Final Year Undergraduate Project Report Writing Guidelines Who is the report written for? Structure Referencing Only the written work is graded so make.

Final Year Undergraduate ProjectFinal Year Undergraduate Project

Report Writing GuidelinesReport Writing Guidelines

•Who is the report written for?

•Structure

•Referencing

Only the written work is graded so make sure you get individual advice at your progress meeting about how to evidence any other deliverables.

Format guidelines etc. on website.

School of ComputingFACULTY OF ENGINEERING



Who is the report written for?

•Three markers

• Supervisor

• Assessor

• External Examiner

•The Computing audience

• Future SoC students

• Students at other HEIs

• researchers




Report Structure

•Introduction

•Understand the Problem (20)

•Produce a Solution (40)

• Preparation of Solution (20)

• Delivery of Solution (20)

•Evaluate the Solution (20)

•Reflect upon the Project Experience (5) [Appendix A]

[Remaining 15 marks for write-up]




The Introduction

• What is the problem/question you are investigating?

• What is the starting point?

• The evidence that there is a problem?

• Open question(s) in the research literature?

• New techniques / tools that need evaluation?

• What have you done about it?

• How is the remainder of the report organised? What has been provided (e.g. raw data, source code, etc) in annexes?




Understand the Problem

Good Practice•Clear evidence of problem statement and systematic strategy for identifying relevant literature•Comprehensive review of literature•Engages actively with literature and uses appropriate level of detail•Relevance to COMP and research•Approach adopted is compared with other potential approaches and the choice justified


Poor Practice•Problem not put in context with small no. of mainly secondary sources and/or too much WWW.•Literature is mainly descriptive and quoted verbatim•Limited relevance to COMP•Limited engagement with full problem and current practice•No justification of the choice of approach



Produce a Solution - Preparation

Good Practice•Clear evidence of planned approach with revisions clearly identified where appropriate•Local version elucidated•Methodology, design choices and evaluation criteria justified and referred back to problem•Judgements and details about planning in Appendix A


Poor Practice•Limited planning not related to project development•Development choice not related to problem•No justification of the methodology to be followed•Limited discussion on how this plan will solve the original problem



Produce a Solution - Preparation (2)

Good Practice• What were the hard constraints?• existing codebase in X• must run on platform Y (why)?

• What choices did you make: why?• For writing the GUI, Qt was used, building on taught material in CR20 and thus reducing the learning overhead. As Qt provided all the fetaures required, other alternatives (e.g. wxWindows) were not considered further.


Poor Practice• I used Python because its a widely-used language with tools for doing this and that, and runs on many platforms. The program was written on Windows because its the market leader in systems.



Produce a Solution – Delivery [min reqs do not need reproducing]

Good Practice•Clear evidence (e.g. limited screen shots, source code) of meeting and exceeding minimum requirements•Discussion of appropriate sections of implementation and necessary changes•Background reading related to design and methodology as appropriate


Poor Practice•Insufficient evidence of meeting minimum requirements•Relies on extensive and repetitive screen shots/diagrams/source code•Minimum requirements not exceeded.



Evaluate the SolutionGood Practice

•Criteria close to publishable quality•Criteria explained and justified•Clear evidence that there was an evaluation plan which was adopted.•Verdict offered withstands scrutiny


Poor Practice•Reasonable but only basic criteria are defined•Only limited attempt to justify evaluation criteria•Only limited evidence that criteria have been applied.•Only limited conclusions about solution.



Evaluate the Solution (2)More Good Practice

• The best algorithm currently known is O(n2 log n) [ref].• By implementing path-compression [ref] for the union-find structure, run-time on the trial dataset improved from 49s to 14.4s.• Figure X shows the change in runtime against data size, suggesting that the improvement is linear in the problem size.


More Poor Practice• The problem is very hard to solve efficiently.• By changing the code it was possible to make the program run rather faster.• This helps even when the data becomes bigger.



Evaluate the Solution (3)More Good Practice

• A trial was conducted on a sample of N users to test the following hypothesis: ...... Summaries of the sample population are in Annex B. The following protocol was followed: ... The raw data (see Annex C) were analysed using <insert statistical test>. Figure Z shows the results, allowing us to confirm the hypothesis. (Or, if it was refuted, what happened?).


More Poor Practice• The program was trialed on a group of users who were positive about the interface.• On average, 65% of users completed the task faster.• Feedback from the survey was generally positive ...



Appendix AGood Practice

•Mature reflection offers useful guidance to future students attempting projects in the topic area or using similar methods


Poor Practice•No Appendix A!•‘Obvious’ and generic observations about assessment of the process•Not particularly reflective or insightful



Write-UpGood Practice

•Very well written•Shows an appreciation of the intended audience•References organised appropriately•Appropriate use of appendices for unpublished background and verifying material•Competently proof-read


Poor Practice•Minimally competent (or incompetent) use of grammar and spelling•Unclear structure with too much or too little use of subsections•Exceeds page limit without prior permission – beware waffle!•Poor balance of material between content and appendices



Report Structure

•CLEAR through clarity of thought

•SIMPLE so your ideas can be grasped at first reading

•CONCISE: avoid waffling [page limit is a maximum not a target!]

•UNAMBIGUOUS: beware of potential misinterpretation




Report Structure

•Professional and generally technical. Format and content are important. Avoid colloquialisms, slang, text speak, chatty style.

•Do not compare to previous reports. Even if a report got over 70, that’s up to 30 marks they didn’t get and you don’t know the marks breakdown.

•Discusses full problem, focuses on solution and relates back again to problem. Sell your achievements!




Report Structure

•Flows appropriately [reproduction of aim, min req etc. were project issues not problem issues so can be left out as long as planning is still clear]. Always ask WHY you are putting something in. Was the Waterfall Model REALLY an appropriate methodology to report on given it is hardly ever adopted? Write a single integrated report.

•Be very careful how background material is referenced. Avoid too many verbatim quotes as this does not show engagement or relevance of literature. OK for definitions or if you need to critically evaluate.

•http://www.leeds.ac.uk/AAandR/cpff.htm




Referencing

•It ensures that the reader can check the facts

•It acknowledges the contribution of others

•It helps others wishing to undertake similar projects or to pursue associated lines of enquiry

•Is a good thing! You are not expected to know everything and should build on sources

•Referencing refereed sources is vitally important.




Suppose you’ve read the following during your research:

“The scope of the relaxation scheme used involves employing both the Gauss-Seideland the Jacobi line relaxation schemes on the same grid, but without any overlap, depending on the position of the grid point (i; j) on the computational domain.”

From:Goodyer CE and Berzins M, “Parallelisation and scalability issues in a multilevel EHL Solver”, School of Computing Research Report Series, 2005.05, 2005




The supervisor sees

The scope of the relaxation scheme used involves employing both the Gauss-Seideland and the Jacobi line relaxation schemes on the same grid, but without any overlap, depending on the position of the grid point (i; j) on the computational domain.

References:

<None>

Action:Action:




The supervisor sees


References:<None>

Action:Action:

Reported for plagiarism. No acknowledgement of source.Reported for plagiarism. No acknowledgement of source.




The supervisor sees


References:

[1] Goodyer CE and Berzins M, “Parallelisation and scalability issues in a multilevel EHL Solver”, School of Computing Research Report Series, 2005.05, 2005

Action:Action:




The supervisor seesThe scope of the relaxation scheme used involves employing both the Gauss-Seideland and the Jacobi line relaxation schemes on the same grid, but without any overlap, depending on the position of the grid point (i; j) on the computational domain.

References:[1] Goodyer CE and Berzins M, “Parallelisation and scalability issues in a multilevel EHL Solver”, School of Computing Research Report Series, 2005.05, 2005

Action:Action:

Reported for plagiarism. No clear acknowledgment of source in text.Reported for plagiarism. No clear acknowledgment of source in text.




The supervisor seesGoodyer and Berzins [1] discuss this issue. The scope of the relaxation scheme used involves employing both the Gauss-Seideland and the Jacobi line relaxation schemes on the same grid, but without any overlap, depending on the position of the grid point (i; j) on the computational domain.


Action:Action:




The supervisor seesGoodyer and Berzins [1] discuss this issue. The scope of the relaxation scheme used involves employing both the Gauss-Seideland and the Jacobi line relaxation schemes on the same grid, but without any overlap, depending on the position of the grid point (i; j) on the computational domain.


Action:Action:

Investigated and probably reported for plagiarism. No quotation marks.Investigated and probably reported for plagiarism. No quotation marks.




The supervisor seesGoodyer and Berzins [1] discuss this issue: “The scope of the relaxation scheme used involves employing both the Gauss-Seideland and the Jacobi line relaxation schemes on the same grid, but without any overlap, depending on the position of the grid point (i; j) on the computational domain.”


Action:Action:

None but no engagement with source.None but no engagement with source.




The supervisor seesGoodyer and Berzins [1] discuss the issue of relaxation schemes and for the problem discussed here it is appropriate to apply both the Gauss-Seideland and the Jacobi line relaxation schemes on the identified domain.


Action:Action:

None. Excellent student! Paraphrased and engaged to show an None. Excellent student! Paraphrased and engaged to show an understanding.understanding.




Referencing•Make the extent of the source clear.

•Use quotation marks where material is verbatim.

•Identify original and refereed sources.

•Do NOT use Wikipedia for definitions. Anyone could write this. It may be helpful as a signpost.

•Refer to the location of published articles even if they are reproduced on the Web. Web sources are generally frowned upon but some are more trustworthy than others and may be necessary with newer research.

•Use lots – but engage with the sources and put them in context.

•Use ‘personal communication’ if necessary on the same basis as material not referenced.


Final Year Undergraduate Project Report Writing Guidelines Who is the report written for? Structure Referencing Only the written work is graded so make.

Documents

limited evidence

current practice

original problem slide

project experience

preparation of solution

format guidelines

problem size

delivery of solution