Top tips for writing successful proposals (NERC focused)
Robin Hogan, 29 October 2012 (see
What is a good idea for responsive mode?
The problem needs to be exciting but not too “out there” that it will put off any reviewers. It also needs to be mature enough that it is demonstrable that it will work (e.g. one paper published) without being widely known.
Preparation
1.Try to avoid responsive mode (success rate less than 20%); prefer Directed Programmes!
2.Start early: don’t underestimate the time it will take.Remember the University deadline is 2 weeks before NERC’s.
3.Ask Linda Tse to set up the Je-S form and do the costings.
4.Carefully read the guidelines (& the call for Directed Programme).
5.Read successful proposals: -> RED -> preparing applications.
6.Remember there are many other models than 1x unnamed 3-year postdoc, e.g. (i) 2x 2-year postdocs focusing on different parts such as observations and modelling; (ii) add a fraction of a senior postdoc’s time; (iii)name an experienced postdoc (Marguerite has a list of postdocs nearing the end of their contract); (iv) request a higher spinal point if it can be justified.
7.Large grants only: consider adding a PhD student in addition to a postdoc:only 20% of the cost of postdoc and can do more speculative science. The student can’t be working on a mission-critical part, but their work can nonetheless significantly add to the overall project aims.
Collaborations
8.Find a suitable Co-I to strengthen a project with multiple aspects, but avoid a rag-bag of multiple unrelated aims; the best proposals flow naturally from one key idea.
9.Hold a 2-hour workshop to brainstorm with colleagues.
10.If you have external Co-Is then beware! They will be working to a later deadline than you as most other Universities do not enforce an early internal deadline.
11.Approach project partners early, particularly the Met Office who are much more likely to contribute in-kind support if you tell them you are planning a proposal 12 months in advance.
12.Strong letters of support from project partners are important; consider proposing some text for a partner to use as a starting point, containing (i) praise of the innovative nature of the proposal and the track record of PI, (ii) an explicit statement of how the partner will contribute to the project, (iii) a financial value on the in-kind contribution, and (iv) a statement of how their organisation will benefit from the research.
Proposal mentors and other people to provide comments
13.Identify and talk to a mentor early; a duty of a Met academic is to be a proposal mentor if asked. Obviously sometimes people are genuinely too busy so ask Robin to nominate someone if you can’t identify anyone.
14.A good mentor has (i) knowledge of the specialist area, (ii) previous success with that funding body, and (iii) an ability to be brutally honest in a tactful way!
15.Typically there are two demands on a mentor: an early meeting to discuss the scope of proposal followed by a late draft for mentor on which to provide written comments.
16.Seek other comments too, even from busy people; just 10 minutes to skim through a proposal is good practice for the panel – can they get the main arguments in that time?
17.Robin and Charlotte can provide comments if you give them sufficient notice.
18.Take careful note of all comments! If someone was confused it is likely another will be too.
Evaluation criteria
19.Is the problem to be addressed well-formulated?
20.Is it important, with clear external beneficiaries?
21.Is it a research problem, or routine application of a known technique?
22.Do the proposers have a good idea on which to base their work?
23.Is it exciting and explained in enough detail to tell if it will work?It is absolutely not enough merely to identify a wish-list of desirable goals.
24.Does the proposal explain clearly what work will be done?
25.How would it be possible to judge whether the work was successful?
26.Do the proposers discuss previous work and its limitations?
27.Is the proposal cost-effective and ambitious, but realistic?
28.Is it simultaneously comprehensible to non-experts but contains technical details that are convincing to experts? Keep technical material in well-signposted sections; avoid in summary.
29.Is it interesting to read?
30.Is it easy to skim-read?
Style
31.Use the recommended style: 11pt Arial, 2cm margins (as this document). UseHelvetica in LaTeX: \renewcommand{\rmdefault}{phv}.
32.Right and left justify the text (as this document): it looks more professional.
33.Number sections and subsections in the main 8-page part. Use bold for section headings.
34.Proof read very carefully; use Word’s spelling and grammar checker, and a spell checker if you use LaTeX.
Layout of a NERC case for support
35.In a NERC standard grant, it is stipulated that of the 11 pages, the first 2 are your previous track record and the last one is the outline data management plan.Typically the penultimate page is used for management blurb and references, which leaves a 7-page “core”. It can be difficult to know how best to arrange the material in this part. I find it helpful to note that each paragraph in the core should fall into one of five categories: (1) motivation: why this work is important, (2) previous relevant work and its limitations, (3) new idea and/or key objectives, (4) what will actually be done, and (5) wider impact of the work. These parts obviously make most sense in the sequence shown, but many proposals mix them up in a way that can make it difficult for the reader to quickly get an overview of the whole thing. My recommendation is therefore that you start with a one-page executive summary containing all five categories in order, followed by the detail of the first four categories, in order (a detailed description of the impact of the work is contained in the separate “Pathways to Impact” document).
36.In this model, the sections of the central 8 pages could be as follows:
- Executive summary (1 page)
- Background (1-2 pages) – could be split into several sectionswith specific titles
- New idea (1-2 pages) – includes figure(s) and uses a specific title
- Work plan (3 pages) – split into subsections for work packages
- Management and collaboration (1/3 page)
- Staff development and training (1/3 page)
- References (1/3 page)
37.Depending on the individual case, consider these optional sections (each around 1/3 page):
- Risk assessment and mitigation strategy, e.g. a satellite dataset not being available
- Unsuitability for other sources of funding
- Participation in international research programmes
Previous track record
38.Put investigators first, not the institution: you are the stars!
39.Start Investigator track records with a few lines showing the date ranges during which you have had particular appointments; that way it is easy to compare your achievements to where you are at in your career.
40.Follow up by the number of papers you have written and total number of citations received.
41.Most of the track record should be a list of your research achievements, citing your own papers.
42.Use strong words and phrases such as “discovered”, “pioneered”, “for the first time” etc.
43.Make sure you focus on achievements relevant to this proposal, and say so.
44.Mention a few leadership roles in research projects and international activities.
45.Follow this by a “relevant recent references” section, but slim down if there are many investigators (and use 8pt font).
46.Finally the institution details; there is a standard paragraph about the Department of Meteorology, which can be followed by the specific strengths you will be drawing on, e.g. the strong modelling support provided by the presence of NCAS Models and Data.
Executive summary
47.Your panel will have a huge pile of proposals to get through and most on the panel will only read a few in detail. The rest they will just try to get the gist of. The purpose of the executive summary is therefore not that of an abstract, which is to entice people to read the whole document; rather it is to give people too lazy or busy to the gist of all aspects of the proposal.So it must be understandable to the non-expert.
48.You should lay out the five aspects of the proposal discussed earlier in order in separate paragraphs, although there is no need to use subtitles; the arguments from one should flow naturally to the next.
49.Use bullets or an enumerated list somewhere in the executive summary, for example to list your key objectives, key scientific questions to be tackled, or the main tasks to be performed.
50.Don’t try to explain everything in detail: you have the rest of the proposal to do this.
51.End with a flourish! The reader should get to the end of the page with the sense that this is an exciting project that would make a difference and it would be cool if it could be funded. One way is to say that this project is timely because of a new modelling or observational capability, to give the sense that funding this project now will help to keep UK science at the forefront.
Provocative prose
52.Throughout the proposal, do what you can to make your proposal easy to skim-read: use short sentences and italics to add stress important things like key questions, new ideas, main deliverables.
53.Make use of bullets and numbered lists where helpful; a good number of items is three.
54.Make sure the verbs in these lists are strong and varied. Try to avoid weak verbs such as: improve, study, understand, use, investigate, aim to, try to; prefer strong verbs such as: evaluate, tackle, design, generate, perform, derive, formulate, quantify. It is frequent for each item of a list to begin with “improve” which shouts INCREMENTAL RESEARCH to the reader!
55.Avoid acronyms: don’t confuse your reader just to save a line or two. Only use acronyms for very common ones, e.g. ECMWF. Never use “UKMO”: it is “Met Office”!
Background section(s)
56.Use a descriptive title, e.g.: “Background: The challenge of…”.
57.The first part elaborates on motivation and including human and economic factors (e.g. the floods in England in 2007 cost the economy over £3bn), not just scientific enquiry.
58.The next part reviews previous work, but should stay focused on the limitations of this previous work that will be addressed by the proposal.
59.A figure can quickly convey the nature of the problem (e.g. a cross-section through a sting jet). Don’t think about all the words that you have to exclude to fit in the figure; the reviewer or panel member will definitely look at the picture but will often not read the text. Nowadays, colour is fine in a proposal, but as in a paper don’t use it unnecessarily (e.g. for lines in a graph that could be distinguished by line style); consider the 8% of male reviewers who are colour-blind.
New idea section
60.Use a descriptive title, e.g.: “A new approach to…”
61.Is there a figure from preliminary work that shows the idea will work? Remember that if you have done enough preliminary work to get this far, you may need to clarify why funding is now needed (presuming that the preliminary work was done without funding).
Work plan
62.If necessary there may be a need for a more general methodology section first before getting to the nitty gritty of work packages.
63.Split the activities into work packages and use a descriptive title for work packages (so that they are easy to skim read).
64.A Gantt chart can be useful if you have more than 1 researcher or have a field experiment and the timing is important; otherwise it is probably enough to follow the title of the work package by the range of months in which it will be done.
65.Include as much detail as you can in the description of each work package.
66.Work packages must have deliverables that should be stated at the end of each work package in italics. There are many types of deliverables, for example
- “A paper on … to be submitted to …”
- “A new technique to … described in a paper to be submitted to …”
- “A dataset of … released to BADC”
- “A code / software library to … released under a free-software license”
- “A new parametrization of … in the Met Office model available to future users”
- “A web-based … targeted at …”
- “A prototype instrument …”
- “A COMET module to educate forecasters in …”(
67.Be ambitious about the number of papers: e.g. 2x (postdoc years – 1). Better to look ambitious than unambitious. State where the papers are going and possibly stress “open access”. Consider targeting some papers at a wider or different audience, e.g. BAMS or ERL. Associated with this you should request a substantial amount for page charges, e.g. 6k not 2k.
68.It is crucial that the reviewer gets the feeling not only that a lot will be delivered, but also that a variety of things will be delivered. Therefore deliver more than just papers. Otherwise there can be a sense that the only outcomes will be academic knowledge stored in technical papers that only a few will read.
Final blurb
69.Management and collaboration section should that mention weekly meetings will be held between the investigators and the PDRA(s) to review progress and strategy. It should also highlight how collaboration will be fostered with project partners.
70.Staff development and trainingsection should mention (i) access to training courses, (ii) opportunities for teaching and/or project supervision, (iii) annual staff development reviews by someone not their PI, (iv) weekly Departmental seminars and group meetings, (v) attendance at national and international conferences, and (vi) encouragement to engage in paper/proposal reviewing.
71.Referencessection: try to shrink this as much as possible, e.g. using 8-pt text, two columns, and not duplicating references in the track record. Consider removing titles if short of space.
Filling in the Je-S boxes
72.Use Word to check grammar/spelling or proofread very carefully. These often contain typos!
73.Don’t duplicatetext from your case for support; rather see it as an opportunity to make the case in a different way. All sections must be understandable to a non-expert.
74.Read the guidelines for each box in the Je-S help pages (buried within the Je-S system).
75.Objectives box: present more than three short objectives: use more of your 4000 characters!
76.Academic beneficiaries: think wider than just your narrow field.
77.Summary: note that this used to be headed “Make understandable to an intelligent 14 year old”.
78.Impact summary: this is non-academic impact only! Mention who will benefit and how, but not what you will do to ensure this happens: that’s for the Pathways to Impact document.
Pathways to impact
79.Many Pathways to Impact documents people write are very underwhelming. There is some evidence that these documents have been used to decide on funding at the threshold, so see this as an integral part of the proposal and spend some time getting it right.
80.Read Charlotte Johnson’s document of tips andthe NERC “good” and “bad” examples, and seek advice from Kathy Maskell (Walker Institute).Possible “users” of your research include:
81.Met Office, ECMWF etc.Activities could be giving a Met Office seminar, participating at or hosting workshops, or writing a COMET module for forecasters. To know who to contact at the Met Office you could talk to Peter Clark or a MetOffice@Reading manager.
82.Industry such as insurance, aviation, energy, instrument manufacturers etc.Participate in industry workshops, hold a stakeholder meeting in London, join a Knowledge Transfer Network. Talk toPier Luigi Vidale, David Brayshaw or Lizzie Froude.
83.Policymakers usually addressed via the IPCC process; also go on a NERC policy workshop.
84.The Public. Consider press releases, public lectures, school visits, NERC public engagement training andcreating web sites. Use the specific skills of your named postdocs.
85.Research staff benefit from career development, and this counts as impact so mention it.
Resources
86.Your Justification of Resources must state why something is needed.
87.Always ask for generous travel, computer, page charges; they are unlikely to be queried as they are much less than the salary costs. The worst outcome is that they are reduced, but the success of the proposal will be unaffected.
88.PI time is typically put at 10% and less for Co-Is, but consider asking for more (with justification); the University lumps this together as overhead, some of which comes back to your SDA as an “incentive” payment, and some to the Department which pays support staff.
Good luck! Don’t be discouraged if you are not successful first time!
1