New checks render results ‘meaningless’, says marker

Test papers are being marked inaccurately because of lax assessment of markers under the new system, according to senior examiners
4th July 2008, 1:00am
Test papers are being marked inaccurately because of lax assessment of markers under the new system, according to senior examiners.

A key stage 3 maths team leader said one member of his team of nine markers had been allowed to continue examining despite making 10 marking errors when she completed “benchmarking”.

Benchmarking is an online marking exercise designed to check examiners are working to a good standard. The benchmarking error limit was supposed to have been set at eight and those who fail the test twice are meant to be barred from marking.

But the team leader said he thought all markers were being allowed to continue, whether or not they were over the error limit.

He said of his team member: “She did not get any papers at all until less than two weeks before the deadline and then marked 350 papers in a very short period.

“Her third and fourth benchmarks were very quickly done, with lots of basic errors. So benchmarking did not stop this marker, as we were told it would. I have concerns for the quality of her marking as it was so rushed.”

A KS2 science marker with eight years’ experience said this year’s results would be “meaningless” because examiners were not able to discuss errors with supervisors and adjust their marking accordingly.

Previously, markers sent in samples of scripts they were marking as they went. The supervisor could then pick up on any common errors. Under the new system, markers carry out benchmarking checks against standard scripts online.

The KS2 science marker said there was little opportunity to correct marking in the light of errors.

The benchmarking exercise, and a similar one conducted before markers start work, do not result in any level being given to a child; they are a hypothetical exercise designed to assess the examiner.

A KS3 writing marker said: “Under the new system, the actual scripts are never sampled by another marker.

“An unscrupulous marker could spend hours over the benchmark scripts ... then race through the live scripts and no one would be any the wiser till those boxes spilt their contents over the staffroom floor.”

But National Assessment Agency figures, presented to Parliament last month, do not suggest a big change in the numbers of markers being barred. Last year, 72 out of 9,816, or 0.7 per cent, were thrown out in this way. In 2008, the figure was 81 out of 9,991, or 0.8 per cent.

An NAA spokesman said that benchmarking had led to 151 errant markers being stopped this year, compared to only 92 last year.