How to write a strong abstract (for a conference)
(based on my own experiences, including three years as a track chair for the Association of Collegiate Schools of Planning (ACSP), reading abstracts and making decisions:

1. Have a clear research question.  Put it up front.  No need (or space) for a long wind-up.

2. Keep your narrative focused, with a clear set of steps -- and a convergence between question, topic, literature, methodology.

3. There is not much room (word count limit), so keep the writing tight.   Mention a few key writings to orient the project / place in context.

4. Do include citations.  Depending on the abstract submissions guidelines, they are either part of the abstract text or entered separately.   (My experience as a track chair: the lack of citations signaled that the abstract writer either didn't put enough effort into the abstract or didn’t know the literature/context).

5. Use active, assertive voice (no passive voice!).   And talk about what you do -- not what you hope/plan to do.  (e.g., "I use the cases of Portland, Seattle and Cleveland to compare transportation funding between urban and suburban districts to document the disparities in per capita spending by race and class"  rather than "I plan to examine three case studies across in the US to compare whether there is racial discrimination in transit.")   It takes a bit of confidence to talk actively about work you may not yet have done, but informed, measured confidence is key here.  Don't be vague or too speculative or rhetorical.

6. If you are converting a big dissertation proposal into a short ACSP abstract, it may be hard to edit down.  Instead, build from the ground up.  Start fresh, with a clean sheet/clean screen.   Copy and paste text into the abstract where absolutely central.  And then edit down to the most efficient, concise minimum.

7. Pick the right track:   track chairs can move things to other tracks, but that is a slow and not always successful process.  So pick wisely.  (though many papers can fit well into 2-3 tracks).

8. Avoid the mistake of simply describing the broader topic and why it is important for planning.  Do not have your abstract read like a literature review.  Be sure to assert your own voice and contribution, both your argument/outcome and your approach (methodology).   I am reminded of three elements identified in a short SSRC publication  (Adam Przeworski and Frank Salomon, On the Art of Writing Proposals, Social Science Research Council, 1995 rev., 1988, page 2): 
   “...every proposal reader constantly scans for clear answers to three questions:  What are we going to learn as the result of the proposed project that we do not know now?; Why is it worth knowing?; How will we know that the conclusions are valid?”  

9. Proofread before submitting.  Be sure all cited author names are correct.  Fix passive voice, typos, vague ideas, redundancies, dubious claims, verb tense disagreements, unnecessary qualifiers, etc.   And have others read your text before you submit.

Good luck!
Scott Campbell
PS happy to hear your own experiences and advice.  Please send a comment.