Skip to content

Latest commit

 

History

History
51 lines (31 loc) · 4.44 KB

craft-extended-abstracts.md

File metadata and controls

51 lines (31 loc) · 4.44 KB

(craft-extended-abstracts)=

Extended Abstracts

This chapter provides guidance on how to structure and write an extended abstract; in particular, what sections to have and what information to convey in them.

The advice is designed to be general good practice but it is unlikely to cover all cases and there may well be good reasons to ignore parts of it depending on the nature of any given project. And you should of course follow the instructions on length and structure as given in the call for extended abstracts that you're responding to. Furthermore, despite being phrased as "do this" or "don't do that", the points made below are designed to be sensible starting points–not rules that must be followed.

This chapter has drawn on numerous sources, including a thoughtful piece by William Pugh.

Much of what is relevant in an extended abstract is relevant for the whole paper too. So you should definitely read {ref}craft-writing-papers before reading this chapter. In particular, you should read the "Overall Contribution" section of {ref}craft-writing-papers as, with an extended abstract, you are similarly trying to convey to your reader what the marginal contribution is that you are making (or promising to make) relative to pre-existing knowledge and tools.

What is an extended abstract?

An extended abstract is not simply a long abstract. An extended abstract should contain references, comparisons to related work, any key equations, and other details that you might expect to see in the research paper but not in an abstract. Think of an extended abstract as a paper printed on a postage stamp.

There are two things that make extended abstracts tricky: first, you may not have written all of the paper yet; second, you need to tell a story in a relatively small number of words and writing concisely is hard. Don't expect this to be a quick or easy exercise—it will almost certainly be a useful exercise though.

The key to a good extended abstract is to omit anything that is not relevant to the core narrative of your work; you need to create a single, consistent story. You can also think of an extended abstract as one of those film trailers that gives away the whole story (but hopefully the reviewer will want to 'see' the full picture anyway!).

A typical length limit would be two pages, but you may find as many as six pages in the wild. Sometimes the limit is expressed in words or characters; typical limits might be 1000 words or 5000 characters.

A typical structure for an extended abstract might be:

  • Title
  • Introduction: provide an overview or background of the study, as well as a statement of the problem and the objectives of the study. Cite recent relevant literature where necessary.
  • Methodology: report the most important elements of the methods you're using to answer the problem. Can include equations, especially if they are key (for example, the main specification).
  • Results: reporting on key results only. Figures and tables can be included, but space will be limited so they must be core to the narrative.
  • Conclusion: this should cover any principles and generalisations that can be inferred from the results and any limitations of the work.
  • Acknowledgements: as in papers.
  • References: just those you need to support the extended abstract.

What not to put in an extended abstract

What should you not put in an extended abstract? Here's a list of what not to include:

  • future work
  • details of proofs or implementation that should seem plausible to reviewers
  • ramifications not relevant to the key ideas of the abstract
  • appendices

Tips for making your extended abstract strong

  • The value add needs to be clear and loud from the get go: is the work a significant advance over previous work?

  • If the work has a specialised application, provide the context and explain how this development will push forward the field more broadly.

  • Don't make it too long! Reviewers will notice if you try to fit 20 pages of material into the 2 page limit. It's a tough communications challenge, but one that will ultimately help you sell your work.

  • Don't treat the limit like a target, treat it as a limit. Everyone is busy; if you can convey the key points more concisely, reviewers will be grateful.

  • A reviewer is much more likely to be reading an extended abstract alongside many others at the same time, so an ideal submission will have caught their attention within a few minutes of reading.