Dashboard Spy readers know that whenever Stephen Few, a leading expert in information visualization and dashboard data design, publishes his “Visual Business Intelligence Newsletter”, I’m right there waiting to soak in his precious gems of knowledge. His recent issue presents the 5 purposes for which we display quantitative information.
First, let me tell you where to download his latest issue, then I’ll tell you what he said.
Type of Quantitative Display (This is a direct pdf download.)
Few starts with his purpose – “to describe these types & provide basic guidelines for their effective design”.
We present quantitative information for various purposes. Each purpose requires that we design displays in particular ways to achieve relevant outcomes. What are these purposes? If we exclude quantitative displays that are used for mathematical purposes (e.g., a mathematical proof), we can identify five fairly distinct but at times interrelated purposes for which we display quantitative information:
- Exploratory Data Analysis
You should read the article for yourself (as well as subscribe to get future issues), but here is a quick snippet of explanation of each type of quantitative display:
Narrative displays are used to inform, explain, or persuade. Narrative (story) is an explanation. It answers questions: “How did we get to this state?” “What caused this to happen?” Its primary goal is communication, resulting in understanding and informed action.
Exploratory Data Analysis:
Exploratory data analysis (EDA) displays support the following two purposes:
- Data exploration to find facts of potential interest in a set of data
- Data sensemaking (a.k.a., data analysis or descriptive statistics) to determine what the facts mean.
Understanding the business situation is the immediate goal of these activities.
Monitoring displays support the following purposes: 1) Maintaining ongoing awareness of what’s going on and how well things are doing, and 2) reporting situations that require prompt action, either to correct a problem or take advantage of an opportunity. As such, displays of this type may prompt and support decisions.
Lookup displays, often called operational or production reports, are used to look up facts that are needed to do one’s work. As such, the information is usually presented in tables (i.e., arranged into columns and rows). As such, they communicate primarily through verbal channels (words and numbers that must be read) rather than visually (graphics).
Displays of this type are used for predictive analysis (a.k.a., predictive statistics or “what if” analysis) to anticipate what might happen in the future given specific conditions, based on an understanding of what has happened in the past.