Pattern 18: Priming and interference ** [Abstract]
AKA:

Good designs should take account of the cognitive abilities of human users. How can we exploit what is known about human memory to improve out designs?

Therefore

Ensure that you understand all relevant psychological phenomena. Whenever possible take account of priming and interference effects when you design the site. Most visitors will not have the opportunity for rehearsal, so exploit transfer effects where possible. Use widely understood metaphors.

Now consider using STRUCTURED MENUS (19). FOLLOW STANDARDS (36) to exploit rehearsal and transfer effects and avoid interference effects. Prime the user using THE RHETORIC OF ARRIVAL AND DEPARTURE (20).

Contributors and sources
Frank Buschmann, Kevlin Henney, Nora Koch, Oliver Vogel, Uwe Zdun

According to current theories of cognitive psychology, long-term memory (LTM) stores knowledge and data. To be used these memories must be activated and this takes time. They are normally stored in an inactivated state though not necessarily in a separate location from those immediately accessible in what used to be called short-term memory. Working or activated memory is limited and transitory. Items in working memory can be accessed directly and quickly. The activation of the linkages to these memories decays over time. Miller (1956) discovered that activated memory could hold a maximum of between five and nine items or chunks before filling up in some sense. These limits depend on the meaning of the material and an expert’s chunk can represent far more knowledge than that of a novice. It is a crude mistake to interpret this as meaning, for example, that there should never be more than 7±2 items presented on a screen. For example, when a novice sees a display of a chess board he sees 32 pieces. A grand master, however, may see only five high level game patterns, which she has chunked through long practice. Further, the units are stored in classification, composition and other structures which assist recall. Rehearsal of cues helps storage in LTM and retrieval, through reinforcing commonly used activation paths in the brain’s neural network. Another way of activating these paths more quickly is priming where recalling one item helps to activate another semantically related one. Priming helps activate concepts in working memory (WM). When a user moves from one system’s interface to that of another there are transfer effects due to both rehearsal and priming. These transfer effects can be both positive (beneficial) and negative (harmful) from the point of view of usability. The more that user interface designers are aware of memory characteristics the better they can do their job.

Other memory effects that may be significant include interference, which occurs when priming may activate the wrong things or at least activate memories that interfere with what should be recalled. This is an example of a negative transfer effect. Positive transfer effects can exploit the ability of users to classify their knowledge and this helps with the consistency and coherence of an interface. Generally speaking, positive effects will occur when the designer copies the structure of existing and well-known tasks. The success of the desktop metaphor can be regarded as evidence for this proposition, as can the popularity of tricks such as 3D button controls that appear to depress as they are clicked. A classic of negative effects was the use of the F3 key for the help function in WordPerfect when most packages use F1. When a WordPerfect user presses F1 the wrong things are primed and when trying to recall what key gives help there is no such support. An example of the exploitation of positive transfer effects is the support for Lotus syntax in Excel.

Figure 18.1 Example: Priming memory of typefaces [refer to book]

Priming effects occur between semantically related words. For example, saying CAT primes DOG; showing ITALIC may prime bold. Another example of the beneficial exploitation of priming effects in a user interface occurs when the actual appearance of an item is used to reinforce the memory of what selecting that item may imply, often using a visual cue or an icon. The figure illustrates this with an example of two different menu designs for the typeface selection within a word processor. The rightmost menu clearly helps the user remember what typefaces like Courier and Times Roman look like. However, this can go too far. A user unfamiliar with the Greek alphabet might well be very confused by the Hellenic appearance of the Symbol face. Priming is a context-sensitive effect and designers should understand this. For example, CAT could prime X-RAY rather than DOG in the context of medical diagnosis.

Browse the language	What is Wu?	Look at an example pattern sequence	Structure of the patterns
Comment on Wu	Contributors	Return to TriReme home page	Links to related sites