Six Do’s And Don’ts For Designing Chart Based Reports

Click Here To Know More About:

Submitted by: Jason Clauss

In this age of plenty, there surely exists no shortage of ways to adorn your data, but what if those 3D isometric bar graphs and piggy bank-shaped financial projections were in fact confusing your message? Maybe they are.

The real question is, what are you achieving by aggressively decorating your data?

Are you trying to impress your boss with your graphical chops?

Unless you work for the Pointy Haired Boss, all you’re doing is making it harder to figure out what’s going on.

Are you trying to cover up for a lack of useful data?

Someone will figure it out soon enough.

Are your reports strictly for your own consumption?

Then quit playing around and focus on the data!

[youtube]http://www.youtube.com/watch?v=OHRRXzIvJ7g[/youtube]

What is “chartjunk”?

Anyone who follows Edward Tufte knows about something called “chartjunk”. Chartjunk is a loosely-defined term encompassing any and all graphical bling-bling that clutters data visualizations be they intentional distortion of data for aesthetic purposes or extra illustrations added around the data to make the entire graphic more “exciting”.

Steps to creating awesome data visuals

With today’s software, anyone can make a graph in seconds. For those of us who have something useful to communicate, that’s a very good thing! But it also means that people with nothing useful to say can now say nothing in the form of a partially-transparent, 3D exploded donut with a nested 3D, not-so-exploded pie.

You cannot become an expert overnight but you can make a quantum leap in your own effectiveness by learning three things to avoid and three more to always do. I’ll start off with three don’ts so that you might cast off the dead weight of bad ideas before raising yourself up with good ones.

Three to avoidNo 3D!

Despite living in three dimensions, we really only see in two. Depicting graphs in anything but a head-on perspective distorts the shapes, making it very hard to read. Isometric perspective is bad but natural perspective, with its converging lines, is worse! This surrealist pie says it all.

Don’t add distracting textures or reflections.

Simple colors will communicate your point just fine. Anything more may actually be counter-productive. For instance, hatch lines are not just hard to read but potentially painful thanks to Moire vibration. Meanwhile, those cute reflective effects available in some visualization applications can be downright misleading. Look at this example by graphics guru Stephen Few and determine at first glance, if you see two or three slices in this pie.

Don’t add weird or confusing shapes.

Anything that distracts the user from the data is bad. In fact, as a rule, anything that uses area rather than length to encode values is a bad idea, let alone irregularly-shaped objects. I cite this infographic of Labor Day wedding locations; the size of the cake denotes popularity. Are we supposed to consider only the height of the cakes, or does the width have meaning? Are we looking at the general volume of the cake? Stephen Few often criticizes pie charts, but even he would admit they are far better than a cake chart.

Three things to embraceUse colors logically.

Colors can either enhance or hinder the chart, depending on how they are used. Most important of all is consistency, ensuring that a color isn’t given two different meanings (red for bad, and red for emphasis, for instance). Use common idioms such as red-yellow-green for status graphics. Conversely, you might want to offer a version of your visualization readable by those who are colorblind – they make up as much as 10% of the male population and a smaller percent of the female population.

Use the right type of graph.

Using the right type of graph is a valuable UX affordance to the reader, framing the information in such a way that they properly interpret it. Failure to do so could result in the reader misinterpreting the information, so choose your medium wisely.

For example, line graphs portray information as a continuous stream and work well to show patterns over time. Bar graphs portray information as discrete entities which works better for categories or individual time periods. Newer styles of chart like spark lines and bullet graphs are based upon lines and bars but condense data for rapid consumption.

Supply context.

Simply throwing numbers at the user helps no one. For instance, a gauge graphic that indicates sales for the month are at $500,000 with no further information is useless. How does that figure compare to the competition? How does it compare to the previous month, or year? How does it measure up against predictions? Any of these would be helpful context, turning meaningless data into useful information that you can act on.

Therefore, make sure that every one of your graphics provides adequate context. Compare your numbers to benchmarks, and as mentioned above, use color coordination where appropriate.

These six tips will make you much more effective at communicating your data visually.

About the Author: To learn more about

self-service reporting

or

ad hoc reporting

, please visit the ActiveReports Server web site at activereportsserver.com.

Source:

isnare.com

Permanent Link:

isnare.com/?aid=1263572&ca=Computers+and+Technology