Other PhD supervisors bend the rules by passing "dicey" theses to examiners who are known to be lenient.
Such unethical practices in higher education have been highlighted by two academics who have studied PhD supervision in social science and education departments.
Dr John Hockey, a research fellow at Cheltenham and Gloucester College of Higher Education, refers to the rewriting of PhD students' work in the latest issue of Research in Post-Compulsory Education.
His claims are endorsed by Dr Christine O'Hanlon, a senior lecturer at Birmingham University's School of Education, who has produced an unpublished paper on PhD supervision. She estimates up to 15 per cent of doctoral theses are so poor they have to be substantially rewritten by the supervisor before they can go to an examiner.
"I will be open about it because it is time this was exposed," she said. "There's an awful lot of dodgy practice. Sometimes the rewriting is needed because an overseas student's English is not good enough. But some home candidates are also weak."
Dr O'Hanlon said some supervisors provided more help than they should because they felt intensely responsible for a student. "You form such a strong bond during the four years you are together and believe you are letting them down if they don't pass.
"I know of one student who became almost suicidal when she realised that she might not get her doctorate. Almost all her work was rewritten for her - at least three or four chapters."
A minority of the supervisors Dr Hockey interviewed told similar stories. He believes the rule-breaking occurs because PhD supervisors are expected to be both guide and critic. Inexperienced supervisors often struggle to gauge the correct amount of help they should offer, he says.
But the second dubious practice that he identifies - seeking out lenient examiners - cannot be attributed to lack of experience. As one supervisor told him: "When you get as far as selecting an external examiner what you do is you judge the PhD and say, um, the better the PhD the more prestigious the external examiner. The more dicey the PhD, have I got a friend who'll be lenient?"
Many academics will challenge the claims. "A lot of the talk about rewriting students' theses is rubbish," one professor of education said this week. "A PhD thesis can be 70-100,000 words long, so where would they find the time?"
"A Complex Craft: United Kingdom PhD supervision in the social sciences", by John Hockey, Research in Post-Compulsory Education (vol. 2 no. 1 1997 - publication delayed). The full text is accessible via Triangle Journals' website found at: www.triangle.co.uk