- Never ever read a prepared text out loud. Speak from brief notes or a slide show. It is better to have five lines of summary text in ball-point pen on the back of your hand than a manuscript.
- Do you have reason to believe that a considerable part of your audience a) does not know anything about the subject, b) does not yet care about the subject? Then make sure that you start your talk by setting the scene with some background info (a few maps, a time line, a landscape view or two), and above all explain why you care. Why is your subject interesting?
- When presenting a scientific study, do not rattle off data and descriptions. Start with questions, explain how you came up with them, and then present only enough data (in overview format) that you can make a plausible argument in favour of some answer to each question. If you have no questions or no argument, do not agree to make a presentation.
This is good, solid advice, if you ask me. Podium presentations need to be somewhat dynamic in order to keep an audience interested, especially those comprised of people that are not necessarily there to hear you in the first place, as is true of too many general sessions at the SAAs, for instance. A derived side effect is that you're unlikely to get good feedback on a paper that's not clear, succinct and well presented. I don't know how I feel about a complete lack of reading... from my perspective, I think that writing a conference paper using colloquial language and reading from it (note: not reading it) can be a good alternative, especially for presenters still new to the practice or whose first language is not English. Ideally, a paper written that way can be rehearsed ahead of time, with the resulting presentation being relatively fluid, comparatively jargon-free and more engaging than a straight read of an academic paper.
For people who don't have a lot of experience presenting (and who have a lot of ideas or material they want to discuss), I actually think that writing out what you want to say is good practice. That's mainly because it takes about 2 minutes to read one page of double-spaced text at a pace that won't lose your audience. So if you have 15 minutes to present a paper, what you have to say should fit in 7.5 pages. Under the best of circumstances, you can then boil down that presentation into a series of key points tethered to specific graphs and figures.
That said, when it boils down to it, why would you choose a paper as the format to present your work? There's a stigma that somehow places papers 'above' posters in terms of prestige and effectiveness in the academic imagination. To a large degree, I think it derives from the fact that professional organizations still implicitly devalue the poster option when they ask questions like "would you be willing to present your work as a poster if there is no room in the regular presentation schedule" during the submission process; I've never seen the opposite option being offered (would you be willing to present your poster as a paper), which I think is pretty telling.
And that's really a shame because, in my experience, the one constant about these two formats is that you unquestionably get more feedback, questions and one-on-one interaction as a result of presenting a poster. From my perspective, the relative merits of posters and papers depend on the context: what are you talking about, what audience are you trying to reach, how comfortable do you feel speaking in public.
First, if you're presenting research that's mainly theoretical, then yes, a paper's probably the way to go. For just about anything else, I think a poster can be just as good - and in many cases, maybe even better - a venue as a poster. If a conference presentation is a condensed version of a research paper boiled down to its most essential three or four graphs, what do you think is the more effective way of presenting these data: on separate slides that the viewer can't go back to, or as a composite on a poster which allows the viewer to jump back and forth between them? In addition, since they're hard copies, posters have the virtue of being more amenable to tweaking to emphasize certain aspects or make a point. By that I mean that I've seen posters comprising all sorts of interesting data visualization strategies that are unavailable to slide and PowerPoint presentations. For instance, I distinctly remember one poster whose presenter overlaid sets of transparencies over the stratigraphic profiles that was the main figure on her poster in order to better show concentrations of artifacts and organic compounds at a site. And at last year's Paleo meetings, there was at least one poster presenter who had set up a laptop next to his poster on which he was running the simulation he had used in his study so he could walk people through the various steps of his analysis. All of that to say, that I think you have a lot more freedom in terms of how you present your work, which overall makes for better communication and more memorable work. Also, each poster means one less potentially horrid, clunky and poorly designed PowerPoint presentation which are unfortunately too common at meetings (and yes, I mean you, Dr. Lightredfontondarkredbackground... you know who you are and you owe me some new eyes)!
But to me, the real plus of giving a poster is the interaction and exposure it can give you. For one thing, you usually have 2-3 hours during which your poster is exposed, which in theory is that much longer than 15 minutes to have people come see it. In other words, posters are easier to fit in a schedule than individual talks... since talks are almost always off schedule for some reason, this means that even people who make it to the time your podium presentation's supposed to take place may not end up seeing it. In that light, posters are good because you're at the same spot for a long time, which makes it easier for people to come see it when they have a bit of time. Additionally, many people will end up strolling through poster sessions as a productive way to kill time between papers they want to try to catch, which means that a wide range of people will at the very least see your poster. And if you see someone whose opinion you're especially interested in, you can always ask them for it when you see them walk by. And you never know who will walk by, really, which means that the likelihood of making useful, unexpected contacts is always there (and for that reason, I usually try to have 11x17 versions of my poster to hand out, along with my contact info).
Which brings me to my last point, namely that posters actually give you the chance to field questions people may have about it, which is not always possible in paper sessions, where papers are often tightly squeezed, with little time for discussion of specific aspects of individual papers. A poster can give you the opportunity to get much better and direct feedback than a paper will. What is more, you also don't have to fret about engagingly reading your poster to an audience: people can either read through it themselves or ask you for a condensed summary, which is more than OK to give in colloquial language since you're actually interacting with someone as opposed to addressing an anonymous audience. So, especially if you're a first-time presenter or not especially comfortable at public speaking, that makes for a great, less nerve-wracking experience. And if you end up standing in front of a poster than no one stops to look at (which I've never seen happen, really), just ask yourself whether that's much worse than the alternative of actually having to read out your paper to an empty room... at least with a poster, you can leave early!