What a fantastic conference: great topics, great venue, and great people. This report does not do justice to the presentations, which were packed with valuable information. Here, I give only a flavour of insights that I gained.
The STC Visual Communication Weekend was in Cambridge, UK, 23-24 June 2007.
The subtitle of this session was 'the return on investment of visual information design and usability testing'. Patrick Hofmann, in his inimitable way, showed us the value of using pictures and decreasing the number of words in our documentation. Although Patrick repeated some of the content from previous presentations, this session was a valuable contribution to the conference. For an example of cost savings, see 'The ROI of usability and information design' in Making cents of making sense: a review.
The use of pictures to create wordless instructions is not always possible. One example is in a troubleshooting environment, where we give people reasons and explanations for the choices that they must make. Usually, the more decision-making activity there is, the less successful pictures become.
Users' feedback on the images and the icons that we design is vital. Possibly, a customer does not want an information designer to speak directly with its customers. Therefore, what can we do? In the consumer market, it is possible that friends, colleagues, and relations use a company's products. Therefore, one option is to ask them for feedback.
Golf swings, juggling, and dance were some examples from José de Souza's fascinating insight into how movement is shown on a static medium.
Motion is shown in many contexts, for example, comics, art, and science. Historically, no guidelines were available for showing motion in technical documents. José gives a taxonomy as part of his doctoral research at the University of Reading (www.reading.ac.uk/typography/).
Some basic dimensions of motion are direction, displacement, and trajectory (that is, the path that an object takes). José exemplified them using tracks in the sand. Other dimensions such as time, force, and acceleration were not discussed in this introductory session.
Different ways of showing motion exist:
- Single significant moment. This shows the most important part of a movement. This type of illustration can be ambiguous, and possibly, symmetric movements such as juggling are not clear.
- Before, during, and after. As the name indicates, this is three images. Use this to illustrate the essential parts of a movement.
- Composite image. This is a series of images. Use this to illustrate the parts of a movement individually. An advantage is the flexible layout, for example, as in a cookbook. Disadvantages include the following things:
- A lack of clues about displacement
- Possibly, readers do not see the apparently subtle changes
- Possibly, the layout defies traditional reading order.
- Synoptic. This is a series of images that appears in one frame, unlike a composite image. Use this to show how the whole sequence fits together. An advantage is that displacement is clearly shown. A disadvantage is that sometimes, to show a sequence is difficult.
Different graphic devices to show motion exist. To show a future motion, use an arrow from (a representation of) an object. To show past motion, use an arrow towards an object. (Sometimes, readers misinterpret the arrow to mean 'look under the physical object'.)
Motion can be given more emphasis by using motion lines (loosely, arrows with shading). For example, when showing a trajectory, the further back in time a part of a line represents, the lighter the shading on the line.
In the first part of this session, Phylise Banner Klein gave an amusing account of her working life. High-power financial managers were usually not amused with her straightforward manner ("no, look at it this way"), but she lasted longer than her predecessors did, and she gained the respect of people in the financial industry.
The visual codes we possess are strong. For example, Phylise asked us to doodle—the constraint was to use only one curved line that the started and ended at the same point. We added a triangle on an outside edge, and a circle inside the doodle, and we ended up with a representation of a bird.
Phylise gave a taxonomy for representing information:
- Observation: "I looked, and this is what I saw." This is typically used in scientific analysis.
- Induction: "I looked, and this is what I thought."
- Methodology: "This is how I think it works."
- Classification: "Think of it this way."
When we visualise information, we need to collect the raw data, and then transform it. This compares to the scientific method, in which we show what we perceive. The visual designer's art is to remove information that does not add to the viewer's understanding.
Over the years, I have heard of Conrad Taylor, so it was a real treat to see and hear him in person. He started by explaining the functions of a line. What can a drawn line do? It can depict outlines (forms), inner lines, textures, and shading.
Some edges are real, for example, on a knife. Other edges are artefacts of the way in which we look at something, for example, the edge of an orange. External edges can become internal edges, for example, an external edge on a box viewed from one position becomes an internal edge (appears inside the borders of the box) when the box is rotated.
How does software do edge detection and edge enhancement? Conrad explained that where an image is composed of pixels (raster graphics), software identifies an edge by comparing the properties (for example, brightness) of adjacent pixels. If the brightness differs by some specified value, the pixels are identified as belonging to an edge. After an edge is identified, it can be enhanced. Light pixels are made lighter, and dark pixels are made darker. This fools us into perceiving a clearer picture.
This was not just another 'how to' presentation; Matthew Ellison (www.ellisonconsulting.com) explained the whys and wherefores of using screen captures in documentation.
Screen shots are good for showing and explaining the layout of screens, and for identifying the location of controls. Use a screen shot when it helps people, not because a style guide says, for example, that every page must contain a screen shot.
Screen shots can add to the cost of localization, because typically, new screen shots are necessary for each language into which a document is translated. One possible solution is to 'greek' (blur) the text using an image editor, so that the screen shots do not need to be localized. It is not a perfect solution, but usually, it is acceptable.
Different image file formats exist, and Matthew explained their relative merits. He also discussed some of the tools for taking screen captures.
Reading a form is different from using a form. Using forms that appear on websites as examples, Caroline Jarrett (www.effortmark.co.uk) demonstrated the problems that people can have with poorly designed forms.
A practical exercise demonstrated that people look in different places on a form when they have different tasks to complete. To help people complete tasks, the 'page furniture' (buttons, navigation, links) must be where users expect it to be, because when they swap tasks, they look at the page furniture.
Do the details of a form matter? Caroline explained that readers do not care whether colons are used at the end of field labels, provided that the labels are consistent in the form. However, for the case of the text, ISO 9241 part 17 gives definitive advice: use sentence case for labels (rather than using title caps or capitalising every first letter).
Early in his career, Philip Ball changed from observational drawing to informational drawing. Now he is a medical artist (www.maet.org.uk); his illustrations are stunning.
When Philip first started, he created all his illustrations using traditional methods. Now he usually draws with the help of a computer. Sometimes, he creates drawings manually; for some reason, vets prefer hand-drawn illustrations.
Different types of illustration serve different needs.
Black and white drawings are used to present information. Traditionally, they were created using a dip pen, which means that the artist can draw lines only by moving the pen towards himself or herself. The illustrations contain a code (cipher), for example, veins are drawn differently from arteries (stipple versus lines), and medical people understand these conventions.
Watercolours, which frequently are monochrome, are used to show operative sequences (for example, how to remove a kidney). Typically, the background is a halftone.
Photographs are not used as frequently as people think. For example, it is not possible to photograph the areas of a foot that touch the floor. However, photographs are used as a source from which to create illustrations. Similarly, video recordings of medical procedures frequently are the source for Philip's illustrations. He draws from the video, and discusses the drawings with the surgeon.
An advantage of drawn illustrations over photographs and videos is that the artist can focus on the important information. The artist does not show the background clutter. Also, the illustration is idealised, instead of being an exact representation.
Frequently, art editors are involved, but the medical artist and the medical practitioner usually interact directly.
An icebreaker exercise, which was based on a real situation, showed how difficult it can be to design a graphic that depicts an object or situation in the real world. How does one design a graphic that represents the use of a western toilet when the audience has never used such a device? It is not easy!
Problems with pictures include too much detail; lack of standards (in a given document); unsuitable re-use; lack of meaning or unintended meaning; symbols that are consistently misunderstood. Patrick explained how to solve these problems. Two parts of the presentation that I found specially useful were about flow charts and screen shots.
Sometimes, flow charts are badly designed. For example, possibly the start point and the end point are not clear. Patrick recommended that the primary process or procedure has a central path, and that error conditions, exceptions, and subsidiary paths branch from this.
Is the best layout for a flow chart from from top to bottom or from left to right? No language goes from the bottom of a page to the top, but some languages go from right to left, therefore, the best option is to create a flow chart that goes from top to bottom. The only exception is with a flow chart that expresses time, in which case a left to right flow is acceptable.
Two large problems with screen shots are that they are not action oriented and that they do not show what is important. Solutions include shading parts of the screen shot that are not applicable, blurring the text, and cropping the image. One method for showing location is to use a thumbnail (small image) together with the primary image.