"The key may be tracking the performance of those who supervise doctoral students." summarizes Matthew Reisz for Times Higher Education.
Universities can improve their Ph.D. completion rates by using metrics to assess the performance of supervisors, Australian academics say.
Richard Russell, a former pro vice chancellor for research operations at the University of Adelaide, told the Third International Conference on Developments in Doctoral Education and Training that, while everyone knew academics who “should never have been allowed to supervise Ph.D.s,” institutions that tolerated this were “failing in their duty of care.”
At the event, organized by Britain’s Council for Graduate Education, Russell explained how leading research universities in Australia’s Group of Eight agreed to require the formal training and registration of doctoral supervisors as long ago as 2004-05. More recently, Adelaide realized that it “needed to optimize candidatures to hold our levels of funding and scholarships.”
All supervisors were therefore assessed on their number of past students, current “load” and an index designed to capture “outcomes versus opportunities.” The university was keen to reward supervisors for “timely” completions, other completions and “student rescues,” when someone about to abandon a thesis was persuaded to stay on. It wanted to penalize noncompletions and withdrawals due to dissatisfaction with supervisors, but to remain neutral about early withdrawals, student-initiated withdrawals for nonacademic reasons and failed rescue attempts.
The result, Russell said, was a much more effective system for classifying and tracking the performance of supervisors. This has led to problems being addressed earlier, the removal of “totally unsatisfactory supervisors” and an 8 percent increase in timely completions.
Source: Inside Higher Ed