The article investigating PhD drop-outs using a sample of "successful doctoral candidates" ("Doctorates: why so many give up", THES, November 19) reminds me of a (probably apocryphal) story about an air force trying to decide where on the bodies of fighter planes extra armour would be most useful. They decided to count the bullet holes in different parts of planes returning from combat, and to put the extra armour where there were most holes. The results of their study surprised them; for example, it seemed that they should not put extra armour over the engine or fuel tanks, as very few holes occurred there.
Eventually somebody noticed the flaw in their study: they were counting the holes only in the planes that came back.
Perdita Stevens Division of informatics University of Edinburgh