StuartABerger.com

In a previous post, I wrote about Death by Powerpoint and how it arises when a talk (scientific or otherwise) becomes about the slides. A talk should always be primarily about what you say and the story you tell. Slides have a role to play. But that role should always be supporting, not starring.

Since slides do have a role to play in a talk, it is important to get them right. While a stellar slide might not elicit gasps of approval from the audience, a poorly designed slide can result in audience members despairing of absorbing its information and just giving up.

What are the prinicples of slide design? I mentioned some of them previously. Visually uncluttered and balanced are important. Simple is almost always better than complicated. Ideally, each slide should have one – and only one – main message. Data should be presented clearly, with easy to read labels (even from the back of the room) and be graspable in the time that the slide is shown. Headings should be used to clarify what is on the slide. And each section should have a conclusion explaining what you think is the main message.

A simple slide does not have to be simplistic. Rather, it should enhance and clarify the point you are trying to make. In our world, data and its presentation play a big role. When a presenter shows data, it should address the question or hypothesis posed and should advance the narrative that the speaker is conveying. Ideally, it should be obvious to the audience what the data show and mean. Context – the thought process behind the experiment, should always be provided. Going from one experiment to the next without a clear logical flow requires the audience to ‘fill in the blanks’. This exercise can be exhausting and fraught with confusion.

The most common forms of data presentation involve comparisons – control sample vs experimental sample(s), changes over time, the effect of an experimental manipulation. Most people can grasp multiple comparisons but complex comparisons amongst more than 4 samples are challenging to absorb in a short amount of time. I recommend either breaking the data into more manageable comparisons or redesigning its presentation format so it can be described and absorbed. Color has an important role to play as it can be used to distinguish amongst different groups. The key here is to be consistent in the use of color from slide to slide. Make it easier for the audience, not harder!

Biomedical research increasingly employs ‘omics technologies that generate very large data sets. These data require computationally intense methods and creative approaches to presenting them. I often see complex dot plots based on single cell expression data generated from sophisticated dimensional reduction techniques that contain ’clusters’ of cell types. The technology is remarkable and yet it is rare that presenters explain the analysis required to identify and validate the cell populations. When presenters display multiple samples on a single slide, it is almost impossible for the audience to grasp and absorb the content. I recommend that the presenter take time to explain the analysis, and then use simpler methods to present key differences in the samples.

Summary slides are also critical. At key points in the talk, particularly after a lot of data has been presented, taking a pause to summarize the findings allows everyone in the audience to get on the same page as the presenter. Conceptual models or diagrams can also be effective. I will write more about this topic in the near future.

Feel free to reach out to me either here or at my LinkedIn site if you would like help preparing an upcoming talk. And please like, repost or comment to spread the word.