THESIS WORKSHOP - PowerPoint Presentation

Slide1

THESIS WORKSHOP

Dr Sharon Goldwater(School of Informatics, University of Edinburgh)Murdo Gillanders (Postgraduate Office, College of Science & Engineering, University of Edinburgh)

Slides adapted from Jon

OberlanderSlide2

Plan for today

What is expected of your thesis?How to get startedInteractions with your supervisorMotivation and mechanicsThe viva and afterwardTHESIS WORKSHOP2Slide3

Discussion

What makes a good (or bad) thesis?Who will read it and why?THESIS WORKSHOP3Slide4

Examiner’s report form (U of E)

Is the thesis an original work that makes a significant contribution to knowledge in or understanding of the field of study?  YES NO

Does the thesis contain material worthy of publication?

YES NO

Does the thesis demonstrate adequate knowledge of the field of study and relevant literature?

YES NO

Does the thesis show the exercise of critical

judgement

with regard to both the student’s work and that of other scholars in the same general field?

YES NO Is the presentation and style of the thesis satisfactory?   YES NO

THESIS WORKSHOP

4Slide5

State your contributions explicitly

Don’t make the examiners work to infer what’s new/good.State your contributions in intro AND conclusion (AND in each chapter!)Be obvious! (Section heading, bullet list, etc.)THESIS WORKSHOP5Slide6

Publication-worthy material

Easy if you have already published papers!Aim for 2-3 conference papers, ideally in high-quality venues.But: if time is short, prioritize the thesis.THESIS WORKSHOP6Slide7

Demonstrate knowledge of the field

Include background material thatexplains your methods (or those you build on)justifies your choicesacknowledges other approachesDON’Tprovide a brain dump of everything you knowinclude details of other work just for the sake of itTHESIS WORKSHOP7Slide8

Demonstrate critical judgement

In background: evaluate, don’t just describe.In your work:Claims are supported by evidence.Arguments are well-reasoned.Limitations are acknowledged.THESIS WORKSHOP8Slide9

Presentation and style

Clarity and organization matter!Define terminology; be consistent with notation.Use examples to illustrate difficult concepts.Make figures/tables easy to read, with captions that stand alone.THESIS WORKSHOP9Slide10

Unified body of work

Different styles of thesis:Unifying taskUnifying theoryUnifying methodTHESIS WORKSHOP10Slide11

What examiners like to see

Quality of work:Focus on interesting/important problem.Groundbreaking theories/methods/results.Publishable in high-quality journals.Clear openings for future work.Well-reasoned, well-designed, reproducible experiments; preferably hypothesis-driven.Quality of thesis:Logical presentation of data and arguments.Clear, easy-to-interpret figures and graphs; good English.Short! (No extra words)

THESIS WORKSHOP

11Slide12

What examiners don’t like to see

Uninteresting or unimportant problem, or interest and importance not properly introduced.Poor reasoning, badly-designed experiments with missing controls, insufficient analysis.Work that only repeats or confirms well-established findings made by others.Sloppy presentation of graphs and illustrations.References missing or incorrectly cited.Grammatical and typing errors, convoluted sentences, waffle.THESIS WORKSHOP12Slide13

Typical thesis structure

AbstractIntroductionBackground~3 content chaptersDiscussion/ConclusionReferences(Appendices)THESIS WORKSHOP13Slide14

Introduction

Introduce the problem and motivate its importance.Outline your hypotheses and approach.State any important assumptions.Summarize your main results/conclusions.Outline the structure of your argument/ dissertation.THESIS WORKSHOP14Slide15

Background material

Should have a clear point, and only as much detail as needed to make that point.Explaining a method you use/build on: details ok.Clarifying the novelty of your approach: focus on similarities/differences to other work.Justifying your choice of method: focus on pros and cons of other work.THESIS WORKSHOP15Slide16

Background material

Don’t tear down others’ work.May be your most interested readers!OK to “build on” or “be inspired by”.Cite your examiners, carefully.No blatantly unrelated papers, but make sure not to forget relevant ones.THESIS WORKSHOP16Slide17

Content chapters

May include additional background material, experiments, proofs, algorithms, etc.Make sure the introduction/conclusion of each chapter explains how it fits into the bigger story.THESIS WORKSHOP17Slide18

Discussion/conclusion

Summarize main results/conclusions.Discuss any limitations or areas for future investigation.Reiterate why the work is important.THESIS WORKSHOP18Slide19

What is expected of your thesis?

How to get startedInteractions with your supervisorMotivation and mechanicsThe viva and afterwardTHESIS WORKSHOP19Slide20

THESIS WORKSHOP

20DISCUSSIONWhat is the aim of your project?Why are you doing this? What is interesting about it?Slide21

THESIS WORKSHOP

21ThinkWhat have you actually found out so far?Slide22

THESIS WORKSHOP

22Now write for 5 minutes covering:What you are doing and whyWhat you have discovered so farSlide23

THESIS WORKSHOP

23You should have a good start for your abstract for only 15 minutes workSlide24

Initial outline

Decide what will go in each chapter.Make an outline that includes your main points, not just section headings.Are you making a coherent argument?The order of results need not be the order in which you obtained them.Discuss with supervisor before you start filling in text.THESIS WORKSHOP24Slide25

Outline with section headings

THESIS WORKSHOP25IntroductionBackgroundMethod MySys EvaluationExperiment 1: Noise Data Results and DiscussionExperiment 2: OtherData Data Results and DiscussionConclusionSlide26

Outline with main points

THESIS WORKSHOP26Intro:- previous chapter showed X, but Y still unresolved. Here, do Z to answer Y.Background:- previous methods did A or B.- A is better for Q, B is better for P, but both have potential problems with noise....Experiment 1: Noise- describe standard data set and method for introducing noise- results for varying noise levels show MySys is more robust to noise than

OtherGuy

- but could be an artifact of this data set

Experiment 2:

OtherData

- shows that Experiment 1 results generalize.

…Slide27

How your supervisor can help

Thesis is your responsibility. BUT: your supervisor has more experience, and can help.Agree an outline before you start writing.Including: what background material is needed.Provide full chapter drafts for feedback.Point out places where you have questions or have made changes.Don’t ignore advice. If you disagree, discuss it.Ultimately, you will need to defend your thesis.THESIS WORKSHOP27Slide28

What is expected of your thesis?

How to get startedInteractions with your supervisorMotivation and mechanicsThe viva and afterwardTHESIS WORKSHOP28Slide29

How to get on with it

Write something every day; set goals.Do it first (before email, Facebook, preparing for reading group, …)Turn off the Internet.Don’t look up references, details of results, etc.: leave holes/notes/bullets (use a to-do macro).Write first, edit later.Find a thesis buddy (or two).THESIS WORKSHOP29Slide30

If you are stuck

Talk to your supervisor.DON’T run away and hide.DON’T wait until you have “something to show”.Consider setting shorter/more specific deadlines together.THESIS WORKSHOP30Slide31

Editing

Read through entire chapter/thesis.Have I defined terms before using them?Is my notation/terminology consistent?Is my argument clearly structured?Is it clear what the point of each section is and how it fits into the rest of the chapter/thesis?Are cross-references correct?Are bibliography entries complete and correct?Ask a friend to proofread.THESIS WORKSHOP31Slide32

Good scientific writing

Choice of methods should be justified.Claims should be supported by evidence.Don’t over-interpret results.Distinguish main claims from speculations.THESIS WORKSHOP32Slide33

How much detail?

Your work should be reproducible.Conference papers often leave out some key details; don’t forget to add them back in.Don’t necessarily include every experiment.Failed attempts can be mentioned in passing or omitted entirely. Discuss with supervisor.Write for an intelligent Ph.D. student.Yourself three years ago, or the next office down.THESIS WORKSHOP33Slide34

No waffle, no padding

Be precise.Avoid vague comparisons: “Similar” in what way? “Better” along what axis? “More X” than what?Be concise.Does the material support a point?Am I repeating myself? (Brief summaries OK but don’t repeat whole sentences/ideas/arguments; use backward/forward references.)Can I say the same thing in fewer words?THESIS WORKSHOP34Slide35

Writing resources

Online (mostly sentence-level mechanics/clarity):http://homepages.inf.ed.ac.uk/sgwater/resources.htmlhttp://homepages.inf.ed.ac.uk/imurray2/teaching/writing/Books (check library):Style: Toward clarity and grace by Joseph M. Williams. (General writing; organization, clarity, mechanics)The Craft of Research by Wayne C. Booth, Gregory Colomb, and Joseph M. Williams. (Scientific investigation and scientific writing)Useful for all writing, not just thesis. Start now!

THESIS WORKSHOP

35Slide36

Avoiding plagiarism

Plagiarism: the act of copying or including in one's own work, without adequate acknowledgment, intentionally or unintentionally, the work of another, for one's own benefit.THESIS WORKSHOP36Slide37

Acknowledgements and citations

Acknowledge/credit any work done by others, incl. collaborative work.Explicitly acknowledge each individual piece of work in a thesis done by another person.Reference and fully cite other people’s text and arguments (published or unpublished). It is better to be overgenerous than to risk accusations of plagiarism Formal guidance: see University website or postgraduate handbookTHESIS WORKSHOP

37Slide38

THESIS WORKSHOP

38The perils of cut and paste…..Original text:“The number of doctorates awarded in the UK in all subjects increased by about 18 per cent between academic years 1995/96 and 1999/2000. The biggest growth areas for UK students were social studies and law (an 80 per cent increase in each), creative arts (over 110 percent growth), and education and leisure (130 percent growth). Over the same period, UK-domiciled students gained around 600 additional doctorates each year in biological and related subjects. However, the number of doctorates awarded to UK-domiciled students in the physical sciences fell by 9 per cent between 1995/96 and 1999/2000.”from “Set for success”, The Report of the Sir Gareth Roberts’ Review, 2002

Avoiding plagiarism in the background/review sections of your thesisSlide39

THESIS WORKSHOP

39Is this plagiarism?Example 1…… The number of doctorates awarded in the UK in all subjects increased by about 18 per cent between academic years 1995/96 and 1999/2000. Over the same period, UK-domiciled students gained around 600 additional doctorates each year in biological and related subjects. However, the number of doctorates awarded to UK-domiciled students in the physical sciences fell by 9 per cent. …….

If you cut and paste from a

source,

it is plagiarism

unless you include quotation marks

and

attribution.

Example 2

…… The number of doctorates awarded in the UK increased by about 18 per cent between academic years 1995/96 and 1999/2000. The number of PhDs awarded to UK-domiciled students in the physical sciences fell by 9 per cent over the same period, and UK-domiciled students gained around 600 additional doctorates each year in biological and related subjects. …….

Changing the odd word here and there or changing the word order without quotation and attribution is also plagiarism.Slide40

THESIS WORKSHOP

40Is this plagiarism?Example 3…… According to Roberts (2002) “The number of doctorates awarded in the UK in all subjects increased by about 18 per cent between academic years 1995/96 and 1999/2000. Over the same period, UK-domiciled students gained around 600 additional doctorates each year in biological and related subjects.”.

Cut

and paste

is

not plagiarism if you

include the material

in quotes and provide a reference.

BUT too

many direct quotations can make your material difficult to

read.

Example 4

The Roberts Report (Roberts, 2002) documents an 18% increase in the number of doctorates awarded across the UK in the five academic years from 1995/96. Most subject areas recorded this increase, with 600 additional doctorates each year in the life sciences and large increases in social science and law. The exception was in the physical sciences where the number of doctorates awarded fell by 9%.It is not plagiarism if you paraphrase the material, as long as you include a citation for the original source. The same applies to arguments/explanations (not just facts).Slide41

THESIS WORKSHOP

41Web sources and figuresIf referencing a web page you must include both the URL and ACCESS DATE since web pages are subject to change.keep a paper or e-copy of the material in case it is later deleted from the site.The web is a big source of plagiarised material (and easy to detect) so be particularly careful with web-based material.

Be careful with inserting figures from published material – you should either seek the permission of the copyright holder or adapt the figure and cite the source.

On both topics seek the advice of your supervisorSlide42

THESIS WORKSHOP

42Questions to considerWhat if I’ve already published a paper?Must say so in the thesis.What if I’ve worked as part of a team?Make clear which parts of the work or your own, and which were done by others.What if I get help with editing?Using other people's edits/suggestions is fine if the main text is yours and you acknowledge the help. If in doubt, discuss with your supervisor.Slide43

What is expected of your thesis?

How to get startedInteractions with your supervisorMotivation and mechanicsThe viva and afterwardTHESIS WORKSHOP43Slide44

What happens in a viva?

Varies a lot depending on examiners.Almost certainly not what you expect.Use it to your advantage.RELAX. Have a conversation. If you don’t understand, ask for clarification. Your goal: demonstrate your independence as a researcher and ability to defend your work.Get feedback for writing papers/future work.THESIS WORKSHOP44Slide45

Typical format for viva

Begin with general questions/comments.Proceed through thesis in order with more specific questions/comments.Only rarely are these questions “tests”, usually they reflect genuine uncertainty/curiosity.THESIS WORKSHOP45Slide46

Some questions to be prepared for

What is your thesis? (Main claim in 1-2 sentences).What is (are) your most important result(s)?What are the limitations of your approach?What would you do next?THESIS WORKSHOP46Slide47

Examiner’s report form (U of E)

Is the thesis an original work that makes a significant contribution to knowledge in or understanding of the field of study?  YES NO

Does the thesis contain material worthy of publication?

YES NO

Does the thesis demonstrate adequate knowledge of the field of study and relevant literature?

YES NO

Does the thesis show the exercise of critical

judgement

with regard to both the student’s work and that of other scholars in the same general field?

YES NO Is the presentation and style of the thesis satisfactory?   YES NO

Specific criticisms for transmission to the candidate:

THESIS WORKSHOP

47Slide48

Most common recommendations

(Reg a) Award Ph.D./doctorate. The candidate satisfies the requirements for the award of a doctorate degree as laid down in the Programmes of Study as appropriate and the degree ought accordingly to be awarded.(Reg b) Candidate to Carry Out Corrections. The candidate satisfies the requirements for the award of the degree except that editorial corrections are required or stated minor deficiencies in the thesis must be remedied. In the opinion of the examiners, the candidate will be able to remedy these shortfalls without further supervision and without undertaking any further original research. The amendments are to be completed within three months, and to be subject to certification by the internal examiner(s), and by the external examiner (where the examiner so requests), before the degree is awarded.

THESIS WORKSHOP

48Slide49

Major corrections

(takes time but usually ok in the end)(Reg c) Candidate to Rectify Major Deficiencies. The candidate is substantially deficient in one or more of the requirements for the degree, but appears capable of so revising the thesis as to satisfy them. The candidate ought therefore to be invited to resubmit the thesis in a substantially revised form along lines indicated by the examiners within a further period of study which should normally not exceed 12 months, but exceptionally shall not exceed 24 months.THESIS WORKSHOP49Slide50

Award of MPhil rather than PhD

Candidate is substantially deficient in requirements and appears not capable of revising the thesis to satisfy them, but may satisfy the requirements for MPhil.(Reg d) Award MPhil. (MPhil with no corrections.)(Reg e) Candidate to Aim for MPhil -- No Thesis Resubmission Is Necessary. (MPhil with minor corrections, 3 months.)(Reg f) Candidate to Aim for MPhil -- Corrections Needed but Thesis Resubmission Is Necessary. (MPhil with major corrections, 12 months.)

THESIS WORKSHOP

50Slide51

Fundamentally deficient

(Reg g) Award Masters by Research. The thesis and/or the candidate's defense of it in oral examination are so fundamentally deficient in respect of all or any of the requirements for the degree that the candidate ought not to be awarded the degree of PhD nor MPhil. However, the work is of sufficient quality to merit the award of Masters by Research.(Reg h) Fail. The thesis and/or the candidate's defense of it in oral examination are so fundamentally deficient in respect of all or any of the requirements for the degree that the candidate ought neither to be awarded the degree of Ph.D. nor a taught professional doctorate, nor to be invited to resubmit the thesis for any degree of the university.

THESIS WORKSHOP

51Slide52

THESIS WORKSHOP

52

University of Edinburgh Report of Examiners after an Oral Examination for the Degree of PhD by ResearchSlide53

THESIS WORKSHOP

53A final message….At the end of the examination process nearly all candidates obtain a PhD(mostly after completing corrections and modifications required by their examiners)Slide54

THESIS WORKSHOP

54