Covid corner: A levels algorithm fiasco

The recent A level results fiasco gives data science a bad name. It’s a really useful illustration of how important it is to consider how algorithms can propagate existing inequalities.

Data scientist Sophie Benne has taken a fine tooth comb over the ofqual sport on the algorithm used and it makes really interesting reading.

“Standardising the grades at a national level means that the high grades of these students effectively pushes down the grades of students in larger schools, who are already subject to more downgrading. Since larger schools are more likely to be state schools, and small schools are more likely to be independent schools, this increases unfairness and disparity even further[also]

  • The model does not take into account variability in grades from one year to the next.
  • The algorithm does not seem to properly account for differences in the value-added across schools.
  • The reliance on relative rank order
  • No quantification of uncertainty.”

This case study should be taught in universities looking at fairness in AI.

📖 Read more here (3,868 words) 📖