Repairing a broken text slide and then converting it to a powerful visual

Let’s take a look at a hopeless text slide and repair it. We’ll do it in two phases. First we’ll turn it into a comprehensible text slide and then we’ll convert that slide into a powerful visual.

This slide has 70 words, well above the “40 word rule” recommended on this site. Initially it’s fine to put everything down in long form, but then the text should be edited down so it can be quickly read by the audience, while they simultaneously listen to the speaker.

So let's start with what the author intended…

The title indicates the slide is about the growth and centralization of the Cancer Clinical Trials (CCTO) staff. The first bullet says that in 2009 only 22 staff members reported through the CCTO, but that 50 staff were members of 13 independent departmental teams also conducting cancer clinical trials. The second bullet says that decisions were taken to grow the overall staff and to centralize all staff under the CCTO. The third bullet says that staff in the 13 departmental teams were converted to CCTO employees through attrition, rather than simply transferring them into the CCTO. But the 4th bullet says that staff remaining with their departments would be dual supervised by both CCTO and departmental managers. By 2015 attrition had reduced the number of departmental teams to three with a total of 11 staff. The CCTO meanwhile grew to 166 staff members.

Now let’s turn this into a text slide that can be easily read, and when combined with the speaker’s voice, comprehended.

This slide shows some changes we can make without losing content. There is no need to keep repeating “clinical trials staff,” or “staff.” The title takes care of that. Take off the adjectives, “small,” “total,” etc. Note "new policies" is two words shorter than what it replaced. Economize whenever possible. The slide should contain the core information, which can be embellished by the speaker’s voice.

So let's look at the repaired text slide:

We’ve cut the number of words from 70 to 37, still a bit clunky, but now it's possible for the audience to simultaneously read the slide and listen to the speaker (as he/she points to each bullet and explains what was intended).

Text slides are poor visuals, so when possible, we should always try to visualize text slides. Here is the above text slide in a more visual format.

A couple of things to point out. First, color is used to distinguish between CCTO staff and departmental staff. Color is not a decoration but a way to distinguish different kinds of information. The dual supervised staff are colored in a combination of red and blue. Circle areas are sized to reflect the number of staff, providing the audience with an instant recognition of the distribution. The issue of change by “attrition” is not contained on the visual, but that can be addressed by the speaker’s words. Not every last detail needs to go on the slide.

A to Z: Data gathering to presentation

The following figure gives a quick summary of the process from data gathering to presentation of results.

Experiments (data gathering research) in the laboratory generally produce mountains of data. A critical part of the presentation process (also part of the research process) is to take that mountain of data, figure out what is important and decisive, and understand the proper context for that data (context is symbolized by the layout of the pages on the grass). You cannot proceed to the presentation until you have figured out what small portion of your data is important and explanatory.

The key point is that a slide presentation is not a "data dump." It concisely summarizes the thematic lessons learned from the research.

Successful vs. Effective Presentations

The audience was getting restless. The job candidate had flipped through about 70-80 complicated slides in the first 45 minutes of their job talk. He then looked up and announced, "I'll have to speed up if I'm going to get through all my slides." Impressed by his thoroughness, the department made him an offer.

Increasingly, it is possible to achieve a degree of success in biomedical research despite an inability to effectively communicate scientific information. Audiences are increasingly satisfied with gaining only superficial understanding from a talk and massive volumes of data give the appearance of comprehensiveness. Leaders evaluating performance are themselves often creatures of the quantity over quality culture.

Talking faster to cope with time constraints may make the speaker feel good but it does nothing for the audience. Ditto for microscopic fonts or those slides with 8 or more graphs. Sadly, such speaking strategies are increasingly "adequate."

Three paradigm shifts are needed to move from successful to effective presentations:

1) Presentations must be designed for audience needs and they must be evaluated by what the audience learns, their increased ability to raise interesting questions, and to see the connections with their own research.

2) Presentations must be designed around a narrative showing. Only the data relevant to the narrative should be shown. We must move away from "data dump."

3) Presentations need to identify uncertainties, potential sources of error, and future strategies for closing these gaps.

An effective speaker will also be successful, but a successful speaker may not be effective in today's biomedical research culture.