Comments by the National Postgraduate Committee on the 1992 Research Assessment Exercise (1992)

This section...

Also in Old Policy Responses Archive...

Written by James Irvine


The proposed timetable is very tight. We note that in future it is proposed to run these assessment exercises every three years. While this may be seen to provide the most up to date information, a longer period between assessments may in fact provide a better compromise considering the amount of work involved (as noted in point 9). Also, a longer period between assessments would allow more time for recommendations and changes to be carried out. A three year period is such that a PhD student recruited in the October after the assessment is unlikely to produce published work before the following assessment.


The definition of 'work of direct relevance to the needs of commerce and industry' should be clarified so that it is clear that work done in collaboration with industry and on industrial contracts should be included. This may include collaborative studentships and Teaching Company schemes as appropriate.

Departments which are simply set up as service departments for other parts of the institution or with other non-research aims should not be assessed. It may perhaps be appropriate to independently check the status of departments which institutions feel should be excluded, but it must be clear that an institution will not be penalised for justifiably excluding any department.

Subject units of assessment

If peer review is the chosen method of assessment, the subject units must remain sufficiently tight to allow this to be done as fairly as possible. Reducing the number of assessment panels risks the possibility of compromising effectiveness.

Method of assessment

Objective data provided by the institutions should form an important part of the assessment. For this reason, the information requested must be tightly defined, and the panels given guidance so that it is assessed in a consistent manner.

Definition of quality and Commonality of approach

It cannot be assumed the panels will 'recognise quality when they see it'. This approach will simply result in the same number of definitions of quality as there are panels. Full guidance and/or training must be given to panels to ensure a fair treatment of institutions. There is no review or appeal process available to institutions who feel that they have not been assessed fairly, and the proposed timetable does not allow adequate time of one. There should therefore either be auditing of the panels themselves to ensure consistency, or some departments should be re-assessed concurrently with the main exercise to prove the validity of the results.

Information to be requested

This should be more rigorously defined.

'Completed PhDs' should be expanded to define the time taken to complete (i.e. completions within four years of first registration), to ensure commonality between institutions.

Under 22.i., relevant MPhil degrees by research should also be included.

We fail to see the justification for restricting the number of cited publications to two out of the three year period. Survey panels should be familiar enough with their own field to be able to choose for themselves which publications to study in detail, and members of staff should be given a full opportunity to cite their work.

While postgraduates employed as Research Assistants will have their output assessed as members of staff, Research Students will only be recorded with reference to their number and source of funding. There should be a provision to include the research output of research students.