ÁñÁ«ÊÓƵ

"The big marking experiment" (18 September) raises the following questions. The reported span among grades may be encouragingly narrow, but those differences might matter all the more in a finely calibrated system, for example the American alphabetical scheme (A, A-, B+ and so on). Even if reliability among graders were relatively high, a system beset by grade inflation would merely yield reliably overstated grades. But just the same, one must laud an inquiry into grading inspired by a professor named Martin Luck.

Abbott Katz, London.