By Caitlin Kirby, Associate Director of Research & Interim EDLI co-Director
I was trained as a graduate student to embrace failure as an expected part of my professional career. That training embodied itself in the outfit I wore for my dissertation defense in 2019. I made a skirt out of a selection of rejection letters that I had received throughout my graduate studies. This idea ended up resonating strongly with people, and the picture still floats around on the internet occasionally. As with any creation that gets shared, interpretations will vary. Two threads emerged of how people thought of rejection based on their reactions: the display was either a reminder that my success was because of all my hard work and prior rejections, or in spite of all those who had failed to note my aptitude and had rejected me. I personally advocate for the latter interpretation, with an additional recognition that a healthy dose of luck plays a hand in getting accepted or rejected.
As I have progressed in my career and now write grants for the Evidence-Driven Learning Innovation (EDLI) team, my pile of rejections has, of course, grown. Our funded grants are easily and widely celebrated (such as our work in AI ethics and blended courses). However, I feel that our unfunded grants also deserve celebration, and simply listing these grants as “unfunded” in our tracking spreadsheet doesn’t give them the recognition they deserve. Many of these failed, rejected, and unfunded ideas represent important pieces of learning, growing, and shaping the ideas that have gone on to be successful projects. This post pays homage to a select few of our unfunded grants to acknowledge the value they provided, although it was not in dollars.
Good grant writing requires a strong understanding of the funder and their review process, a deep knowledge of the scholarly literature on the topic, procedural knowledge of grant administration following an award, a strong network of collaborators for the work, and the ability to write out a project clearly and very thoroughly. All of these get better with practice, requiring failures along the way. Here are some of our favorites:
Finding Footing: Blended Courses Unfunded Grants
When I first joined EDLI in my grant-writing role, I had to fail several times to find my footing with the grant writing process. Most of these early failures centered around efforts by instructors at MSU to offer hyflex and other flexible blended modalities for students, particularly in the wake of the COVID-19 pandemic. We submitted to the Spencer Foundation Large Education Research grant program in 2022 and 2023 for this project, and to the National Science Foundation Improving Undergraduate STEM Education (IUSE) program in 2022. Continued feedback and revisions on this project resulted in a successfully funded grant through the IUSE program when we submitted in 2023. The work and feedback on the three prior grants was foundational to our success, improving our collaborative team, writing skills, and overall project plan. These rejections resulted in $400k of funding for EDLI.
Improving our Understanding: Generative AI Data Analysis Unfunded Grant
Next on the chopping block was a 2023 Spencer Foundation Small Education Research Grant where we proposed exploring the capabilities of generative AI programs (like ChatGPT) in processing student midsemester feedback comments for instructors. There are several potential advantages in using generative AI for this purpose, such as being able to process large amounts of student feedback, filter out more harmful or emotional language from students, and receive pedagogically recommended action items based on feedback.
Writing this grant allowed us to dive into the literature on generative AI’s development, training, and limitations, which we have since integrated into our practice on midsemester evaluations, future grants, and our scholarly work across campus. Planning out the project helped us solidify an approach to human-and-AI coding, a process that Hala Sun uses in midsemester evaluations (for interested instructors) to improve efficiency and outputs. Hala and a collaborator also published on a similar process in a methods paper on how to utilize generative AI as a supplemental research partner in conducting qualitative content analysis. Writing the grant required us to thoroughly understand the background of generative AI’s development and uses in education, a foundation that has helped our team apply for five additional grants related to generative AI (several of which are also unfunded, some of which are pending). Finally, we integrated this work into our understanding of generative AI that we bring to our conversations and consultations with instructors in EDLI’s colleges, and publications like this one on building an approach to generative AI.
Connecting with Collaborators and Learning a New Agency: An AI in Education Grant
We have a grant that is still pending with the Institute of Education Sciences (IES) focusing on academic outcomes of generative AI tutoring programs (so while it’s not technically rejected, its status is currently “unfunded”). Writing this grant introduced us to two new processes: building out a process for inviting collaborators across institutions and learning the process for submitting to a new agency, IES.
This grant required collaboration across multiple institutions, and we identified several potential partners early in the process. However, there is a complicated dance of bureaucracy in proposing these multi-institutional partnerships, and we had two partners drop out relatively last-minute. These challenges gave us access to some potential partners to reach out to earlier in subsequent grant processes, and taught me just how far in advance of the deadline we need to be formalizing agreements with collaborators. EDLI is collaborating with multiple institutions on a current submission to IES, and I am glad to say that I am sending out official forms over one month from the deadline to better coordinate these crucial relationships.
Additionally, with most of our prior funding coming from the National Science Foundation, the IES process was new to our team. Additional project components, such as a cost analysis, and project documents, such as a structured abstract, are each completed with particular formats to various agencies. The work we did on this first grant to IES has set a foundation for writing to the particular requirements of this agency and others that might use similar language or formats.
Other Tips for Submitting Unfunded Grants
Unfunded grants, while always an initial disappointment, are worth much more than as a line item in a spreadsheet or an annual review, because of the hard work they represent that leads to grants that do get funded. I will end with a few tips for how others can improve their writing of unfunded grants, too:
- Start early and often: All of our EDLI post-doctoral scholars were PIs on small education research grants submitted to the Spencer Foundation this fall. We will celebrate the skills they built and projects they created regardless of the review outcomes. (At least one of these grants is already yielding benefits, with the research design that Min Zhuang and Richard Marks developed for supporting accessibility using Custom GPTs submitted as a presentation to MSU’s Center for Teaching and Learning Innovation spring conference.)
- Spend time as a grant reviewer: Joining review panels for granting agencies that you plan to apply to is incredibly helpful in learning some of the peculiarities of each agency, as well as understanding that the process isn’t personal and often does come down to luck, such as the expertise of your particular panel of reviewers.
- Celebrate (or at least recognize, if you can’t yet feel celebratory about a disappointment) the benefits of each unfunded opportunity.