Popular Science Presentations

[ajb]

I am thinking about giving a short presentation to an amateur astronomy association. They have asked what I do at work, I don't think I can explain much about that but I was thinking about giving an introduction to general relativity. For those that know something about me, will know that I study mathematical physics using geometry and general relativity, I though could be a reasonable way of introducing some "geometric concepts" to astronomers. I have a basic plan in my mind, but I will work on it before I give the talk.

 

 

As my audience has little mathematical sophistication, I was planning to present most of the ideas very intuitively and very pictorially, sacrificing mathematical rigour for ease of understanding.

 

 

So, what I wanted to ask is; who has given talks for a general audience before? What advice could you give?

 

Thanks

[ecoli]

use a lot of pictures and graphics in your presentation

[swansont]

(Much of this applies to technical talks as well)

 

Pictures and graphics are good, but they should not be cluttered/too densely packed with information.

 

Text should be big (20 point or larger), and standalone text should be a summary, not a transcript. The audience is there to listen to you, not read. Reading takes more time, and it's hard to read and listen at the same time, and get anything out of it.

 

One common mistake is too many slides. A good guideline is no more than one slide per minute of talk time.

 

Practice your talk, especially if you are lacking experience in giving them.

[timo]

Closest I've done was presenting my project in the "computational physics" course which was about GR simulations (the project where e.g. this picture comes from). T'was a 5-10 minute crash-course presenting the relevant formulas to 3rd-4th year physics students. That's probably not comparable, but I had the feeling that it was way too much for them, even though I only really needed the equation of motion and the Schwarzschild-metric (which I didn't even motivate but simply stated).

 

An idea would be consistently showing the concepts on a visualizable manifold, where 2D manifold in 3D are probably the best/only choice. E.g. you can nicely show the parallel-transport of vectors using a pencil and moving it around an apple (or an arrow on earth if you're doing it on a diagram). And blowing up a balloon if you're covering that, of course. You should probably not be afraid using stone-old analogies. If people saw them before that might actually help since they can then put them in context.

 

Nucleosynthesis or dark matter might be easier topics for that you can cheat a lot of understanding by showing Feynman diagrams. Especially modern experiments on dark matter searches might be interesting to hobby astronomers.

 

Keeping things as simple as possible of course is the default-advice. You hopefully do that in all your talks.

 

Would probably interesting if you'd post your talk here on sfn, afterwards.

[ajb]

I am not sure if I will use slides as there is a nice white board there! I also think that models could be useful, in particular to demonstrate that the sphere is indeed curved and the plane not by intuitively explaining parallel transport and curvature.

 

For sure, I see no point in putting up loads of text and I have seen that many times in research seminars! No-one reads them anyway.

 

I have given a few postgrad talks, but not one for a more general audience.

 

The idea would be to "explain" the field equations. By this I mean heuristically decode the equations and explain some geometry along the way. I thought

 

0)Quote the field equations

1) Manifolds + Metric as "space-time"

2) Vector fields and curvature (pencil and sphere job!)

3) Say something about matter, but not much.

[insane_alien]

because you are dumbing it down a bit(not to say the audience is dumb though) it will lose some of its accuracy but try to keep it close. telling them to ask for clarification if they don't understand something during the presentation. although, if you do that you need to be careful not to go off on a tangent.

 

as ecoli said, pictures are gold. use them wisely though. if they are too complex nobody will understand them and if they are too simple then they will not convey any information.

 

another thing i do is not to make a full plan like 5minutes on a 7 on b and then 3 on c.

 

just have a list of topics and/or you want to get through and wing it. it is much better and you don't have to cut yourself off.

 

a bit of humour in the presentation will also help, both you and the audiences attention. it doesn't have to be stand up comedian level stuff just a little gag or one liner related to the presentation. google will help with this. don't use too many though.

 

also, don't be afraid of putting some maths in. only, for the variables you will be better using words than the proper symbols. a lot of people have a mental block on algebra if it is like that. and if it is really complex take some time and describe the basics. most of them will get it.

[ajb]

The talk has been set for next week, Thursday 13th September.

 

A copy of the slides (may not quite be the finalised version) can be found here.

 

Any comments, suggestions etc are still welcomed.

[insane_alien]

had a quick glance at the slides, looks good. hope it goes well.

[timo]

Ok, first on-the-fly notes while reading through it:

Page 4:

- Capitalization of "Red shifting" looks strange. Since you are the native english speaker I'll skip propositions of how to write it.

- The word "Frequency" is misaligned

- I don't understand the first point: What is a "higher gravity well"? Perhaps rephrase that ("higher inside a well" or something like that?).

- I'd not reuse the term "lens" in point 3 or at least add additional information like "can bend light rays and thus effectively act like lenses".

 

Page 5:

- Typo: "Reimannian".

 

Page 6:

- Semicolon instead of ":" after "Examples".

- Do manifolds really necessarily locally look like Euclidean space? The local metric of spacetime surely isn't (it's riemanian, in appropriate coordinates).

- Perhaps add an emphasize "=> dimension of manifold" to the last point.

 

Page 8:

- First point misaligned.

- Perhaps make g red as in the text to emphasize or write "A metric g...".

 

Page 10:

- Usage of symbols for coordinates is inconsistent between text and picture ([math] \phi \leftrightarrow \varphi [/math]).

- Double-check the metric. It seems wrong to me that ds as a function of dtheta would be independent of phi. Note that the nomenclature in your picture is rather uncommon (at least to me); usually, theta is the angle to the z-axis.

 

Page 11:

- Same spelling error on Mr. Riemann, again.

- Typo: "Pesudo-"

 

Page 13:

- I don't know what you are going to say verbally, but the break from Page 11 and 12 to page 13 is tremendous. Effectively, Gravitational bending is excluded by a global coordinate system (t,x,y,z) with the global [math]g=\eta[/math].

- Very nice picture (the lower one).

 

Page 17:

- The statement "vector can be rotated as it moves along the loop" might be misleading: After all, you want to parallel-transport it (which is actually not rotating it). Of course, once you come back to the original point, the vector will have changed, but I'm not sure if your statement really catches it. Don't have a better proposal how to formulate it, though.

 

Page 19:

- I proposed this when writing a comment on page 17 and then deleted the proposal seeing you actually did so, already. I'd go 3 right angles chosing the vector parallel to the first direction of parallel transport, though. I think that's easier to visualize and self-test for the audience. You might want to bring an apple and a pencil for your talk .

 

Ok, that's from my first readthrough; haven't thought about the actual content, yet (not sure if I really want to, not sure if you'd really want me to ...)

 

EDIT: Btw, how long are you planning to talk?

[w=f[z]]

Ask the audience questions. Make them feel like participants instead of spectators.

 

Cheers

[ajb]

Atheist ,

 

thanks.

 

 

Typos are always hard to spot when you write something yourself, but obvious once they have been pointed out. You have also raised some points that I need to clear up, or really just be a little more careful.

 

As for the length I have 40 mins. I will indeed be using "props" to demonstrate parallel transfport and curvature in 2-d.

 

once again cheers