Intelligent Enterprise

Cones are further subdivided into three types, each of which detects a different range of the color spectrum: roughly blue, green, and red. The fovea is simply an area with an extremely dense collection of cones. As a result, light that shines on the fovea can be seen in extremely fine detail. We're capable of seeing up to 625 separate data points in a one-inch square area, such as a dense collection of dots in a scatter plot. Perception of visual stimuli detected by parts of the retina other than the fovea is much less detailed, but it's capable of simultaneously processing vast amounts of information throughout one's span of vision, ready to notice a point of interest that invites greater attention (for example, the peripheral approach of a speeding car), which then leads to a quick shift in one's gaze to that area of interest. Rods and cones translate what they detect into electrochemical signals and pass them on, through the optic nerve, to the brain where they can be processed. Our eyes sense visual stimuli, then our brains perceive that data, making sense of it.

FIGURE 1 - Diagram of the visual perception system. (Eye and brain graphics created by Keith Stevenson.)

Role and Limitations of Memory

Just like computers, our brains use various types of storage to hold information while it's being processed and, in some cases, to store it for later use. There are three fundamental types of memory in the brain: iconic, short-term, and long-term. Iconic memory is similar to the graphics buffer of a computer, for it briefly stores what the eyes see until it is either moved into short-term memory (also known as working memory) for conscious processing or is discarded as nonessential. Short-term memory is like random access memory (RAM) in a computer: readily accessible for high-speed processing but limited in capacity. Information that's deemed worthwhile for later use is moved from short-term memory into long-term memory where it's stored and indexed in one or more ways for future retrieval, just like on a permanent storage device of a computer (for example, a hard disk). Despite a common misconception, we don't retain memories of everything we experience in life. Long-term memory is limited in its capacity, but amazing in its flexibility. It maintains and constantly rearranges a complex network of links between memories in an effort to keep memories available and optimally useful.

FIGURE 2 - Example of a graph that exceeds the limits fo short-term memory.

Short-term memory is where the real work of sense-making is done. New data is passed in from the world through the senses and old data is swapped in from long-term memory, working much faster than the conscious speed of thought to help us make sense of the world. Given our extraordinary cognitive abilities, it's incredible that all of this is achieved using short-term memory that can only hold from three to seven chunks of data at a time. This limitation must be considered when designing data presentations.

Figure 2 exhibits a common problem in graph design: the meaning of the nine separate data sets — represented by the nine differently colored lines — can't be concurrently held in short-term memory. The readers are forced to shift attention back and forth between the legend and the lines of data to remind themselves over and over what each line represents. If you want someone to make sense of the graph as a whole, then you must limit the number of data components that encode distinct meanings to seven at most — and safer yet, to no more than five.

Fundamental Attributes of Sight

The rods and cones that populate the retina are tuned to detect a limited set of visual attributes, such as shape and color. When we perceive an object, that perception is constructed from a combination of these simple visual attributes. Even though an object as a whole might take some conscious effort to identify, the basic visual attributes that combine to make up that object are perceived without any conscious effort. Perception of these basic visual attributes is called "preattentive" processing, in contrast to the conscious part of perception, which is called "attentive" processing. Preattentive processing is extremely fast and broadband in that we can simultaneously perceive a large number of these basic visual attributes, called "preattentive attributes." Preattentive perception is done in parallel, but attentive processing is done serially and is, therefore, much slower. Here's an exercise that illustrates the difference between these two types of visual perception. Take a few seconds to count how many times the number "5" appears in Figure 3.

FIGURE 3 - Example of attentive processing.

The visual differences between the shapes of the various numbers that appear in Figure 3 (for instance, the difference in shape between a "3" and a "5") are too complex to process preattentively. To count all the 5's involves serial attentive processing. Now count the 5's in the same set of numbers in Figure 4.

FIGURE 4 - Example of preattentive processing.

This time perception was easy and immediate, because the 5's were encoded with a different preattentive visual attribute from the other numbers — in this case, a different color. Why is this important to note? Because if you want to visually encode information in a manner that can be perceived instantly and easily by your readers, you now know that you should visually encode the data using preattentive attributes, and if you want some of the data to stand out from the rest, you should encode it using different preattentive attributes.

Here's a list of the preattentive attributes that are of particular use in visual displays of data:

FIGURE 5 - Preattentive attributes of visual perception most applicable to data presentation.

These visual attributes aren't perceptually equal. Some are perceptually stronger than others. Some can be perceived quantitatively and can therefore be used to encode numeric values, and others can't. The graph in Figure 6 falsely assumes that quantitative values can be encoded as variations in hue.