Guidance for Individual Projects
The Individual Project is, for most HetSys students, the first chance to work uninterruptedly on their research project, and where the student’s sole focus becomes goal-oriented novel research. In the context of the HetSys PG Diploma, successful completion of the project not only forms a progression gateway but also provides crucial input for the peer-to-peer UQ exercise that forms the second part of PX915.
The goals of the project should be discussed jointly at the start between the student and the supervisor and written up as a short document containing the following: i) a brief project overview (around 2-300 words), ii) a list of bullet point “deliverables” / “goals”, iii) an initial suggestion for what the documented “reproducible result” will be, to be handed over for the peer-to-peer exercise and iv) an outline workplan. Note that it is no problem if plans for the project evolve as it proceeds, depending on outcomes: the submitted outline is simply a record of initial plans. This document should be uploaded to the HetSys website by the end of week 4 of the summer term (Friday 22nd May 2020). Supervisors and/or students will be contacted if any concerns arise when these are reviewed.
The project goals should be structured such that:
- The work can produce demonstrable, discussable new results during the roughly 12-15 week timeframe before the progression vivas in September. Note that students will likely wish to take some time off in this period and that in normal years supervisors and students may also be busy with conferences and other travel (less so in 2020!). This should be incorporated into the setting of realistic goals for the project.
- The project should demonstrate evidence of some or all the main goals of the CDT training programme. These are: spanning of disciplinary barriers and time and/or length-scales, incorporation of uncertainty in modelling, and robust software engineering.
- It should be possible to create and document a clearly-defined “protocol” for evaluating or reproducing one result from the work, associated with an estimate of the uncertainty or error associated with this result.
The project report and viva are not assessed for credit within a CDT module. This means it can form a seamless part of the overall PhD project and be presented in the final thesis. If this work was for credit in the PG Diploma, then under university rules it could not later form part of the PhD thesis.
However, this project is in effect the final progression gateway to the PhD, so we assess its success in two ways in order that we can be confident students are ready to commit to a further three years of research on this topic: a brief report on the findings structured as a research paper, and a progression viva with the CDT management team to discuss this report.
We request that students write up their findings as if preparing to submit them as a research paper. This does not imply that a publishable body of impactful work should have been completed and polished. Instead the point is to learn how to structure and present your work as a narrative. The writeup should be around 5-10 pages, formatted as if for submission to a journal, and should contain a similar level of detail in the background and introduction to a typical research paper in its field. That implies i) a brief literature review with key references and background motivation that underpin the paper, ii) a methodology section that refers to other published work where appropriate, rather than trying to explain all details from scratch, iii) a detailed results and discussion section that highlights and explains key findings, iv) conclusions and plans for further work.
Supervisors are requested to keep the level of proof-reading of text to a minimum but to advise on contents, presentation of graphs etc (like they might for an undergraduate project, rather than for a collaborative paper). The report will form the basis of discussion in the viva, and we would be delighted if students then continued to expand on the work such that it was eventually published.
The report must provide information to enable full reproduction of at least one relevant result from the work, e.g. as supplementary material detailing steps to achieve this. This could entail downloading a ‘protocol’ script from an online repository (note: this does not have to be made public), compiling or locating binaries for any packages required, and running it on SCRTP hardware. This can involve HPC calculations, but these should be possible within moderate computational resources available to anyone in the CDT. An estimate of the uncertainty and/or error of this chosen result must also be provided in the report. This is necessary to allow the peer-to-peer assessment as part of the PX915 module.
Viva examinations will usually be 20-40 minutes in length and will be arranged in September following the end of the 12-14 week period associated with the Individual Project. The examiners will be 2 members of staff from the CDT management team, not including the project supervisors. Vivas in 2020 are likely to take place by videoconference unless Covid-19 restrictions are significantly lifted by that time.
The student will be asked to present a brief overview of their project and its main findings. There will be follow-up questions to probe the student’s understanding of the project and the way the work fits into and builds on the CDT training programme.
The appendix on reproducibility will be checked to determine whether it can be followed within reason by another student – if this is not the case, improvements may be requested.
General guidance for the students on oral examinations can be found on the website:
An audio recording will be taken of all oral examinations. Candidates will be asked to state their name and candidate number for the recording at the start. This may be provided to external examiners, so it should not record extraneous conversations.
Marking and Feedback
Feedback forms will be filled in during the viva by the staff – these will be provided to the students once all vivas have been completed for the cohort. Suggested minor corrections to the report should be completed and a final version uploaded to the website before the beginning of Term 1: this version will form the basis of the information provided to whoever is using this work as the input for the peer-to-peer UQ exercise.
The viva will be marked on a pass/fail basis, on the basis that the student must demonstrate a sufficient understanding of their project, and engagement with the goals of the CDT, and taking into account confidential feedback from the supervisor(s).
In the event that the viva examiners deem the student to have failed the viva there will be one opportunity to revise the report and resit within a few weeks. If satisfactory improvement is not observed, then the student’s progression to the PhD will be referred to the CDT management board (though we do not expect this to be a common occurance!).