaluation system to collect robust and reliable data necessary to support curriculum reviews and decision making”.

The Scottish government has said it accepts all 12 OECD recommendations, but today - under questioning from Mr Rennie - Ms Somerville maintained the “pupil-by-pupil” data collected on the back of the tests is valuable so that “variation in outcomes” can be identified at a national, council and school level.

She also said: “We are seeing an increased use of [the tests] by teachers and they are finding them more useful as the years have gone on.”

Ms Somerville used another OECD report - this time from 2015 - to justify retaining the tests and the gathering of the Achievement of Curriculum for Excellence Levels (ACEL) data.

The 2015 report, she said, had found that “light sampling” of literacy and numeracy at the national level - as carried out via the Scottish Survey of Literacy and Numeracy (SSLN) - had not provided sufficient evidence.

She said: “We have been told by the OECD in the past that sampling does not provide the right evidence to allow national agencies to be able to take the right decisions and the right course of action.”

Mr Rennie admitted that the SSLN, which was published for the last time in 2017, had not been perfect but he maintained that a sampling model could work, if the sample was larger.

“The crucial word in that was ‘light sampling’. You could have heavier sampling and still keep a sampling model rather than the current model you have adopted that allows for crude league tables to be published,” he said.

Mr Rennie argued that allowing teachers to continue using the tests - but stopping the collection of ACEL data - would be “the best of both worlds”

He said: “Why won’t you consider that, so we can stop having these crude league tables that I think demoralise schools and pupils?”