Preparing your PhD Proposal
Nancy M. Amato
What is a PhD Proposal really?
• A brainstorming and planning process • A succinct writeup of your proposed research goals, strategies, jus=fica=on, contribu=ons
• A good =me to get feedback and direc=on from experts
• Some=mes a stressful period of graduate school
Why write a PhD Proposal?
• An important, required step towards gradua=on!
• Focus your research direc=on – Wri=ng can clarify your thoughts
• Establish contact with commiFee • Obtain feedback from commiFee • Side effect: Proposal wri=ng experience
The Proposal Process in 1 slide • Select your commiFee with your advisor
– Get them on board – Get defense date
• Write proposal itera=ng with advisor: the proposal will be a contract with your commiFee
• Give proposal to commiFee • Prepare presenta=on of proposal
– Prac=ce several =mes – Get the =ming down
• Prepare for hard ques=ons – reread proposal as commiFee member • Present and defend proposal
– Reviseproposal based on feedback & formally submit
• Celebrate this big milestone! – Thank commiFee, meet with advisor (and commiFee members) to discuss ques=ons and
direc=ons
• Move forward on research!!
How long should a proposal be?
• Check your department & UniversiF rules – Texas A&M: at most 10 pages
• Some do 2 versions, one for commiFee and one for university
– Columbia: 30pp – no more – Others may be shorter
What is the role of the CommiFee? • Throughout the process …
– Guidance and understanding of what to expect – Feedback – Eventually, reference leFers
• At the proposal – Make sure you know what you’re talking about – Make sure you know the previous literature – Make sure you know the tough issues and have some answers
– Make sure your plan is reasonable
Selec=ng the PhD CommiFee
• Know your department & University rules • Ideally
– PhD advisor(s) – Faculty in the topic area to provide useful direc=on
– Faculty outside topic to provide high level feedback
– External member – someone in research area who could write a ref leFer
Proposal Presenta=on Tips • AFend others’ proposal defenses in your area • Thank commiFee, introduce yourself and background • Prac=ce presenta=on many =mes • Be polite during interrup=ons with ques=ons • Have someone take notes of ques=ons • Be open to sugges=ons • Prepare for ques=ons • Don’t be afraid to have no answer for a ques=on. Ask for
direc=on/help on those ques=ons. • Be confident. Don’t look to advisor for answers.
Challenges, Frustra=ons, Misconcep=ons
• “The proposal is just a hurdle. I can just propose ideas off the top of my head now and then figure out what I really want to do later.” – the lazy path
• “How can I propose something when I don’t know the details yet?” – the unknown path
• “I don’t know how to organize the different parts of the research on the page.” – writer’s block
• “I’m not ready yet. I might as well solve the problems and then present them.” – postpone, postpone, postpone….
What does a good Proposal Accomplish?
The proposal should clearly answer: • What is the problem you are studying? • Why is it important? • What results have you achieved so far and why to
they maFer? • How is this substan=ally different from prior work? • How will you systema=cally evaluate your work? • What do you need to do to complete your work? • What is your =meline?
When is the best =me to write/present a Proposal?
• When you’ve completed the rest of your requirements
• When you have a clear idea of the problem you want to solve
• When you have some preliminary work done to demonstrate promise of your approach
• When you have some no=on of the major subproblems to be solved
• When your advisor recommends
What to do before I start wri=ng?
• Think about what YOU want to accomplish • Write a succinct thesis statement/hypothesis
• Discuss your ideas with others • Present parts of the research at seminars, workshops, PhD workshops, conferences
• Think about 3-‐4 major contribu=ons/papers
• Formulate these contribu=ons in wri=ng
How do I get started wri=ng?
• Look at examples in your department, with same advisor, in your area
• Breadth of project proposed – Separa=on into subprojects – Wri=ng style: problem statement, hypothesis – Organiza=on
• Overall outline and flow • Within each proposed project sec=on
– How they mo=vated their topic and foreseen contribu=ons
Think about the Audience • Your CommiFee
– Not necessarily all in your general topic area – Not familiar with your specific problem – Not aware of your prior work – Not aware of your skills, infrastructure
• Implica=ons – Background: terminology, problem,… – State of the art related to your problem – Convincing mo=va=on for importance – Demonstra=on of feasibility/promise of success – Is a star=ng point for your disserta=on/thesis
How to organize a proposal? • Introduc=on
– Problem statement and importance
• Background and State of the Art • Proposed Research
– Subsec=ons on each research contribu=on • Evalua=on Plans • Research Plan • Summary of your Contribu=ons and Timeline
The Introduc=on
• General, high level problem for people outside area to appreciate
• Quick overview of what state of the art does not address
• Thesis statement – specific open problem and proposed strategy
• Brief overview of key insights and why your approach is promising
• List of your likely contribu=ons
Proposed Research
• Overview of project – maybe a figure • Specific project in steps • For each –
– Problem – Strategy – Details known now – Plans for remaining challenges – Evalua=on plan
Evalua=on Plans • Experiment Design:
– Ques=ons you will ask to judge success of your approach – Independent variables – what is being varied/compared
• Eg, your technique versus other techniques – Dependent variables and measures – what is being measured
• Effec=veness – precision and recall, f measure • Cost – efficiency
– Human subjects? – Methodology
• What ac=ons are you going to perform to conduct the experiment?
Research Plan
• What steps do you plan to take next? • What will you save for post-‐thesis work? Why?
Conclusion
• Summary of contribu=ons to the state of the art – intellectual merit
• Repe==on of broader impact on society
Sample Ques=ons at Proposal Defence
– Topic too large? too small? doable in the =meframe? Focused? What problems haven’t you foreseen? What happens if your planned experiments fail (backup plans)?
– Evalua=on (plan, sta=s=cs, validity) – Related work missing? – Prac=cality/scalability – Vision of where this can go… – Certain people have go-‐to ques=ons
• Watch your commiFee members on other proposal defenses • Especially if your thesis touches on work they have done or know a lot about
• Ask your advisor
Top Related