Node-link diagrams are an effective and popular visualization approach for depicting hierarchical structures and for showing parent-child relationships. In this paper, we present the results of an eye tracking experiment investigating traditional, orthogonal, and radial node-link tree layouts as a piece of empirical basis for choosing between those layouts. Eye tracking was used to identify visual exploration behaviors of participants that were asked to solve a typical hierarchy exploration task by inspecting a static tree diagram: finding the least common ancestor of a given set of marked leaf nodes. To uncover exploration strategies, we examined fixation points, duration, and saccades of participants' gaze trajectories. For the non-radial diagrams, we additionally investigated the effect of diagram orientation by switching the position of the root node to each of the four main orientations. We also recorded and analyzed correctness of answers as well as completion times in addition to the eye movement data. We found out that traditional and orthogonal tree layouts significantly outperform radial tree layouts for the given task. Furthermore, by applying trajectory analysis techniques we uncovered that participants cross-checked their task solution more often in the radial than in the non-radial layouts.

We investigate the readability of node-link diagrams for directed graphs when using partially drawn links instead of showing each link explicitly in its full length. Providing the complete link information between related nodes in a graph can lead to visual clutter caused by many edge crossings. To reduce visual clutter, we draw only partial links. Then, the question arises if such diagrams are still readable, understandable, and interpretable. As a step toward answering this question, we conducted a controlled user experiment with 42 participants to uncover differences in accuracy and completion time for three different tasks: identifying the existence of a direct link, the existence of an indirect connection with one intermediate node, and the node with the largest number of outgoing edges. Furthermore, we compared tapered and traditional edge representations, three different graph sizes, and six different link lengths. In all configurations, the nodes of the graph were placed according to the force-directed layout by Fruchterman and Reingold. One result of this study is that the characteristics of completion times and error rates depend on the type of task. A general observation is that partially drawn links can lead to shorter task completion times, which occurs for nearly all graph sizes, tasks, and both tapered and traditional edge representations. In contrast, there is a tendency toward higher error rates for shorter links, which in fact is task-dependent.