Pilot test rogue results revealed

The test will now be more like the Sats children are used to

The government has explained some of the "unusual" results from the first pilot of new tests that might replace the Sats in England's schools.

Schools Minister Jim Knight told MPs that in a reading test, younger children did better than older ones.

He said other oddities might have been because teachers had entered children for the tests when they were not ready.

Mr Knight said if single level testing is not a positive evolution from the current system, "We won't do it".

The pilot tests were taken by 22,543 children around England.

The delivery of the results to their schools was delayed while officials looked into "unusual patterns" in them.

Failure rate

Mr Knight was quizzed about this when he gave evidence to the Commons children, schools and families committee's ongoing inquiry into assessment, and has sent the MPs a letter about the issue.

That was just a bit odd Jim Knight Schools Minister

He said an evaluation of the December pilot and the next one, in June, would be published in the autumn.

Teething troubles were inevitable and he did not want results to come out in a "drip, drip" fashion.

But he said the most significant unusual outcome had been variations between Key Stage 2 and 3 children - several years apart in age - taking the same test.

"So let's say they were taking a Level 4 writing test, the Key Stage 2 students doing significantly better - taking exactly the same test - as the Key Stage 3 student.

"Now that - that was just a bit odd," he said.

Mr Knight flatly denied that there had been a high rate of failure.

And he said the decision to make them more like existing Sats for the next round of pilots in June had been reached before the first ones had been taken, and was not a response to the results.

Wrongly entered

Labour MP Lynda Waltho asked him - "cutting to the chase" - whether the pass rate in the tests had been lower than expected.

Mr Knight said it had been better in some tests and worse in others - again he said the issue was the unexpected pattern of results.

In the usual national curriculum tests which all children take, they can achieve a range of levels.

In these new single level tests, teachers assess when their pupils are ready to take the test at a certain fixed level and the test is there to confirm whether or not that judgement is right.

One of the reasons advanced for the odd results is "inappropriate entry".

Mr Knight said a number of pupils had been entered with a teacher assessment lower than would have been expected if they were indeed working securely within the level being assessed.

Other possibilities put forward are:

• pupils perceived the tests as harder than they were used to
• pupils, especially the older ones, may have been less motivated
• a number did not get to the end of the papers
• markers were unaccustomed to marking scripts at a single level from pupils of different ages

In a wide-ranging session dealing with other aspects of tests and league tables, the government was challenged by members of the committee to come up with a more intelligible "value added" progress measure.

Committee chairman Barry Sheerman said the MPs represented constituents with a wide range of educational attainment and they all needed to be able to understand it.

"Contextual value added is unintelligible - only understandable to people who have higher levels of qualification. That's unfair," he said.

A simple banding system was suggested as a way forward.

Mr Knight said the government would naturally reflect on any recommendations the committee made.