Every time I attend the MPSA / APSA conferences I am surprised by the level of scholarship discussed here. On other hand, the variance in the quality of the presentations themselves is equally surprising. Here are some random tips on both substance and format:

1. Do not  spend more than 2 slides or 3 minutes with your introductory / motivation / lit review slides. Your audience is specialized and they do not need you to repeat what “they already know”–they want to hear your empirical or theoretical contribution.  With only 15 minutes to deliver, why not cut to the chase? If a slide titled “My argument” or “my hypotheses” comes until minute 6 or so, it’s already too late: you probably lost my attention by making me think on everybody else’s findings.
2. For empirical papers, it’s better to start asap with your main hypotheses, perhaps discussing the extant literature around your specific research question. You can also mix the explanation of your data with some lit review (Y, X and Z variables matter because AA and BB say so, or X was relevant for sample period M but not for N).
3. For reasons that escape me, empirical papers avoid talking about summary statistics or data sources. How am I supposed to interpret your point estimates or marginal effects over  a baseline I am not aware of? Again, discussing summary stats is a great way of combining lit review with that great dataset of yours. 
4. Your estimation technique looks fantastic but can you provide your audience with a sensible interpretation of at least one or two key coefficients? If you cannot do this then you are just showing off. And by the way, do not insult our intelligence with bullets with variable names and funny up/down arrows or +/- signs: your expert audience wants to see those point estimates or marginal effects and its corresponding little stars…
5. Try to devote about half of your time to the actual exposition of YOUR findings. It’s much better to discuss the implications of your findings, spiced up with some further lines of research, than to focus on what others have said or on ideas that you cannot test with your current data.
6. Never say things like: “so these are our findings, but we really don’t know why!” (yes, it’s a true quote). If you have a puzzling finding, you better come up with one or two plausible interpretations. If you don’t have it, it means that you have not done your homework thinking theoretically about your topic or, even worse, you may have serious identification issues with your model or the validity/reliability of your variables.
7. Avoid dark backgrounds in your presentations. Especially avoid dark fonts over dark backgrounds: black fonts over blue background only look cool on your sharp monitor screen. When conference rooms have plenty of light, light backgrounds with dark fonts look much better. Have you noticed that journals, books, and kindles use black over white? Trust me, it works.
8. If you are sophisticated enough to estimate fancy models in stata or R, you should be able to put a readable table or figure in a powerpoint (you know, like the ones made by marketing freshmen).  A table with more than 6 columns or more than 10 rows it’s too much data for one slide. If the “ugly big regression table” is unavoidable, you should be able to put a little red square or somehow highlight the key coefficients on the table. Or follow that biguglytable by a slide that “zooms into” the estimates of interest. (This also applies to that fantastic R graph with barely readable axis labels.)
9. By the same token, you should be able to produce a pdf file instead of an unstable powerpoint file (mac and windows powerpoints are not always compatible, and pptx files may look weird in “free powerpoint viewers”). Moreover, please use the “full screen” option available in all systems (F5 in powerpoint, ctrl-L in acrobat reader). Again, if you have a Ph.D. but cannot use the most basic presentation software, it’s really a bad signal.
10. Train yourself to find the “page down / page up” on any keyboard to advance your presentation. If you can’t, spend $30 on an usb remote control for your presentations. Don’t be cheap: you want to make a living by conveying ideas and this is the one time you may have an actual audience! People have no idea how clumsy they look when they say “does anybody know how to move to the next slide or go full screen?”
11. Some people like to use lots of photographs in their presentations, especially in case studies (“and this is what a municipal building looks like in Tlacoquemecatl”). What are you trying to convey with them? Haven’t you noticed that journals do not publish photographs? There is a reason for that. Similarly, never paste long paragraphs in a slide, it indicates that you are too lazy to narrow down your fancy prose into a catchy bullet.
12. Besides having your presentation in an usb drive (no, not the one with 6 months of work and MPSA09 viruses), always email it to yourself, or bring your own laptop, and a printed handout of your presentation (for true emergencies). Year in and out, cheap hotel notebooks will freeze at random or fail to recognize your jump drive.
13. It’s ok to have a printed handout to rehearse in the elevator –or while other presenters talk– but please do not read it during your presentation: it shows a lack of confidence and/or preparation. Similarly, do not read from the screen. You spent years reading the literature and months preparing this paper, right? It should flow naturally in 15 minutes.