To Display, or Not Display, Data that is the Question?
When choosing to represent data in graphical form there are many options one has at hand. There are pie charts, Gantt charts, bar charts, line graphs, linear regressions, normal distributions, and countless others; the list really is endless. With this many possibilities of representing data, it can become confusing to weigh the options of which type will be best in which situation. To choose one it is best to consider what is discussed by Kostelnick and Roberts in their text Designing Visual Language Strategies for Professional Communication. They start by defining a data display, "data displays show quantitative information by transforming textual elements--usually numbers--into images." They go on to add, "data displays are extra-level in nature (Kostelnick, 245)." This means that their informative power is based in the imagery, and not the text contained in the graph. The final key to a data display that Kostelnick and Roberts emphasis most, is the ethics of the display. They describe the ethics as the ability of the graph or chart to most clearly present the data with out possibility of misrepresentation. While the process of choosing data displays is fairly straight forward, when explained by Kostelnick and Roberts, there are millions of terrible graphs waiting on the internet.
This bar graph is an example of an unethical graph. One, the bar graph representation is unfit for the data they are displaying, a pie chart would have been much more effective since their data is in ratios. Two, another problem with this chart is the fact that the x and y axis of the graph are not labelled, and there is no title of the chart. Axis labels and Titles are necessary on every data display, otherwise it is impossible to tell what the graph represents, as is the case with the display above.
Another example of an unethical graph is the UK Internet traffic to Social Networks line graph. This chart, while the correct choice for this type of data, is unethical because the y axis is scaled incorrectly, therefore, it over-represents the percentage of traffic growth by not including all the percentages down to zero.
With so many options, and countless data sets to represent, please remember to keep your data displays ethical.
Add a comment