Advice

Conference Season Special #3 – Enhancing Your Slides & Presenting Data

Visual Aids

Images are really useful for increasing understanding and making your work much more memorable, compare these slides:

Slide1

This is a weak slide because there is no visual evidence to help the audience remember what you are talking about

Slide2

This is a much stronger slide because it shows the audience the thing that you are talking about and shows a map of the presentation.

Evidence-assertion

First of all you can get rid of the generic title…RESULTS or CONCLUSIONS. Write a headline for the slide; just like a newspaper article – a short memorable sentence that succinctly summarises what the slide is about, you don’t let it get more than 2 lines; usually about 32 or 36 and

Then you put strong visual evidence beneath you headline, so the visual evidence could be a diagram, map or figure – so what we are doing is hooking a fact to a pertinent explanatory image – analysis of the recordings of eyeball movements when people are viewing webpages shows that we are more likely to read image captions than any other text on a webpage.. And you can use this for two purposes:

  1. Explanatory not descriptive. Tell the audience what this slide is about and what they need to take away from it. Audiences come in and out and they daydream; so if they miss a crucial sentence they can read what the point is.
  2. It forces you to focus the content on the one slide down to one main point; that doesn’t mean that there aren’t ancillary points that are going to come along that I am going to discuss; this is the one point. It stops you from clogging up the slide.

Examples:

 Slides3

Slides4

A longer description and more examples are available here:

http://writing.engr.psu.edu/slides.html

There is an interesting variant here:

http://scientific-presentations.com/2010/03/16/effective-variant-on-the-assertion-evidence-paradigm/

Colour

When we talk about colour we can describe it in terms of hue, and intensity, hue is the colour, intensity is the lightness or darkness of a colour.

Hue affects readability; some colour pairings make text vibrate and can cause eyestrain; this tends to happen when you use complimentary colours; opposite each other on the colour wheel.

VibratingColors

No the middle image isn’t a GIF! It’s static.

Appropriate Contrast

Contrast can also effect readability; if a background is too dark and a text too light it can make the text sparkle; this is true even when working in black and white, using a slightly lighter background colour like a dark grey can greatly improve readability. On the other hand if the contrast is too gentle it’s difficult to see the words; the trick is to find a good balance between light and dark.

In screen based media bright colours also affect readability; it is better to go for duller colour combinations to mute the contrast.

Black text on a white background is the default in word and other computer programs, because it has been the standard in print for a long time, but it causes eyestrain. A light grey background with dark grey text; is a much more sympathetic onscreen combination.

Venue brightness should dictate whether you go for a light or dark colour scheme

Dark venue = light background colour scheme

Bright venue = dark background colour scheme

Colour-blindness:

Be aware of colour-blindness, the most common form is protonopia, where the red cone is missing or not working, so it means that reds just kind of blend in with green; you can check your colour scheme for appropriate contrast online:

http://www.vischeck.com/vischeck/vischeckURL.php

Suggested colour-schemes

Dark Background with Light Text and Graphics Background – a dark blue (navy shade) or dark purple Text and Graphics – white or yellow Accent Colours – red, lime green, camel orange, light blue

Light Background with Dark Text and Graphics Background – warm beige Text and Graphics – dark blue, black, dark purple Accent Colours – dark green, burgundy
The trick with contrast is to make sure there is enough, but not so much that you produce eyestrain.

Slide typography

Use basic system fonts.

There may be times when you won’t be able to use your own machine to display your presentation file, so basic fonts such as Arial will help you to avoid system headaches when transferring between machines.

The following fonts are web-safe or screen safe fonts; they exist on the vast majority of computers and you can expect them to load correctly if you use them:

  • Arial
  • Impact
  • Comic Sans
  • Times New Roman
  • Courier
  • Trebuchet MS
  • Georgia                        
  • Verdana

Some of the fonts in this list (comic sans, courier and impact) have strong connotations and cannot be described as neutral, and so should not be used in slide design.

Use a maximum of 2 fonts

As a basic approach to typography in PowerPoint you can’t really go wrong if you stick to Arial or Verdana for both title and normal text. Both of these fonts are highly legible on screen because they have a generous x-height (that is the height of the lower case x).

Font size

Use a minimum of 24pts.

Check readability

To check readability get up from your desk and walk three metres from your screen.

Designing Data

Tables

Use a table when:

  • It is necessary for the audience to compare individual values
  • When the audience needs to see precise figures
  • When multiple units of measurement are used

Which of these tables is easier to read?

Table 1

France Germany UK
Number of professors 10,005 13,976 14,435
Population 65,700,000 81,890,000 63,230,000

Table 2

Country Professors Population
France 10,005 65,700,000
Germany 23,976 81,890,000
UK 14,435 63,230,000

 Table 2 is easier to read because:

  • The data that is for comparison is in columns not rows (if you use row for comparison values, as in table 1, the readers eye has to jump backwards and forwards to compare values).
  • The numbers are right aligned; it makes it much easier to compare the size of numbers if the digit columns line up.

What else can we do to improve the readability of Table 2?

Country Professors Population

(Millions)

France 10,005 65.7

Wherever possible you should remove digits and state the unit instead (as in third column), because additional digits do not make it easier for readers to understand the size of the number. More digits equates to more processing.

Row striping

PowerPoint helpfully does the row striping by default. Row striping is good for readability because it helps us to see the footprint of the table itself distinct from its containing slide and distinguish between rows.

Graphs

  • Probably the most efficient way to represent most aspects of data in the smallest amount of space.
  • They tell a story that it would take hundreds of words to tell.

American statistician and professor of political science Edward Tufte, In his book The Visual Display of Quantitative Information (1983), coined the term ‘chartjunk’ to mean anything that does not represent data or information about the data; lots of chartjunk like graph lines and unnecessary background panels is generated by the graphing software such as Excel.

A default graph produced in Excel:

graph1

graph2

The pink stuff is chartjunk; it doesn’t aid our understanding of the data. The audience is asked to untangle this information and determine for themselves what is meaningful here.

 The same information can be presented more simply:

graph3

Accepting the default chart that programs like Excel produce is questionable behaviour because when you do a presentation you will be judged on the work that you present not on the work that you have actually done. So you spend all this time creating data and actually doing research and then you choose to represent your findings by plonking an off-the-shelf graphic on a slide; it is really pedestrian and it makes it easier to ignore.

An upside to this situation is that if you put in a bit of thought and apply some of the principles of visual literacy detailed in this handbook you will stand out.

Key Point: Good graphics present their message simply

Tufte said: ‘Above all else show the data’ (1983)

Use colour

If you make all of the graph greyscale apart from the important bit you mute all the surrounding information that is not the main story. Desaturate all the supporting information so it moves from the background to the foreground; create a hierarchy with contrast. When all the parts are different colours the effect of any one colour is lost; if you limit the palette to black and white hues a burst of colour commands attention.

4

Another way to use colour is to vary the intensity of just one shade.

5

Be careful with visibility when you do this.

Use weight

6

You can experiment with making the important line thicker than all the other lines.  

Gridlines

Whenever a graph has gridlines we need to question whether they are really necessary. Historically gridlines have become part of the mental furniture of scientists and engineers because they simplify the task of plotting data by hand greatly. The original gridlines on grid paper were always printed faintly so they did not distract from the data curve. They were also printed in special green non-copying ink so that when the hand-drawn graph was photocopied the gridlines were not reproduced.

Avoid heavy grids and enclosing chart boxes.

Intrusive gridlines:

7

Assistive gridlines:

8

Tips for graphs and charts

  1. Mute the grid/labels/supporting elements – make the gridlines dotted and pale grey.
  2. Always provide as much scale information (but in muted form) as is needed.
  3. Fill the available space (remove other elements from the slide)
  4. Change the title to describe what it is about. “This graph shows…” Make use of confirmation bias.
  5. Change the remove excessive digits and state the unit, e.g. millions, instead.
  6. Don’t make people turn their heads. If there really isn’t enough space put the label text on a diagonal (rotate 45º) rather than horizontal (90º).
  7. Add annotations pointing out the interesting bit of the data.
  8. Aim to make 1 or 2 points with each chart or graph; just because you can put all the data series in the world in one graph it doesn’t mean that you should, sometimes two smaller graphs on one slide make it easier to understand a comparison.

Confirmation bias

Data slides need an effective title; it should be a text version of the point you are trying to make by showing the data.

As a slide designer you must accept that some people cannot or will not read the data, so you must interpret it for them. If you do this you will also be using confirmation bias to identify the pattern you want your audience to see. When told what to look for most people will not look for an alternative reading. It is similar to writing a headline for a newspaper article; you need to write a single clear assertion about the meaning of the data. Tell the audience what the data means (e.g. Silicon samples show a flat profile at low temperatures), not a description of what it is (e.g. the mechanical Q-factor of silicon samples).

In the final part of our Conference Season series we will look at ‘Handling Nerves and Audience Questions.’

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s