The article titled Effectively Communicating Numbers is a white paper explaining the power of quantitative and how best to project this power through the right visualization. This includes graphs, tables, graphics and other visual representations of data. The main two lessons the author Stephen Few wants the reader to learn are simple in nature but more difficult to execute. This is due to either preconceived notions of how data has been presented before or the lack of experience presenting data in a clear and precise manner. The lessons are how to match ones message with the right display/visualization platform and how to design each component of the visualizations to clearly and loudly showcase the data.

The first lesson is demonstrated in a small example where a hypothetical person has to present data for a meeting. They decide to showcase a complex radar graph to show data relating to three people on 5 attributes. The graph, though visually striking, lacks clarity and complicates the possible message that the data is telling. Tables with simple numeric values would have better represented the data across each person’s attribute score. The author then proceeds to break down the processes of constructing said graphs and data visualizations as followed (note 3-6 are for if data visuals are required):

  1. Determine the message to be analyzed and recognize the data needed to communicate the message clearly
  2. Determine if a table, graph or combination is needed to showcase the message.
  3. Determine the best means to convert the values of the data spread
  4. Find out where to display each variable
  5. Determine what design fits for the other objects from the data spread
  6. Figure out what particular data should be highlighted above the others and how best to showcase it in contrast.

The article then divulges into ‘tables versus graphs, showing when one is superior to another in different circumstances. Another compare and contrast object to observe and analyze when making data visuals is quantitative versus categorical data. Quantitative deals with exact values and what those values means. Categorical is data that is scaled nominally, ordinal or in intervals (think time, places, size categories, ranking, etc.). The author then describes in detail seven common relationships in quantitative business data: Time-Series, Ranking, Part-to-Whole, Deviation Relationships, Distribution, Correlation and Nominal Comparison Relationship. Following the definitions, Few shows the best means to ‘encode’ quantitative data in graphs. This includes points vs lines vs bars vs box graphs. The last third of the article goes into the 6 fundamental steps for when and how to graph data through various examples and demonstrations of both correct and incorrect usage of different graphing techniques.