Artificial Intelligence and Machine Learning

On Site: Does Typography Affect Proposal Assessment?

  1. Article
  2. Author
  3. Tables

Experience from assisting in the review of 30 proposals to a major funding agency suggests that mundane aspects of proposal formatting have an effect on proposal assessment. Why do these apparent connections between mundane formatting and actual funding occur? Here are a few possibilities, any of which might apply to a given formatting heuristic:

  • Causality. A formatting decision may cause a proposal to be viewed more or less favorably, for example, by making an initial impression on reviewers and program directors.
  • Social conformity. Unstated formatting norms might signal group membership (or lack thereof) in the most fundable group of researchers.
  • Sophistication of the proposer. People with the experience and ability to write successful proposals may tend to make certain formatting choices differently from others so that, say, use of a relatively small font size might occur when the proposer simply has more to say. On this view, a formatting property might correlate with high proposal ratings, yet have nothing causal to do with them.
  • Aesthetics. An aesthetically pleasing submittal can’t hurt.

Publicizing the effects of formatting on proposal acceptance helps researchers make better formatting decisions, causing funding decisions based more on content and less on form. This is especially beneficial to inexperienced proposers, those who mistakenly favor the “wrong” formats, those who spend time and energy deciding format issues, and all who want to avoid the potential disadvantage of unfavorable proposal formats.

Here are the analyses of the 30 proposals and the formatting heuristics the analyses suggests:

  • Sans-serif headings. Given that the body of the proposal narrative is in a serif font, are sans-serif section headings desirable? Proposals rated as “Not Competitive” are relatively less likely to have sans-serif section headings. This suggests submitting proposals with section headings in a sans-serif font.
  • Subsection numbering. Proposals rated as “Not Competitive” are the only ones to make the apparent misstep of numbering subsections of the narrative as subsubsections of the overall proposal. This suggests numbering subsections of the narrative independently of the section structure of the proposal as a whole.
  • Font size. Proposal narratives submitted in 12-pt. text were disproportionately low in competitiveness, while the more competitive proposals tended to use smaller font sizes. (Exception: one noncompetitive proposal used 9-pt. text.) This suggests an advantage to submitting proposals with narratives in text smaller than 12 points.
  • Conclusion section. The more competitive proposals tended to have a conclusion (or a similarly named) section of the narrative. This is despite the lack of mention of a conclusion in the official guidelines for narrative content. This suggests including a conclusion section in proposal narratives.
  • Abstract in narrative. Every proposal rated as highly competitive had an abstract or other short introduction in the narrative. Because the overall proposal format required a separate section providing a summary of the proposal, some proposers might assume such an introductory section in the narrative is unnecessary. Proposers of the most competitive submissions, however, did not make that assumption (despite the lack of mention of this in the official guidelines for narrative content, a fact that might mislead some proposers to assume it is unnecessary or even to be avoided). This suggests submitting proposals with brief introductions.
  • Summary reused. Proposals included a summary section separate from the narrative section. A proposer might then reuse this section as part of the narrative. However, proposals that rated as “Highly Competitive” had the greatest tendency to avoid duplication between the summary section and the narrative section.
  • Early submission. The evidence suggests that the more competitive a proposal is, the more likely it is submitted early, possibly reflecting a lack of deleterious “last minute rush” compromises on quality.
  • Suggested outline. Evidence shows that following the suggested outline had no clear benefit. Indeed, the data suggests following the outline very closely (and therefore having no conclusion, for example) is, in fact, deleterious.
  • How proposers referred to themselves. Using the term “we” appears preferable to writing in the “I” or the third person. Over half the proposals ranked as “Highly Competitive” used “we,” compared to less than half of the proposals in the “I” and third-person categories.

The data and analysis here suggest formatting guidelines to use when submitting proposals. These results are more applicable to proposals in the computing field than in other areas because the proposals analyzed were computing proposals. As much or even more, the results indicate a need for more rigorous analyses using more data that will result in clear formatting guidelines. While the most successful proposers are most likely use the “right” formatting guidelines, many others do not. If they did, proposal review would be freed from the distractions and noise introduced by variations in form and independent of content—and content is what proposal review should be all about.


Back to Top

Back to Top


UT1 Table. Proposal Analysis

Back to top

Join the Discussion (0)

Become a Member or Sign In to Post a Comment

The Latest from CACM

Shape the Future of Computing

ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

Get Involved

Communications of the ACM (CACM) is now a fully Open Access publication.

By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.

Learn More