Coming soon: How to give a scientific presentation

[EvoDevo]

I'm working on a nice guide to giving effective science talks. Please feel free to throw in any suggestions you might have that I should include.

Cheers!

---

Members don't see this ad.

 

[Fixed Gear]

Ummm....

-Always introduce your abbreviations. For instance, if you are going to talk about Embryonic Stem Cells, let the audience know that 'ESC' is embyonic stem cell.

 

[Treg]

There seems to be a trend among seasoned speakers (i.e. PIs) to include some totally irrelevant pictures of every day stuff to demonstrate your point. i.e. star wars pics to talk about activated lymphocytes, etc. Has anyone else noticed this?

Treg

 

[EvoDevo]

There seems to be a trend among seasoned speakers (i.e. PIs) to include some totally irrelevant pictures of every day stuff to demonstrate your point. i.e. star wars pics to talk about activated lymphocytes, etc. Has anyone else noticed this?

Treg

It's not just the PIs.

 

[BDavis]

There seems to be a trend among seasoned speakers (i.e. PIs) to include some totally irrelevant pictures of every day stuff to demonstrate your point. i.e. star wars pics to talk about activated lymphocytes, etc. Has anyone else noticed this?

Treg

I noticed this especially with pediatricians; they like to show pictures of their kids/vacations. Nobel prize winners also like to show unusual pictures as well; I heard rumors that the guy who came up with PCR used to show pornographic pictures at some conferences, but I never saw it for myself...

 

[b&ierstiefel]

I've seen a good handful of what I consider excellent presentations. Here are some of the characteristics I found common in them (this perspective coming from one who has a crappy attention span):

(1) A thorough but focused introduction. No need to talk about the whole state of the field but just enough introductory concepts to make your data slides make sense.
(2) Simple slides - I've seen so many slides, especially data slides, that are so busy with too much stuff on it. Assuming the 1 minute per 1 slide rule, this slide is gonna fly by the eyes of the audience. If a very busy data slide comes early in the slide, the presenter risks losing most of the audience.
(3) Layering slides - this kinda expands on (2). What I mean by this is that IF you have a busy slide and you feel it is absolutely necessary to have all that data/text on that one slide, split the slides up into subslides. For instance, slide 1 has 20% of the data...slide 2 has 20% more added to slide 1...and on and on. Basically you are filling in pieces of the puzzle.
(4) Summarize often - after a set of few data slides, consider summarizing with text or a simplified model to really cement your main point in the mind of the audience.
(5) Don't speak too fast - I'm sure we've seen too many speakers do this. I do this especially when I'm nervous.

That's all I can think of for now...

 

[pseudoknot]

Don't have ANYTHING on a slide you don't explain or talk about. Don't make slides full of text and read them out loud. (Personally, I think slides should have little or no text, and I've observed that the more senior/competent people are the less text they have on their slides, but that's just in my experience.)

 

[Dodohead]

Don't make slides full of text and read them out loud.

Agreed with this one. Thats just annoying.

- Choose good colors for your presentation background/text. Be conscientious that some people are color blind, and can't distinguish reds/greens, etc. Yellow is a sucky color for graphs in general.

- Use tasteful animation in your powerpoint slides.

- Have an idea of your audience. In one of the worst sessions I ever attended at a conference, 5/5 of the speakers spent the first 5 min (of their 15 min talk) presenting the same background slide on leukocyte rolling. If you are speaker #5 in that situation, please skip that slide or gloss over it.

- Be prepared with extra slides showing info that you might not have presented, but might be asked about.

- Know the relevant literature. You will likely get questions, particularly, I have found, if you are presenting a poster, that have to do with some guys particular interest in the topic. Its helpful if you know what they are talking about beforehand.

Other suggestions about attending scientific conferences in general...

- Seek out PIs from other institutions that are doing similar work to you, and check out the posters from their students. It helps when submitting manuscripts. I lucked out in that about a month before my paper was submitted, I had spoken in detail about my research with both the editor of the journal we were submitting to, and also, it turned out, one of my reviewers. I think that helped my paper move along quicker.

 

[Nuel]

- Seek out PIs from other institutions that are doing similar work to you, and check out the posters from their students. It helps when submitting manuscripts. I lucked out in that about a month before my paper was submitted, I had spoken in detail about my research with both the editor of the journal we were submitting to, and also, it turned out, one of my reviewers. I think that helped my paper move along quicker.

Interesting!

 

[JETER]

An interesting and very effective method (if done correctly) is to draw the slides.

This is an example of an excellent scientific presentation:

The Oyster and the Quantum

 

[b&ierstiefel]

An interesting and very effective method (if done correctly) is to draw the slides.

This is an example of an excellent scientific presentation:

The Oyster and the Quantum

Brilliant!

 

[leverp2000]

Brilliant!

Hehe, I liked some of the titles: "Bohring Bayesians?" LOL.

Also, the history of the world was great.

 

[tedrik]

where the heck did this come from?!

Only in physics...awesome.

 

[JETER]

I came across this as a physics ug major. It is from Christopher Fuchs (PhD, UNM). Really interesting guy, with great research applications. Works for Bell labs now. This was one of those, "See, physicists can do interesting things" presentations. If quantum information theory is your cup of tea, check out these:

A. Furusawa, J. L. S?rensen, S. L. Braunstein, C. A. Fuchs, H. J. Kimble, and E. S. Polzik, ?Unconditional Quantum Teleportation,? Science 282(5389), 706?709 (1998). [This article was listed in Science as one of the top ten ?breakthroughs of the year? in 1998; see Science 282(5397), 2156?2161 (1998).]

C. A. Fuchs, N. Gisin, R. B. Gri?ths, C.-S. Niu, and A. Peres, ?Optimal Eavesdropping in Quantum Cryptography. I. Information Bound and Optimal Strategy,? Physical Review A 56(2), 1163?1172 (1997).

Interesting stuff. Unfortunately, not too many biomedical type applications, yet.