Predicated Grades: A Predicted Fiasco?

Author: Clare O'Hare and Anna Dolezal

European summers are usually long and quiet. Summer starts and ends with national school exit exams. The end of the exams marks the beginning of summer and of childhood and results day marks the end of summer and the beginning of adult life, with students finding out if they will be going on to university.

COVID-19 threw this usual rite of passage into chaos this year, with exams cancelled across much of Europe. Typical results-day excitement and anxiety were replaced this year by deep anger and confusion as thousands of students, parents, and teachers found out that assigned-student grades were much different to those predicted. Within days, exam results were annulled and, in some cases ruled invalid (see forthcoming post on Norway and the IB).

When it became clear that students were unlikely to return to school after Easter, countries across the world altered or cancelled the exams. In France, the Bac was cancelled for the first time since Napoleon created it in 1808. Even where the exams did go ahead, it was often with modifications that reflected the disruption to education this year, such as giving students a greater number of questions to choose between or examining fewer topics and holding exams in larger spaces that allowed for more space between each student.

In order to assess the impacts of the exam results this year, it is important to understand the importance of these terminal exams in Europe. Unlike in the U.S., most European students only receive final university offers in the summer after they finish school and receive exam results. It is virtually impossible for European students to move directly to university without passing these exams. These exams are a university matriculation requirement and, in some countries, the exam is used as a selection tool, with universities requiring specific exam results for specific degree courses. Because students in Europe tend to apply to specific degree programs, rather than a university in general, in “selective” systems, entry requirements can vary greatly across disciplines within a single university, and students often receive “conditional offers” from universities during their final year at school. Universities are only bound to accept the student if the student achieves the grades set out in the offer.

With exams cancelled, education departments had to find an alternative way to assess students and award diplomas and grades. This scramble to find new ways of assessing students raised questions surrounding both the ethics and the accuracy of the proposed alternatives. In the Netherlands, schools were responsible for awarding grades to students, based on course work and in-school assessment. France took a similar approach. The success rate in both countries was up significantly last year, leading some to question the value of this year’s exams. But even in Germany, where exams went ahead, the resulting scores were a little higher than normal.

1796745946 Medium

Other countries, and the International Baccalaureate Organization, turned to a system variously called calculated, reported, or estimate grades, and this is where the controversy really lies. While there are differences between the systems, on the whole the approach relied on three prongs: a teacher assessed grade, a within-school and within-class ranking, and a national standardization process incorporating historical national grade distributions and school-level data exam results for each subject to produce a grade distribution for each school, which was then overlaid onto the teacher and school assessment.

Parents and students raised alarms about the fairness of the system from the outset, particularly to students from ethnic minorities and economically disadvantaged backgrounds and those attending schools with historically lower grades. While almost everyone recognizes that the task given to examination boards to recreate “normal” exam results this year was both thankless and virtually impossible, students, teachers, parents, and other stakeholders had little to no clarity on the methodology that would be used by the various educational authorities.

After the IB results were released in Norway, the Norwegian Data Protection Authority issued a draft decision that the calculated grade process resulted in personal data being used in an unfair, inaccurate, and obscure manner, and thus the grades awarded to Norwegian students could not stand. In Scotland, England, Wales, and Northern Ireland the calculated grades were eventually scrapped, after almost 40% of grades were downgraded. The harsh reality is that calculated grades in many respects reflected entrenched inequality in the UK and will in some cases exacerbate these inequalities. The disparity between teacher predicted grades and published grades, was as many had predicted, particularly stark for students from disadvantaged areas, who were most likely to have grades downgraded.

In England, a student who predicted an A and two Bs received 3 D grades, squashing her hopes of pursing a degree in veterinary science. Another student from South-West London was looking forward to not only being the first of his family to go to University but to studying medicine at the University of Cambridge. Unfortunately, no one from his school had ever received the grades necessary to study at Cambridge, so the algorithm downgraded his teacher’s assessment of his performance, edging him out of his anticipated place. In addition, the perceived social inequity of the model was increased further by the difficulty in applying a statistical model to small numbers. Thus, grades predicted by teachers were automatically awarded in classes up to five students, ultimately favoring private schools with smaller class sizes. Whatever fears existed that relying solely on teacher’s predictions would lead to massive grade inflation, Paul Johnson of the Institute for Fiscal Studies remarked that the end result was “grade inflation for the already privileged and little or nothing for the rest.”

In Scotland, the education secretary, John Swinney, ordered the Scottish Qualifications Authority to abandon the statistical assessments and revert to those grades recommended by teachers; 76,000 of 124,000 downgraded students saw their marks rise. England, Wales, and Northern Ireland soon abandoned their standardization model giving students the higher of teacher predicted or standardized grade. While good news for students, the “higher of” approach meant more students than expected met their conditional offers. This, on top of new COVID-19 social distancing measures has left universities struggling to find ways to physically welcome all students and finance the increase in numbers. In Scotland, for example, students are entitled to four years of free tuition, and universities estimate they need an extra £25 million in government funding to pay for these extra places, making this saga potentially costly for years to come.

This won’t be the last we hear of calculated grades. The Norwegian DPA will issue a final ruling on the standing of IB grades in the coming weeks, which will give more insights into the relationship between data privacy, automated processing of data, and exam results. And what of students who don’t attend traditional schools? Although homeschooling is fairly rare across Europe, a student in Ireland won a case against being excluded from consideration for calculated grades (the student did not submit that he was entitled to calculated grades, just that he should be considered for such grades). The case may also impact students who attend regular schools but who took an additional subject outside school taught by a relative.

The meltdown of exams in the UK and across Europe highlights many of the ways normal life has been upended by the pandemic. Governments and, in many cases, civil servants, are hastily putting together band-aid solutions to ever-evolving problems. But solutions that increase the uncertainties and additional hardships on marginalized populations that are already hurting the most during this crisis are not solutions. Young people navigating the world of virtual learning and graduating into an abysmal job market, despite not being as vulnerable to dying of coronavirus, now face the crippling uncertainty caused by the mishandlings of these exams. The fiasco runs the risk of undermining trust in the fairness of the education system. This, along with the many instances of politicians acting as though the COVID-19 rules don’t apply to them, creates an atmosphere of mistrust of government, right when this trust is most needed. While many of us hope normalcy awaits us as soon as an effective treatment of the vaccine is produced, the effects of auxiliary crises such as this will likely be felt for years to come.