Evaluation Needs for Visual Analytics

Abstract

Visual analytics present a more complex evaluation challenge than typical usability assessments for software applications. As seen in the nested blocks model, not only are there various parts to evaluate in visual analytics but also there are dependencies between these blocks. That is, changes made as a result of evaluation results in one block can affect the design and implementation of other blocks. Before discussing evaluation techniques for visual analytics, we will first review some of the recommendations made in the original visual analytics agenda (Thomas and Cook, 2005). The book recommends that research on evaluation methodologies should be a part of visual analytics research and development agenda to help focus the research efforts and facilitate the transition of the technology. Thomas and Cook cite the following benefits of evaluation.