A Knowledge and Data-Driven Integrated Measure of Display Clutter with Attention Allocation Factors

Advances in technology have led to the widespread use of information displays in many safety-critical environments, including nuclear power plant control centers, aviation cockpits, and automobiles. Operator information overload can have disastrous effects in safety-critical tasks. Therefore, it is imperative that information displays be designed such that key information is presented to operators without exceeding available attentional resources. Related to this, displays should be designed to reduce display clutter, which is defined as an unintended effect of display imagery that obscures or confuses other information or that may not be relevant to the task at hand. In order to reduce display clutter, there is a need to reliably measure clutter. Many measures have been tested which take into account a variety of knowledge-driven or data-driven factors, but none have been shown to have substantial correlation with task performance. Given this shortcoming of existing measurement approaches, there is a need to develop a new measure that reliably quantifies display clutter and that is substantially correlated with task performance.;The objective of the present research was to develop a novel measure of display clutter accounting for operator task knowledge factors, display characteristics, and attention allocation that correlates with task performance. Furthermore, given the lack of research on display clutter in the driving domain, another objective was to assess driver performance and attention allocation when using high clutter and low clutter in-vehicle information displays for navigation purposes.;The study followed a three-phase approach. In the first phase of the research, navigation display features that contribute to clutter were identified to create low and high clutter display variations. These clutter variations were confirmed using the Edge Density clutter measure, a widely-used data-driven measure. Furthermore, navigation display areas of interest (AOIs) were proposed and navigation task queries were developed that required driver knowledge of each of the AOIs on the navigation display.;The second phase used the low and high clutter display feature combinations in a slideshow-based presentation to assess how driver attention allocation varies among levels of clutter under differing navigation goals. In the experiment, participants were asked to respond to queries representing basic navigation tasks while driving. Participants also rated levels of perceived clutter for each display and the attention allocation metrics were used to confirm the navigation display AOIs from Phase 1. Results indicated that higher clutter led to longer response times, a larger number of fixations, and higher subjective perceptions of clutter. The attention allocation measures also confirmed the hypothesized AOIs for use in the third phase of the study.;The third phase was a driving simulation study in which participants drove a challenging route while using low and high clutter navigation displays to achieve a destination. During the course of the drive, participants were asked to respond to queries for navigation information (similar to the Phase 2 experiment) as well as to provide context-dependent perceptions of clutter. Results revealed low and high clutter displays to be comparable in terms of driving performance and attention allocation, but the high clutter displays led to higher perceptions of clutter. The comprehensive measure of clutter was.