This paper focuses on effective design of graphs to communicate the message using quantitative information. It presents a step-by-step process to achieve this. The author states that business graphs often communicate poorly or sometimes even miscommunicate. The lack of basic graphicacy skills result in arbitrary and ineffective decisions of graph design.
The author first discusses very basic concepts like tables, graph and types of scales used in graphs. He points to the fact that numbers become useful when we understand the relationship between them. Then he explains the 7 types of relationships in quantitative business data. He supports each type with an example graph which helps us in understanding the best way to display it. The author also spends some time on representation of data in the graphs. He states the situations where each of the lines, points, bars or boxes can be useful for encoding the data. The author makes an important point that we should avoid use of 2D areas to encode the data as our ability to compare 2-D areas is not well developed.
Finally we arrive at the crux of the article where the author presents a 6 step process for graph selection and design. Again the author stresses on knowing the message communicated in the graph. Then come the later decisions of identifying the common relationships, choosing the means of encoding, axes, scale and other things. The author shows an interesting way of representing 3 variable data using 2D graphs.
Personally, I found this paper to be very informative which provides a systematic way of designing graphs from quantitative data. However, formulating the process of graph design like this, at times, may result in monotonous graphs and kill the innovation required and expected from a visualization.
0 comments:
Post a Comment