From thesis text to slides. How to ease the process?

Question

I spent a considerable amount of time to write my thesis. Now I have to create the slides for the defense. I thought it was going to be easier, but found important difficulties creating them. I would like to have a method to make it easier, at least the first draft. Starting with a section, I will try to create a slide from one paragraph, although they can be further divided in more than one slide. I thought about:

• convert to bullets each sentence of the paragraph, or set of sentences of related topic. This slide may be further removed.
• in a different slide, replace text from previous slide with images from cited references, and try to memorise the necessary part of the replaced text. This will be specially useful for the introduction. On the other hand, thesis results sections already have their own pictures, so they will be directly taken from thesis.

Do you think this method is OK, or maybe someone has a better one?

Answer 1

You have to cut down brutally. You will have a half hour or so to show the work of a year or more, that is hard.

• Explain what this is all about. A short introduction/reminder (depending on your audience!) is probably needed.
• State your objectives clearly. One slide for this should be enough.
• Summarize your main results. You won't be able to cover everything, but make sure you show enough to cover the stated objectives. Perhaps highlight some tricky point, or a neat technique.
• Summarize your objectives again, explaining how they were met.
• If pertinent, point out possible future fruitful work building on your thesis
• It might be useful to have a slide with publications resulting from the thesis, both already published or in the works (submitted and waiting for approval).

You might consider having some "spare" slides with more details, on which you can fall back when asked to elaborate.

Answer 2

Note that the following approach may not work for every topic and field.

1. Find out what prerequisites you can expect from your audience or how generally understandable your talk shall be. For example a guideline at my department is that a defense shall be directed at somebody who studied the same field (but did not necessarily specialise on the same subfield).

2. Take your most important result. This may be a plot, a number, a model or a hypothesis, but also that some approach that you found not to work.

3. Consider the minimal background information that your audience needs to understand and appreciate this result.

   For example, if your result is a plot: Your audience needs to understand what quantities are depicted on the axis, how the data was obtained and what important steps you made processing it; it needs to understand why the plot was important to make and what major conclusion follows from it. You do not need to tell it about every detail though, e.g., that you rejected three samples because they were highly implausible (and checked thoroughly that this was appropriate).

4. Structure that information and make slides presenting it. This should contain an introduction and a conclusion.

5. Test how long you will need to present this talk. If you do not cram your slides (which you shouldn’t do anyway), a rough guideline is one minute per slide. Actually giving the talk in front of a test audience is a better estimator though.

6. In the rare event that you have time left, add additional details, interesting side results or whatever you seem best fit. Preferrably add something that needs little build-up, given the information that you already intend to present.

It is strongly advisable to talk to your supervisor and other peers, as they can better tell you what part of your work is most important, what guidelines and convention your field and university have and give feedback on a draft of your presentiation.

Do you think this method is OK […] ?

To be frank, I think this method is horrible for the following reasons:

• What is a good structure for a thesis is not necessarily a good structure for a talk. And even if it is, your method restricts how you distribute content on your slides in a way that you will find hard to get rid off, even through intensive refactoring.
• Not all information contained in your thesis needs to be contained in your talk. Moreover, it is pretty normal that certain paragraphs of your thesis are devoted to technical details (and should be only mentioned as an aside) while others deserve a whole slide.
• At no point do you have the top-level structure in mind – which is most important.

Answer 3

Two basic rules for presentations: Think about who you are talking to, and plan top down.

Look at the convention in your university and your discipline and the time constraints imposed by the rules to decide how much time you want to give the different strata in your audience. There is general public, general scientists, scientists from your general area and may be two or three who understand what you did. If your work touches different scientific areas, decide how you want to split your time between them. After assigning a certain amount of time to each part of the audience ask yourself what kind of talk you would enjoy in an area which was similarly far from you. History usually works for experts and layman alike, unless it is commonly known (to the experts).

Top down means start with the large parts, break these into smaller parts and refine until you arrive at what you want to put on every single slide, then write the slide. You will have to cut down, do so by working backward. Decide which goals you want to present and don't fret about unrelated stuff. You don't have to be too precise, the audience will not be able to process all details in the short time they have for a slide anyway. "Under some rather technical condition" or "after some manipulation" usually works here.

Answer 4

It might be easier for you to escape from the pages of your thesis if you storyboard your talk initially. In other words, start by choosing and inserting images, and then add text frugally.