We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.


Straightforward Analytics Visualization: Charts, Graphs, Widgets

Originally published June 28, 2012

My previous articles have largely focused on the creation of actionable knowledge or the means of delivery of analytical results, but less about the presentation. Yet in many cases, the delivery of analytical results is often couched in the context of how that information is to be used – triggering an action, communicating to other business users or monitoring sequences of events. This suggests that the presentation methods for specific pieces of information could be better crafted to best convey a message or trigger the appropriate actions. This idea can be extended to different methods of visualizing and then comparing analytical results.

There are many different types of visualization modes for data, and while this is not intended to provide a comprehensive overview of visualization techniques, it is meant to provide an overview of a handful of ways to present actionable knowledge:

Line Chart: A line chart maps points on a grid connected by line segments. A line chart can be used to show a series of connected values, such as a time series. An example is mapping the rise and fall of gas prices per gallon using the price of a gallon of gas on the first day of each month for the previous 36 months.

Bar Chart:
A bar chart maps values using rectangles whose lengths correspond to the charted values. Bar charts are good for comparing different values of the same variable across different contexts. An example is a chart of the average life expectancy in years across different countries.

Pie Chart:
A pie chart is conveyed as a circle that is broken into sectors representing some percentage of a whole. A pie chart is good for showing distributions of values across a single domain. An example is showing the relative percentages of owner-occupied homes by ethnicity within a zip code area. The total of the components always will add up to 100% and each slice of the pie represents a percentage of the whole.

Scatter Plot:
A scatter plot graphs points showing a relationship between two variables. Typically, one variable is fixed (the dependent variable) and the other is not (the independent variable). In a two-dimensional scatter plot, the X-axis represents the independent variable value and the Y-axis represents the dependent variable. A scatter plot is used to look for correlation between the dependent and independent variable. An example is graphing an individual’s age (the dependent variable) and the individual’s observed weight (the independent variable).

Bubble Chart:
A bubble chart is a variation on a scatter plot in which a third variable can be represented using the size of the item in the chart. An example is graphing the dollar sales volume by the number of items sold, with the bubbles representing the percentage of the overall market share.

Gauge:
A gauge is an indicator of magnitude in the context of critical value ranges. A gauge is good for conveying relative status of critical variables and points that should trigger an action. A traditional example is an automobile’s fuel gauge, which indicates the relative fullness of the tank, as well as an area close to the “empty” measure marked red to indicate the need for refueling.

Directional Indicators (arrows up or down):
These indicators are used for comparison to prior values. Often, these are represented using three images: one to indicate improvement, one to indicate no change and one to indicate degradation of the value. For example, directional indicators can be used as part of a time series presentation of stock prices to indicate whether the end-of-day price is higher, the same or lower than the previous day’s price.

Heat Map:
This is a graph that tiles a two-dimensional space using tiles of different sizes and colors. A heat map is good for displaying many simultaneous values yet highlighting specific ones based on their values. For example, a heat map can display the number of times each particular link on a Web page was clicked and can highlight the areas of greatest activity.

Spider or Radar Chart:
A spider chart displays a series of variable values across a collection of dimensions. Each dimension is represented as an axis emanating from the center with specific gradations. A set of observations can be mapped as points (and connected with lines). Different observations can be graphed using different colors. For example, a spider chart can look at a number of different characteristics of products (price, height, width, weight, mean time between failure) and relative success, allowing the analyst to quickly compare different products and look for correlations of the variable values.

Sparkline:
Sparklines are small line graphs without axes or coordinates. Many sparklines can be used in relative comparison regarding trends. For example, the trends of different stock price histories for similar companies can be compared to determine if there are industry trends relating to stock price.

There are many other types of visual “widgets” that can be used for presentation of information. A good resource for understanding graphic visualization is the classic set of books by Edward Tufte, particularly the first, The Visual Display of Quantitative Information.

Recent articles by David Loshin

 

Comments

Want to post a comment? Login or become a member today!

Posted August 20, 2012 by Lee Feinberg

David, this is a nice primer for choosing an appropriate visual.  I am writing about visualization in the context of how to drive decisions -- what I call Decision Visualization -- on the BeyeNETWORK here http://www.b-eye-network.com/view/16253.  I hope that some of your readers find it helpful.

Is this comment inappropriate? Click here to flag this comment.