Teachers face an “almost impossible task” in fairly awarding A-level and GCSE grades in place of exams during the pandemic, according to researchers who found that students in England from highly educated backgrounds benefited from more generously assessed grades.
The study by University College London and the London School of Economics found that pupils whose parents had degrees were the beneficiaries of last year’s government U-turn that replaced grades determined by an algorithm in favour of grades decided by teacher assessments.
그만큼 survey of students found that even after adjusting for previous results and social background, those from graduate households were 15% more likely to get a better grade from their teachers than from the process using an algorithm created by Ofqual, the exam regulator for England.
Lindsey Macmillan, of UCL’s centre for education policy, said the findings underlined the difficulties that teachers and school face in fairly assessing their students for this year’s A-levels and GCSEs. Earlier this year the government again cancelled formal exams in favour of teacher-assessed grades.
“This shows that while the algorithm had some problems for high-achieving disadvantaged pupils, the alternative that was then put in place didn’t do much better,” Macmillan said.
“Looking ahead to this year, it’s more evidence of the unfairness of using teacher assessments as the final grade. This isn’t to say that teachers are to blame here – teachers are working harder than they’ve ever worked to get the assessments right this year but they’ve been given an impossible task. Some pupils are just harder to assess or predict grades for than others.”
Ofqual’s algorithm had downgraded nearly two out of every five A-level grades that were initially assessed by schools but would have awarded a record 27% of entries with the highest A or A* grades. After a storm of complaints the government instead used the school assessments – which reversed the downgrades and saw 38% of entries given A*s or As.
The UCL-LSE survey, from a nationally-representative sample of 4,000 pupils, including 300 with A-level results awarded in 2020, is the first to compare the social and educational backgrounds of students to the grades they received via the algorithm and assessments.
The researchers said there were several explanations for the 15% differential, including that the schools attended by better-off pupils were more aggressive in their assessments, or that graduate parents were more willing to lobby schools for higher grades on behalf of their children.
The survey found no difference in assessed grades between pupils on free school meals (FSM) and others, “suggesting teachers were careful not to treat FSM pupils differently. But they may nonetheless exhibit an unconscious positive bias towards pupils from backgrounds that tend to be associated with higher educational achievement,” according to the researchers.
Macmillan said schools had more time to prepare for this year’s assessments but that may not help them award grades more fairly.
“Students are getting assessed, it’s still happening, it’s not like exams were cancelled and they are having their usual school work. Anecdotally it sounds like they are getting assessed more than usual,” Macmillan said.
“But the problem is that teachers aren’t able to standardise their assessments across regions, and there’s no incentive for them to do so. There’s only limited challenge to the process so grades are going to vary across different schools and regions.
“Teachers can only deal with the pupils and the data that they have in front of them. This should never have been put on them in the middle of a pandemic.”
Ofqual and the Joint Council on Qualifications on Tuesday published their final exam appeals guidance, with schools bracing for a rush of appeals when awards are published in August.
Ofqual said it would not allow a separate appeals route for students with special educational needs or disabilities, and did not think it necessary to include guidance on possible bias or discrimination in assessing grades.