Horses for courses: Which diagram for which data?
Sound familiar? You have piles of data, most of it in Excel, and are trying to work out how best to get it onto a PowerPoint slide. The simplest way is ‘copy and paste’. But it’s certainly not the best way. Large amounts of data are often confusing, because it’s hard to single out what really matters. There’s also the danger that audience members will be distracted by something that is irrelevant to the point you are trying to make. So: be clear in your mind what you actually want to say, and then present only those data that support your message.
Let’s take company staff numbers as an example. If the distribution across various countries is of interest, a structural representation is appropriate – e.g. a pie chart. If, on the other hand, you want to show an increase in staff numbers during the past few years, a line diagram is a better choice.
The best way to choose the right diagram is by considering the data type:
- Structure, e.g. the market share of different competitors in a country or the breakdown of items in a cost center. Here pie charts or bar charts best illustrate different compositions. Pie charts are particularly suitable when you just have one category of data. If you are comparing more than one category of data, bar charts are clearer.
- Rankings, e.g. turnover of various companies, ranked from the largest to the smallest. These can be presented nicely using a horizontal bar chart. Both the ranking and the relative differences between the companies can be seen clearly.
- Time series, e.g. development of turnover over a number of years. A line diagram is best for a large data set, and a column chart for a small data set.
- Frequency, e.g. the distribution of production losses. Line or column charts are also suitable here, again depending on how many data points are being shown.
- Comparison and correlation, e.g. a comparison of the proportion of men and women in various age cohorts, or the correlation between product price and turnover. This type of data can be illustrated with butterfly bar charts or scatter plots. Adding a trend line to a scatter plot helps to highlight the correlation.
See also Gene Zelazny „Say it with Charts“