I recently secured a NSF CAREER grant on my second attempt. During the preparation of my application, I had the privilege of learning from the successful proposals of colleagues and friends. Additionally, I received valuable feedback from mentors. Certain patterns emerged, and I thought it would be beneficial to distill and share them in the hopes that others may find them useful.

Disclaimer: My submission was to the Security and Trustworthy Cyberspace program, and I sought examples from peers who also applied to this program. While I believe much of what I learned is relevant to other programs within CISE, your experience may differ. It’s worth noting that my first submission did not succeed; I received a Low Competitive rating, although the critiques were constructive. In contrast, my second attempt received a high competitive rating accompanied by three excellent assessments.

Based on my research and submission experience, here’s what a successful SaTC CAREER proposal generally entails:

Introduction (~2 pages)

Clearly outline your objectives and demonstrate their significance and impact. Define the problem's scope, highlight key insights, and describe the proposed activities aimed at addressing the issue.

• Utilize statistics and figures to illustrate the problem and its potential impact (e.g., affected populations, economic implications). Forward-looking trends are often beneficial (e.g., year-over-year growth). Real-world scenarios are also helpful, particularly for systems-oriented proposals.
• I identified three primary framing strategies: (i) solving enduring issues with existing technology; (ii) tackling emerging challenges with new technologies; and (iii) addressing problems faced by underserved communities.
• Clearly reference the intellectual traditions or disciplines that underpin your work.
• Contextualize your CAREER proposal within the larger problem domain (e.g., identify three major issues, and explain how your proposal addresses the first).
• Include a clear vision statement for the proposed work.
• Identify specific challenges in current methods or technologies that hinder your vision (especially relevant if your work pertains to new technology).
• List your research questions and/or proposed activities—this can include impact statements for each question or activity as well as associated outputs and deliverables.
• Present a table or figure summarizing your proposed work and its improvements over existing work.
• Provide a Kairos statement explaining the importance of pursuing this work now.

Background & Related Work (~1–2 pages)

Demonstrate thorough research, showing that your proposal is innovative while building on previous work.

• Define key terms from existing literature.
• Use subsections to cover each area of previous work that you intend to build upon, avoiding redundancy with the introduction.
• Conclude each subsection with a statement of contribution and/or integration.

Integrated Ethos Statements (~1–1.5 pages)

Illustrate that you have taken steps to mitigate risks and show that you are well-suited to undertake the proposed work. This information can be woven throughout the document as appropriate, rather than being confined to a single section.

• Detail any preliminary work relevant to the proposal, particularly if it informs the proposed activities. The more closely related the better, especially if there are risky assumptions in your proposal.
• Highlight previous experience with the methodologies you plan to use. Reviewers often prioritize how well your skills, demonstrated through past work, align with your proposed activities.
• Showcase your experience with related problem domains and technologies that you intend to utilize.
• Point out resources that may provide you with a competitive edge (e.g., existing high-cost equipment).

Plan of Work (~6–8 pages)

Expand on the commitments made in your introduction, detailing what you plan to accomplish and how. This section need not be singular; many opt to structure their work plan similarly to the overview in the introduction (e.g., each research question or proposed activity can be a subsection).

• Explain how your proposed activities address your research questions. Sometimes, research questions are first introduced here rather than in the introduction.
• The plan is typically divided into various thrusts, each representing a top-level contribution.
• Not every research thrust needs to be equally detailed; some can be described more broadly, especially if you expect adjustments during the proposal's timeline.

For proposed systems work: - Different levels of detail can be effective. You might include in-depth technical specifications (like system diagrams and code snippets) or broader descriptions of your intended approach. - Example usage scenarios can provide clarity. - For contributions involving "old solutions applied to new problems," discuss the original context of the solution and its relevance to the new issue. - For "new solutions to old problems," outline the previous problem and explain the necessity for a novel approach or your unique insight. - Consider exploring trade-offs between various new systems or methods if you are uncertain about which may be most effective. - A technical evaluation plan outlining and justifying relevant evaluation metrics is essential.

For user studies: - Detail your participant recruitment strategies, especially for specialized groups. - Discuss potential analytical methodologies (e.g., iterative open coding).

For security-related work, outline your threat model. If one model applies to all proposed activities, it should be articulated in a top-level subsection; otherwise, address it for each system or activity.

If your research plan comprises multiple thrusts, clarify the rationale for pursuing them in a particular order, ideally ensuring that later thrusts build upon earlier ones.

Educational Activities and Integration (~1 page)

Discuss your dedication to enhancing education for students both within and beyond your institution. These activities should align with your plan of work, forming a cohesive effort rather than standing independently.

Educational initiatives can be presented in a separate section or integrated within each research thrust. Common activities include:

• Curriculum development, which may involve creating new courses or refining existing ones, including MOOCs for broader audiences.
• Writing a textbook or contributing chapters to one.
• Developing educational tools (e.g., software aimed at teaching specific topics).
• Conducting workshops to share research findings with local organizations and schools.
• Hosting academic workshops that unite researchers.
• Mentoring students and providing guidance.

Differentiate these from "broader impacts," as some activities may overlap.

Intellectual Merits (~0.5–0.75 pages)

Articulate how your proposed work significantly advances the state-of-the-art in your field. This section typically includes a list of intellectual contributions emphasizing novelty, such as:

• Establishing the groundwork for a new academic discipline.
• Releasing software/hardware artifacts and systems to the public.
• Providing guidelines or design space recommendations.
• Enhancing emerging technologies for practical use.

Broader Impacts (~0.5 pages)

Describe how your work will benefit society beyond academia. Common strategies include:

• Partnerships with industry.
• Effects on specific communities of practice (e.g., developers, policymakers).
• Positive contributions to marginalized groups.
• Community outreach, including K-12 initiatives.
• Advising students, particularly from underrepresented backgrounds.
• Disseminating results through non-academic channels (blogs, social media, press).
• Leading workshops on the topic at academic conferences.

Ensure this section is distinct from educational activities to prevent redundancy.

Timeline & Project Management (~0.5–0.75 pages)

Demonstrate effective project management skills by outlining how you plan to meet the outlined vision. It is usually better to promise slightly more than too little, as predicting specific activities five years into the future can be challenging.

• You might discuss personnel, such as graduate research assistants, and their roles.
• Cover both proposed research and educational initiatives.
• Present the timeline in a list or table format. Two common table structures are: (i) rows for years with two columns for educational and research goals or (ii) rows for proposed activities with columns for years. The former is more space-efficient if activities align neatly with yearly milestones, while the latter is more comprehensive.

Prior Support from NSF (~0.25–0.5 pages)

Demonstrate productivity with any prior NSF awards you have received, which speaks for itself.

Letters of Collaboration

Provide evidence of discussions with collaborators who can assist in realizing your broader vision. Most successful proposals I reviewed had at least one letter of collaboration.

• These can come from organizational partners or other PIs with whom you plan to engage in activities. If your work involves difficult-to-access locations (e.g., hospitals) or populations, securing a letter from such entities is advisable.
• It is beneficial to specify what resources or contributions your collaborators will provide, ideally without demanding exclusive ownership of intellectual property.
• Adhere to standard templates for these letters to avoid rejection without review.

# Reflections on My First (Unsuccessful) Submission (written 2021)

I drafted this post shortly after my initial submission in the summer of 2020. Although I did not receive the CAREER grant that time, the reviews provided insightful feedback.

Personal Reputation Matters:

As this is a single PI proposal, your personal credentials significantly impact the outcome. Reviewers often noted my strengths and weaknesses. For instance, one reviewer commented, “while the PI has an excellent record in X and Y, his knowledge of Z may limit his ability to execute this research.” This observation underscores how central the PI is to the proposal. Therefore, the CAREER proposal may not be the best venue for introducing entirely new areas of research. Leverage your established reputation and focus on your known strengths.

If you do choose to diverge from your expertise, consider obtaining complementary letters of collaboration and possibly including a budgeted senior personnel member. While it might be feasible to compensate for a lack of established reputation in a specific sub-discipline, it’s unlikely you’ll receive the benefit of the doubt as you would for proposals closely aligned with your known strengths.

Scoping is Challenging:

Given that this is a five-year award (longer than most, which may tempt one to propose more) and a single PI proposal (which typically invites less scope), determining the appropriate scope can be tricky. This challenge is exacerbated if you have prior experience with multi-PI proposals, as your scope expectations may be misaligned for the CAREER.

My initial submission was overly ambitious. I usually write proposals for 2–4 years, leading me to propose even more for the CAREER. However, proposals of that duration typically involve co-PIs who contribute additional work. A CAREER proposal should probably propose about the same amount of work as a typical NSF Small award (usually 2–3 years). Keep in mind that CAREER proposals necessitate significant commitments to educational integration, which will also consume time. Don’t be misled by the five-year duration!

# Reflections on My Second (Successful) Submission (written 2022)