NSF CAREER PROPOSAL Experiential Perspective by Sanjukta Bhanja Presented in NSF CISE Workshop at Arizona State University Many Resources: www.clarku.edu/offices/research/pdfs/NSFProposalWritingTips.pdf.
Presentation on theme: "NSF CAREER PROPOSAL Experiential Perspective by Sanjukta Bhanja Presented in NSF CISE Workshop at Arizona State University Many Resources: www.clarku.edu/offices/research/pdfs/NSFProposalWritingTips.pdf."— Presentation transcript:
NSF CAREER PROPOSAL Experiential Perspective by Sanjukta Bhanja Presented in NSF CISE Workshop at Arizona State University Many Resources: Google hits: 1,300,000
My Datapoint Third time Lucky Between 1 st and 2 nd trial, significant changes made 2 nd and 3 rd : Almost no change During the first trial, I taught 1-1 courses/semesters During second and third, I taught 2-2 courses/semester
First Proposal: Self Assessment Cross-layer modeling targeting switching and error Too close to my PhD work (already funded by NSF) Too focused on one problem on research Teaching plans were established
Between First and Second Meeting with Dr. Basu (my division program director) His feedback (CAREER proposal needs to focus on multiple problems in a general direction) Needs to launch a career as opposed to solution to one research problem Already started working on both CMOS and beyond-CMOS devices Created better publication record
Second Attempt Focused on CMOS logic Trade-off between reliability-error and power Used Learning Automata and Bayesian Network Added Thermal errors Also started working on Quantum Cellular Automata Quantum-aware error-reliability and power models Found a few external mentors in my area
Third Attempt Publication records improved beyond second. Almost every component of near-term objectives preliminary data was published. Letters from industry and Academia Two of my peer read the proposal and suggested a few non-technical but critical flow changes Reduce some of the dimensions to cover more depth in each dimension Pictures became much better
General Guidelines With large grains of salt (subjective) Six Blind Men and the Elephant (Illistration from Pawyi Lee, northeastern Thailand)
Research Planning (1) Develop track-record in multiple research thrusts Publishing some of the preliminary research is important Data made available in the proposal This is key for providing confidence to a panelist not working exactly in your area should have potential to launch a career much beyond 5 years Tasks should be closely coupled
Research Planning (2) Have balance between a few near term and far term goals It should not read as condensed version of three/four proposals. Should have right balance of fundamentals and significance Has to have a challenging non-trivial component (non- translational) Balance between risk (ambitious) and reasonability (feasible)
Educational Research (Planning) Education : program-sensitive Developing a new course is OK (too common) but Preferable Integrating research into education Propose and document a successful teaching in your classes Working on an instrument to measure some of the well-known techniques might be great Create a portfolio (teaching) Helpful also for mid-tenure and tenure
Outreach (Planning) Find local resources available for broadening participation, visiting K-12 teachers classroom in GA teach-in Get involved if you can with local 4 year college faculties Participate in SLOAN, McKnight conference Essentially plan on getting a track-record Some letters supporting existing work might be helpful
Networking Target a few conferences where you send papers, attend and network every year. Find a few external mentors; could be CAREER award winners in your division in recent years Visit your program manager Receive Feedback from peer
Writing Do not let a panelist surf for contribution: Particularly important for Intellectual merit, research objectives, task sets and educational goals Clearly state the significance and intellectual merit Many proposals have preliminary results in research but no track-record towards the educational component and broader impact.
Focus on Scientific Merit NSF panelists are against any tone that sounds translational A fundamental scientific theme has to emerge with huge significance. Clearly differentiate between your proposal and existing state of the art Enough support from preliminary data (your own work)
Excitement Writing should capture excitement/ enthusiasm Avoid a laundry-list of tasks coupled with each other Make a case as soon as possible Do not introduce a key excitement at 5th page 5th, 6th, 7th and 8th are in general reserved for fundamental theoretical techniques but the importance of using that amount of rigor needs to be justified in page 1 or 2 and if possible in summary
Support Letters Getting support letters would be critical Excitement shown by a peer and willingness to collaborate would help Chairs letter: An opportunity to get release time for your research (important for those who somehow had less negotiations during hiring) Some common practices in support letter are extremely obvious to panelists
Breadth vs Depth Career proposal should launch a career beyond the five years. So CAREER Proposal would have to be broader than regular proposal. Depth vs Breadth Tricky due to page budget One of my mentors feedback: You have a very strong 20 page proposal but an extremely condensed 15 pages
Pause After writing the first complete draft, give a break of at least seven days. Read and edit again Pause Read and Edit again This is particularly true of those of you that gets upset reading their own paper after the paper appeared in print.
Help Get a technical writer to proof-read Get a trusted peer to read Ask if they understood the tremendous significance of your research. I asked help from my My ex-dean to provide some feedback since he was in NSF earlier.
Untrue Assumptions All the details would be read and understood One need to propose 5 innovations (for 5 years) and then 5 more innovations for long term goals If each innovation is discussed in 0.5 page ->this makes all the innovations look less rigorous Guidelines are fixed between trials Detailed guidelines in formatting can be ignored
Final draft Pay special attention to figures and table captions. Usually first perception is formed after reading the summary and skimming over the proposal looking at figures and tables. Have a good overview picture and great caption that details generic problem statement. Large percentage of effort needs to be on the Summary Get someone to judge the proposal by reading the summary and skimming through the tables and figures
Second and Third time After each failure, take an appointment with your program officer to capture panel feedback beyond panel summary Remember that panelists change, so take feedback of one trial with a pinch of salt but address issues that are relevant. Definitely get feedback from a few colleagues (almost mandatory these days for multiple attempts)
Mentors Have mentors outside your University Mentors inside the department are great Can support your case for less teaching and service load You need mentors in your own area These mentors can help as sounding board and can provide important suggestions for the proposal
Participate in a Panel Effort as a PI and effort as a panelist are not exactly equal Migration of proposals through HC-C-LC-DNC This is also an unique opportunity that your competitors are fighting with each other to get you money.