Evaluating the Impact of User Characteristics and Different Layouts on an Interactive Visualization for Decision Making

Abstract

Abstract

Prediction of Users' Learning Curves for Adaptation while Using an Information Visualization

“…Gaze data has also been investigated to predict long-term user traits (e.g., perceptual speed, visual working memory, verbal working memory, locus of control), as well as task type and task completion time [13,17,41]. Past studies have shown that user characteristics themselves (e.g., cognitive abilities, personality traits) can predict how well a user will perform on, or prefer, a given information visualization system [13,17,23,41].…”

“…Some of the interactive functionalities available to support the decision process include inspecting the specific domain value of each attribute (e.g., the rent of home1 being equal to $500), sorting the alternatives with 1 Video demo: www.cs.ubc.ca/group/iui/VALUECHARTS respect to a specific attribute, swapping attribute columns, and resizing the width of an attribute's column to see how that would impact the decision outcome. For the ValueChart user study (fully described in [17]), 95 participants were recruited (ages 16 to 40) to perform 5 different types of visualization tasks, chosen from a set of low-level data analysis tasks defined by Amar et al [2]. These five tasks (shown in Table 1) require answering questions from different domains for preferential choice (i.e., rental homes, universities, cell phones, restaurants, and hotels) using functionalities of ValueChart (e.g., sorting, reordering, weighting attributes).…”

“…generate aggregate value, compare values) [2], as well as instances of the five Amar et al tasks themselves (e.g., compute derived value typically requires multiple preceding retrieve values). The number of conceptual operations in Table 1 reflects this additional layer of actions per task type, and gives a conceptual measure of complexity (see [17] for more information). Thus, the quantities and types of conceptual operations in a task, in conjunction with the current level of skill a user has with these operations, will impact task performance.…”

“…For purposes outside the scope of this paper (described in [17]), two ValueChart layouts were evaluated: vertical vs. horizontal. The 20 study tasks were first performed with one of the two layouts for one domain, and then repeated with the second layout and a different domain, with order fully counterbalanced across users.…”

“…There is evidence that user performance and satisfaction in working with visualizations can be influenced by individual differences such as cognitive abilities (e.g., perceptual speed, visual working memory, and verbal working memory [13,17,41]), or personality traits (e.g., locus of control [23]). Thus, researches have started investigating how information visualization tools could be made more effective and usable by enabling them to adapt in real-time to some of these user characteristics.…”