%0 Conference Paper %1 Moody2007 %A Moody, Daniel %B Advances in Information Systems Development %D 2007 %I Springer %K diagrams notation %P 481-492 %R 10.1007/978-0-387-70802-7_40 %T What Makes a Good Diagram? Improving the Cognitive Effectiveness of Diagrams in IS Development %X It is assumed without proof that diagrams are superior to text for communication. Diagrams are critical for IS development. People are not trained on how to draw good diagrams, how to layout and present diagrams. Some commonly accepted best practices actually hurt readability. No theoretical or empirical evidence for the syntax used in diagrams. We need to understand (1) the langauge of graphics and (2) human graphical information processing. 1. Bertin introduces visual variables: planar (vertical and horizontal placement) and retinal variables (shape, color, size, value, orientation, and texture). 2. Human graphical inofrmation processing involves perceptual processing (discrimination, configuration) and cognitive processing (working memory and long term memory). Principles for producing effective diagrams can be defined. 1. Discriminability: Absolute (size, contrast, proximity) and relative. 2. Modularity: most diagrams are too complex to understand. 3. Emphasis: most important things should stand out. 4. Cognitive integration: most systems represented by many diagrams. Perceptual integration: cues for navigation, and conceptual integration: forming a coherent mental representation. Techniques for cognitive integration: summary diagrams, nagivational map, signmposting, current context. 5. Perceptual directness: offload interpretation, iconic representations, perceptually direct relationships. 6. Structure: perceptual grouping. 7. Identification: external identification: correspondence between the model and the world; internal identification(graphical encoding) correspondence between conventions and meaning. 8. Visual expressiveness: the number of visual variables used to encode information. 9. Graphical simplicity: the number of conventions used. These principles can be applied to graphical languages or to instance models. They help to design better languages and to modify existing languages.

@inproceedings{Moody2007, abstract = {It is assumed without proof that diagrams are superior to text for communication. Diagrams are critical for IS development. People are not trained on how to draw good diagrams, how to layout and present diagrams. Some commonly accepted best practices actually hurt readability. No theoretical or empirical evidence for the syntax used in diagrams. We need to understand (1) the langauge of graphics and (2) human graphical information processing. 1. Bertin introduces visual variables: planar (vertical and horizontal placement) and retinal variables (shape, color, size, value, orientation, and texture). 2. Human graphical inofrmation processing involves perceptual processing (discrimination, configuration) and cognitive processing (working memory and long term memory). Principles for producing effective diagrams can be defined. 1. Discriminability: Absolute (size, contrast, proximity) and relative. 2. Modularity: most diagrams are too complex to understand. 3. Emphasis: most important things should stand out. 4. Cognitive integration: most systems represented by many diagrams. Perceptual integration: cues for navigation, and conceptual integration: forming a coherent mental representation. Techniques for cognitive integration: summary diagrams, nagivational map, signmposting, current context. 5. Perceptual directness: offload interpretation, iconic representations, perceptually direct relationships. 6. Structure: perceptual grouping. 7. Identification: external identification: correspondence between the model and the world; internal identification(graphical encoding) correspondence between conventions and meaning. 8. Visual expressiveness: the number of visual variables used to encode information. 9. Graphical simplicity: the number of conventions used. These principles can be applied to graphical languages or to instance models. They help to design better languages and to modify existing languages.}, added-at = {2010-11-22T18:29:37.000+0100}, author = {Moody, Daniel}, biburl = {https://www.bibsonomy.org/bibtex/245571a92b92c85844d4df27d21b59200/voj}, booktitle = {Advances in Information Systems Development}, doi = {10.1007/978-0-387-70802-7_40}, interhash = {edc2c34d7c6d8a589fb912cbb1d22837}, intrahash = {45571a92b92c85844d4df27d21b59200}, keywords = {diagrams notation}, pages = {481-492}, publisher = {Springer}, timestamp = {2010-11-22T18:29:37.000+0100}, title = {What Makes a Good Diagram? Improving the Cognitive Effectiveness of Diagrams in IS Development}, year = 2007 }