Target Users' Diagrammatic Reasoning of Domain-specific Terminology

In this paper, we investigate target users' diagrammatic reasoning in a controlled experiment, where participants were asked to search for information in a dual visualization comprising of a concept-oriented graphical (diagram) entry and a corresponding textual (article) entry. During the experiment, users' visual attention was recorded by means of eye-tracking technology. We chose professionals as participants and taxation as our exploratory domain. We show that diagrammatic reasoning is effective and improving on questions related to diagrams only (so-called D-questions). However, significantly longer response time was needed to produce correct answers to D-questions compared to the questions related to articles only (so-called A-questions) as well as questions related to both diagrams and articles (so-called DA-questions). Hence, diagrammatic reasoning of the D-questions is the least efficient compared to A-and DA-questions.