Mirror, mirror on the wall, who has the best visualization of all?

What is a 'good' visualization, one which leads to desired insights? How can we evaluate the quality of a scientific visualization or compare two visualizations (or visualization systems) to each other? In the following, the importance of considering the 'visualization context' is stressed. It consists of the prior knowledge of the user; the aims of the user; the application domain; amount, structure and distribution of the data; and the available hardware and software. Then, six subqualities are identified: data resolution quality, semantic quality, mapping quality, image quality, presentation and interaction quality, and multi-user quality. The Qvis reference model defines a weight value C (i.e.,importance) and a quality value Q for each subquality. The Qvis is introduced as a compact, easy to perceive representation of the so-defined visualization quality. An example illustrates all concepts. The reference model and the graph can help to evaluate visualizations and thus to further improve the quality of scientific visualizations.