Supporting and Exploiting Spatial Memory in User Interfaces

Abstract

Spatial memory is an important facet of human cognition – it allows users to learn the locations of items over timeand retrieve them with little effort. In human-computer interfaces, a strong knowledge of the spatial location ofcontrols can enable a user to interact fluidly and efficiently, without needing to visually search for relevant controls. Computer interfaces should therefore be designed to provide support for developing the user's spatial memory,and they should allow the user to exploit it for rapid interaction whenever possible. However, existing systems offervarying support for spatial memory. Many modern interfaces break the user's ability to remember spatial locations, bymoving or re-arranging items; others leave spatial memory underutilised, requiring slow sequences of mechanical actionsto select items rather than exploiting users' strong ability to index items and controls by their on-screen locations.The aim of this paper is to highlight the importance of designing for spatial memory in HCI. To do this, we examine theliterature using an abstract-to-concrete approach. First, we identify important psychological models that underpin ourunderstanding of spatial memory, and differentiate between navigation and object-location memory (with this reviewfocusing on the latter). We then summarise empirical results on spatial memory from both the psychology and HCIdomains, identifying a set of observable properties of spatial memory that can be used to inform design. Finally, weanalyse existing interfaces in the HCI literature that support or disrupt spatial memory, including space-multiplexeddisplays for command and navigation interfaces, different techniques for dealing with large spatial data sets, and theeffects of spatial distortion. We intend for this paper to be useful to user interface designers, as well as other HCIresearchers interested in spatial memory. Throughout the text, we therefore emphasise important design guidelinesderived from the work reviewed, as well as methodological issues and topics for future research.