This from the folks at Research Trends:
…Research Trends decided to conduct its own case study of scholarly papers published in Cell between 2006 and 2010, and their citations within the same window…
comparing the citation rates of articles of different lengths revealed that papers with titles between 31 and 40 characters were cited the most…
the few papers with questions marks in their titles were clearly cited less…
but titles containing a comma or colon were cited more…
and only one (uncited) paper with an exclamation mark in its title.
Also, on the subject of humor…
An analysis of papers published in two psychology journals…found that “articles with highly amusing titles […] received fewer citations”, suggesting that academic authors should leave being funny to comedians.
Note both analyses are for manuscript citations. If the standard protocol when skimming journals (or having webots do it for you) is to look for keywords and phrases, this all makes a bit of sense.
However, if your grant proposal is sitting on a stack with another 19 or so, and you want to be one of the first read (and believe me, you don’t want to be the last read) your gambit may be to attract some attention. And the best way to do that is to stand out. In Advanced EEB this semester, each student proposed five titles for their NSF Pre-Doc, and we “market tested” each set by vote of hands. Most of the time, it was the short, jargon free, titles that posed a question or a challenge, often with a clever turn of phrase, that won out.
My working hypothesis:
Grant title: maximize the “intriguing” content;
Manuscript title: maximize the information content.
Survive and Thrive in Grad School 