An evaluation of the effects of transfer of practice between various computer menu configurations

The storage and retrieval of information is rapidly evolving as a major function performed by computer systems. One common approach for database information retrieval is the use of menu selection systems. Studies to date have focused on three major approaches towards improving menu retrieval performance: by changing the macro structure of the menu hierarchy, by changes to the micro structure of specific menu pages, and through the provision of training and practice.

The effects of experience have been readily demonstrated in menu selection tasks, where performance improves with extended practice on a particular menu hierarchy. Studies suggest that experience with a particular menu database enhances user performance, because the user becomes familiar with the location of selection alternatives on specific pages. However, recent evidence suggests that the general process of menu selection, independent of database familiarity, is also enhanced with practice.

The current study is concerned with the effects of practice on the general processing of menu items. If this General Processing Effect is independent of the menu's configuration, then the effects of practice should be similar between menu configurations of various breadths (i.e., number of alternatives per page). However, if this effect is related to the specifics of the menu page, then the transfer of practice (positive or negative) should vary between menus of different breadth.

Subjects worked through 2 series of 30 single-page menu searches. Menu pages contained either 2, 4, or 8 alternatives per page. Results suggest that a General Processing Effect exists, where menu selection performance is influenced by prior experience with menu selection. Both positive and negative transfer of practice was found, where prior experience with narrow menus (i.e., 2 alternatives per page) was found to be detrimental to menu selection from broader menus (i.e., 8 alternatives per page). It is hypothesized that the General Processing Effect is due to the use of different search strategies for menus of different breadths. Further research is suggested to validate these hypotheses.