É«ºÐÖ±²¥

Ormond Simpson¡¯s letter says that in 2006 it had been possible to forecast the likely success of a new student entering the Open University, although this system was never used (¡°What are the chances?¡±, 20 August). An OU story that was in the press recently was about a rather less ambitious use of data ¨C an algorithm the university uses to predict individual student failure to submit an assignment (The week in higher education, 30 July). Whichever story we focus on, we are in the territory of being encouraged to study the behaviour of an ant while sitting on an anthill.

For whatever predictive data we focus on, the perplexing fact remains that the OU is very happy to have its methods of tracking failing students widely publicised in the name of ¡°innovations in data analytics¡± but at the same time is recruiting a new intake of students for the autumn when about 40 per cent or more of them are likely to fail their first year of study. They fail because they are ill-suited and ill-prepared in terms of skills and time for university study, especially the rigours of distance learning.

That is the real story, and it is surely immoral that a university should allow such large numbers of students to enrol (80 per cent via a loan) without doing anything to check that those students have a decent chance of success, and to personally counsel them to ensure that this is the best decision for them to take.

Name and address withheld