#71 Are Your Graphs Wasting Everyone's Time?

Software these days can do heaps of exciting and creative things with your data, including summarising it, graphing it and making it flash and ding to get your attention.

Because all this can happen at the touch of a button, it's often hard to remember that software can't reason and therefore can't decide what is a meaningful and useful thing to do with your data.

These principles of useful graphs will guide you so you can decide what is a meaningful and useful thing to do with your data.

Principle 1: answer your question

When you choose a graph to convey information for a specific purpose, be clear about what the purpose is and choose measures that are relevant to that purpose. For example, make sure your graph title explains the measure presented, and make sure the measure presented really is the relevant measure for your decision.

Principle 2: highlight comparisons

Graphs are a visual representation of objective comparisons. When you graph measures, it is important to use techniques that help the user make comparisons in the information you are graphing. For example, use line charts for time series measures, use bar charts to compare departments or segments or other categories with one another, avoid pie charts altogether.

Principle 3: optimise simplicity

Keep graphs simple and easy to interpret by avoiding superfluous symbols, laborious legends, dispensable decorations and volumes of information that take the audiences' attention away from comparisons. For example, remove gridlines, don't bother with data labels and forget the artwork (like little dump trucks instead of data points).

Principle 4: maintain integrity

Help the data tell its story. If this story is unclear, then say so (as opposed to putting the data through contortions and manipulations for the sake of having an answer to your question). For example, be careful how you scale the axes, don't use 3-dimensional effects and don't assume every trend is a nice straight sloping line.

Principle 5: be visual

Pattern, colour and space are the fundamental characteristics of graphs. These characteristics are processed and given meaning through a visual process, so use them in ways that human eyes (and often photocopiers) best respond to. For example, avoid patterns that don't start to move the more you stare at them, use colour to highlight signals in the data and size your graph only just large enough to convey its message.

Tips to transform your graphs, easily and quickly

How easy is it to transform your graphs? Very easy! You just need to know the practical tips that show you how to change parts of your graph's anatomy to make your graph more relevant, informative, simple, trustworthy and easy to interpret for decision making!