Honours grades are based upon the following scale:
| Grade |
Final % |
|---|---|
| First Class: H1 | 80-100% |
| Upper Second: H2A | 70-79% |
| Lower Second: H2B | 60-69% |
| Third Class: H3 | 50-59% |
| Fail | < 50% |
It is expected that any work submitted for assessment will be the student's own work.
Plagiarism in Honours is likely to be detected given the relatively small number of students in the Honours year.
Severe penalties will result which can seriously weaken the student's degree result and also compromise the student's reputation as a scholar.
The final Honours grade is calculated from the results of five lecture sub-units and the Honours Project mark.
At the Honours Examiners Meeting held at the end of the year, marks for individual lecture sub-units may be rescaled if the sub-unit average mark appears significantly out of line with other results.
The mark for the Honours project is based on three elements:
Each dissertation is examined by an Internal Examiner, from the same section as the project supervisor, and an External Examiner from a different section. The Internal Examiner may be the project supervisor only under exceptional circumstances.
The Honours project marks are awarded in light of a written statement by the supervisor describing the input of the student, pointing out any original results, new proofs, new organisation of material, and the extent of library investigations, and according to the assessment criteria.
Each student is provided a brief report outlining the main strengths and weaknesses of the dissertation.
If the Examiners cannot agree on a result, or when their overall project marks differ by more than 10 per cent, the Examiners will try to come to agreement. Should they still fail to agree, the Honours Co-ordinator will appoint a third Examiner who will mark the dissertation and meet the Project Examiners to agree upon a final mark.
In the rare event of continued disagreement, the final mark will be decided at the Honours Examiners Meeting.
For an outstanding, well-structured dissertation in a difficult or challenging area which displays a high level of mathematical/statistical sophistication and understanding, a substantial degree of insight and independent thought, and will typically contain new material. Appropriate theoretical mathematical/statistical tools will be applied with rigour, and results will be interpreted with logic and flair.
As for HD+ but with some small weaknesses, such as where a presentation is not entirely clear on a few points. There may be trivial errors in the mathematics/statistics and some small flaw in arguments.This is typically the highest grade that will be awarded to a dissertation containing no new material.
For a dissertation showing excellence in one or two aspects but with no particular strengths elsewhere. An excellently written dissertation reviewing a current area of research and containing a relatively small number of errors would typically be graded at this level, as would a dissertation containing moments of deep insight but with a number of minor errors.
For a consistently sound piece of work written up in a well structured and clear fashion. A dissertation graded at this level would not be expected to show any deep insight but neither would it contain any serious errors in the use of mathematical/statistical tools or in the interpretation of results.
For a dissertation fairly well littered with minor errors and without any areas of real excellence. An otherwise sound dissertation containing a significant error at some point which does not affect the majority of the work would typically be graded at this level. A sound review of a relatively accessible area of mathematics/statistics would also usually attain this grade.
For a dissertation containing a number of significant errors in the use of mathematical/statistical tools or in the interpretation of results. A poorly written and structured dissertation containing no more than a moderate number of such errors would typically be graded at this level. A reasonably structured review of a relatively accessible area of mathematics/statistics, but containing several non-trivial errors, would also usually attain this grade.
For a dissertation containing a large number of significant errors, or a number of fundamental errors rendering the majority of the material incorrect.