How to Give a Talk

How to Give a Talk

S. Biller, Oxford

write a report and

Step -1: Make a Formal Outline!I. The Importance of Outlines 1) Forms a clear, coherent structure 2) See the “Big Picture” early on in the process. 3) Focuses where time should be spent in preparation. 4) Easy to re-organise so as to improve clarity.

II. Outline Formalism 1) Numbered sub-headings below each major section heading. a) Lettered sub-sub-headings go beneath these. b) Further nestings alternate between numbers and letters. 1. Each nesting should be further indented as well. 2. These are used to fill in more details as research progresses. 2) Multiple sub-headings stake out organisation of principle points.

III. Application of the Outline 1) Extremely useful for both talks and papers. 2) If formalised in sufficient detail, it makes the rest a dawdle.

Step 1:

General Motivation for Study

Overall Context of Result

SpecificDetails

Basic Structure to Guide the Reader:

1. Briefly, this is what it’s about (abstract)

2. This is why I want to do this (preface)

3. This is what I’m going to do (introduction)

4. Look, I’m doing it (body)

5. See, I’ve done it, just like I said! (summary)

6. ...and this is why I did it! (conclusions)

Another look at Structure:

A Few Golden Rules

Good grammarand correctpunctuation etc.lead to good communication(there really is a point to it!)

Unfortunately, although the answer was indeed clear, simple and straightforward,

there is some difficulty in justifiably assigning to it the fourth of the epithets you applied to the statement inasmuch as the

precise correlation between the information you communicated and the facts insofar as they can be determined and demonstrated is such as to cause epistemological problems of sufficient magnitude

as to lay upon the logical and semantic resources of the English language a heavier burden than they can

reasonably be expected to bear.

Epistemological? What are you talking about?

You told a lie.

Make simple, clear statements

Avoid jargon

Carefully define all terms

Whenever possible, distill from multiple sources.

When appropriate, paraphrase individual sources:As noted by Einstein, Eurovision scoring does not obey the normal laws of statistics.

If necessary, explicitly quote individual sources: As Einstein himself stated, “Who ist dis ‘Jedward’ anyvay?”

Common knowledge does not require a reference:Portugal will always vote for Spain; Belgium for France; Norway for Sweden...

(But, if in any doubt, quote one)

Neither Terry Wogan nor Graham Nortontake Eurovision very seriously as a competition.

“Is that STRICTLY true??”

(there are few absolutes in life!)

BeTransparentClearly state what you know.

Clearly state what you do not know.

Swap&

Proofread(Show no mercy!)

Preparing to Talk

• Preparation• Focus• Presentation

A simplified outline presented near the beginning of the talk can be a useful way to tell the audience where you are going:

Outline of Talk:

• Introduction• Middle bit• Conclusions

But not always…

Outline of Talk:

Focus:You cannot say everything and cannot show all thedetails of the work you’ve done… DON’T EVEN TRY!

Pick, at most, 2 or 3 important points to get acrossand concentrate on communicating them clearly.

Give yourself plenty of time by streamlining the presentation to this end.

My 3 points: 1) Make an outline2) Focus on just a few points3) Keep the style simple

Presentation

Simple transition slides can be a useful way to reinforce the structure of the talk

Slide Backgrounds

Might seem cool

(in a geeky kind of way)

Or just pleasant to have around

But are often distracting

And can make things difficult to read

Simple

isBetter

F onts:Use a consistent set of easy-to-read fonts.

Colour is useful to emphacise points and to visually break things up.

There are additional ways to do this as well,

but don’t go overboard.

And remember that if you try toemphacise everything that you haveto say, then nothing is actually emphacised at the end of the day and it completely defeats the whole object you are trying to accomplish.

It’s much better if you just try to emphacise a few well-chosen points.

For different backgrounds, this set can change.

Speaking of colours, just use a few well-chosen ones:

The basics of red, green and blue in addition to black provides a pretty good guide

Remember that not all colours show up well when projected, even if you can see them easily enough on your computer screen

You can put an awful lot on one slide by choosing small fonts and squeezing various different figures into different partsof the page. Some of these can even slide and zoom in, adding more and more to the page. You can also detail complex derivations. In fact, you could write down everything you are going to say and, if you like, even simply read it off the slide word for word. However, if you think about it, who honestly wants to sit through all of that? I mean, it would seem that either there’s no point listening to the speaker or no point trying to read the slide… or perhaps both! And how muchcan the audience really be expected to take in and genuinely comprehend, especially when they are supposed to belistening to you while reading the slide? Little information gets transferred by loading slides with too many details. It becomes impossible to see the forest for the trees and this can have a negative impact on the whole of your talk. A few slides like this and the audience will just give up and use the time to catch up on some sleep. Could you really blame them?

Tricks to help avoid clutter:

Use large fonts.

I generally try to use fonts no smaller than 24 pt,although I think 32pt is usually better.

Of course, you can emphacise things with even larger fonts.

And space things out.

This generally forces you not to put too much on one

page… which is a good thing!

Thus, (c Dt )2 = d2 + (Dx )2

d2 = (c Dt )2 - (Dx )2

invariant

or, more generally,

S2 = (c Dt )2 - [(Dx )2+ (Dy )2+ (Dz )2 ]

''Invariant Interval”

choose frame ''at rest”

= (c Dt)2

“Proper Time”

If you must derive something, one trick is to visually break up the steps:

0n: 1000 events peryear with 1% naturalNd-loaded liquidscintillator in SNO+

Test <mn> = 0.150 eV

maximum likelihood statistical test of the shape to extract 0n and 2n components…~240 units of Dc2 significance after only 1 year!

Klapdor-Kleingrothaus et al., Phys. Lett. B 586, 198, (2004)

simulation:one year of data

While you can display relevant graphs like this:

simulation:one year of data

It’s sometimes better to just talk to a picture on it’s own

To Avoid Clutter:

• Don’t put too much on page• Use large fonts• Break up derivations• Visually simplify• Use bullet points to summarise

More distractions to avoid:

There are many text animation and

SlideTransitionEffectsWhich

are ultimately distractingand incredibly annoying!!!

Simple

isBetter

Rehearsal

(“Simple” doesn’t mean “boring”)

Rehearse your talk!• At least initially, do this on your own

• Take all the time necessary to think carefully about what to say for each slide (don’t worry about details of talk length)

• Prune and iterate

• Run through final version at least 2-3 times

Speak to and engage with your audience…they are your friends!

Conclusions

• Make the transition to “Conclusions” clear• Try to keep summary to one page• Tie things together coherently• Reiterate the major points clearly

1) Make an outline2) Focus on just a few points3) Keep the style simple

• Briefly make any additional observations• Don’t be defensive about questions• Leave yourself plenty of time