Andrew+M+R05

My favorite examples were of the bending bar graphs. I’ve never seen this approach and I’ve never thought of it before and I must say that it is a pretty bad approach. The viewer definitely loses all sense of scale, and therefore impact, with a visual that does not consistently compare to similar data depictions. The meandering river that wraps around the page is the only example I could see as being valid, since a river does meander, but even that example loses it’s visual meaning as a representation of data that is to be compared to other representation of data.

Handling extraneous outliers is always a problem and spans disciplines from art to mathematics. Wile working on my income/expense/school/work graph I was hoping I would not have any dramatic points that would “go off the charts” as they are often design roadblocks. The worst attribute of an outlier is that if you should choose to force it into your current design, it can become too small to see or so big the rest of the data is lost. However, an outlier could reveal a weakness in your approach to presenting data; and if an outlier happens once, it could be a reoccurring trend that has yet to unfold.

Tuft describes William Playfair’s graph as “ingeniously spilling outlying data over…” the edge of the graph, but this approach feels very dramatic or like a cry for attention. Being such a dramatic approach, spilling over should only be used for special representations, such as epidemics or violence. I also feel this approach is similar to those people who give 110%; there is no such thing, if you’re going past 100% then you need to recalibrate your data or how you interpret it.

Outliers don’t have to be a negative occurrence; they can provide an opportunity to rethink how you’re displaying your data. If you have one outlier now, chances are you will find more as data collection and presentation continues. Catching and successfully incorporating outliers early on will prevent you from reorganizing your entire presentation later in a project. They can also inspire a new layout, which can either emphasize or mute the outlier. The rules of hierarchy can be often be used to reduce the visual significance of data by reducing it’s visual contrast, unless the outlier is welcome, then it can tower over your data using it’s size to give hierarchical importance.

It’s important to remember the purpose of creating charts and info graphics, which is to convey information; keeping this in mind will aid in how to treat data that does not paly nice with the data around it. Is this data important to the goal of the graphic or is it a distraction, maybe it shows a flaw in how the data was initially collected. Either way, there’s nothing like a good outlier to bring an entire project to a halt or even be its undoing.