Presenting Your Work (Assignments/Dissertations)

Last Updated on by

Some presentation tips, based on my experience of the things which trouble students (and for which they lose marks) and the things which irked me in previous study-units’ presentations.

JP Presenting @ valletta.ai
Presenting is fun. But it is your responsibility to make it so for the audience as well.

So, in no particular order:

  • Always arrive at your presentation 20 mins early (at least)
  • Your slides and presentation must be clear; highlight what you did and why, and move at an energetic and brisk pace
  • You are doing the presentation to us (and not to your slide-pack); so look and talk to us when you present
  • Whatever you do, do not read from the slides (or the printout of your slides) to us!  Do not talk to the slides.  Do not read from notes either – it makes the presentation clunky.  Above everything else, always interact with the audience
  • Don’t patronize/have a condescending attitude
  • Assume zero knowledge.  There is nothing worse than being told “As you well know ” and the people in the audience having no idea what you are on about
  • When talking, avoid remarks like “etc. etc.“.  Be Specific
  • It is fine to answer a question with “I don’t know“, do not dig yourself in a hole by making things (you are unsure of) up.  Note, however, that if you reply each  and every one of our questions with “I don’t know“, that is probably a sign your presentation isn’t going terrible well…
  • Do not over-claim, just because there is a 0.5% performance difference between your method and someone else does not mean your method is better
  • Lemma of the above; you can robustly claim difference between methods if you have a statistical test which shows a difference
  • Slide numbers please. (“Can you go back on the slide with the table?“, and 70% of your slides contain tables)
  • Images/Graphs are better than words. Tables help organize results
  • When using text, don’t overpopulate your slides. Points, not sentences (no full-stops necessary)
  • Related to the above, avoid cluttered slides
  • Absolutely, no code
  • Static terminal output (typically the result of running your program) is  uninteresting to anyone but yourself
  • No boring/useless implementation details (e.g. “Data was saved in long.txt”)
  • Answer the questions we ask you, and not the questions you wanted to be asked
  • Slides, like your time, are limited – don’t waste any
  • Use all the real estate on the slide (e.g. if you have a few points and most of the slide is empty, enlarge the font to make it more readable)
  • This is a presentation, so no final slide(s) with references (if you get something for elsewhere write it down on the slide where you show it)
  • No spelling mistakes and/or inconsistencies in the slides.  Double check your work.  Carelessness is not tolerated
  • No need for numbers with 14 decimal places, three are typically enough in a scientific setting
  • Show energy, don’t present sitting down (unless you have a valid reason, e.g. a medical condition)
  • If you write it down on a slide – or mention it, it is examinable/assessable
  • Proper use of fonts (no comic sans), sizing (large!) and spacing
  • Practice the presentation in front of each  other/wife/husband/partner/friend/dog before the day.  It is true what they say, nothing replaces experience and experience comes through practice
  • Figures should be properly exhibited (titles, axes labels, no redundancy, maximize data-ink ratio, clear message, units, etc.).
  • No need for figure or table captions, you are expected explain what you are showing
  • Each slide must have a title
  • Make sure to use the correct technical jargon (and don’t make up your own)
  • If you are talking about group work, make sure to have a slide on the subdivision of your work
  • Just because you inserted an image of some text, that does not make it a figure
  • The presentation template (design) is important
  • If you show a mean (of multiple runs) you also need to show some measure of variability (e.g. standard deviation or standard error)
  • For a twenty minute slot, make sure to have not more than 11-15 slides.  The rule of thumb is not more than one slide per minute (less is fine).
  • Leave time for questions at the end (there always are)
  • Future/Further work cannot be marginal, and should be big picture (e.g. not changing a parameter or using some other programming language or technology)
  • At postgraduate study especially, we expect you to look at depth and detail.  For example, if you have reimplemented a method and your results do not match the original method you need to know why (if it is open source, look at it!).  You must have at least, a suggestion for the difference
  • Sometimes it is helpful to have a second, more-detailed slide deck in case you need to explain in more depth (during the examiners’ questions)
  • If presenting an evaluation, you need to compare your results with someone else’s – otherwise it is just material for the results section
  • Always take a printed copy of your write-up to your viva/presentation
  • Lastly and mostly – have fun; presenting is a great opportunity to improve your confidence and public speaking

More applicable to a study-unit assignment presentation (such as ICS5110, ICS5115):

  • Do not waste time on an introduction; we know the problems we set out for you (we wrote the assignment spec!).  The tasks you chose will become obvious during the presentation
  • Theoretically you have 15 mins + 5 mins questions, but this is more of a discussion (so sometimes we ask question right through the presentation if something is unclear).  Also, if your pitch is perfect (or abysmal) we may have no questions for you…
  • Time-keeping is strict and there are marks awarded on whether you are on time (or not). You will be stopped if your time runs out, too bad if you are still on slide #2…
  • Title slide must contain your name, degree, study-unit, email, date, (the new) um logo and the coolest visualization you built
  • A concluding slide with what you learnt at the end of the project is encouraged
  • No silly Thank-you slide at the end (or Questions?).  You can have this in your viva
  • You will probably (although not necessarily) have three sections – each section is somewhat similar in contents (off the top of my head; data processing; viz; evidence; stats; conclusions).  These could be your slide titles
  • Sometimes it is useful to tell the audience where you are at (progression)
  • The partial reason why we hold these sessions is for them to serve as training for your viva, so make the most out of it

Please remind me of anything I may have left out via the comments section below.

Post Scriptum: (Prof.) Gordon Pace got in touch after I published this and noted how time flies, but mistakes students make remain the same – perhaps due to lack of presentation opportunities and training.  He pointed me to a report he authored in 2004(!) which repeats many of the above points.

Learn from other people’s mistakes.

Leave a comment

Your email address will not be published. Required fields are marked *