Replacing bars with objects

A second type of chartjunk is obtained by replacing the rectangular bars in a barchart with pictures of objects. This a much more serious problem since it often visually mis-represents the proportions in the different categories. Are the values proportional to the heights of the objects, their areas on the paper or their 3-dimensional volumes? At a quick glance, most readers would use something between area and volume though it is usually the heights of the bars that actually determine the size of the objects in this type of diagram.

Recreational participation in New Zealand

The diagram below was published to show the percentage of men and women participating in various recreational activities in 1996,

This diagram can be criticised in several ways (e.g. the 'women' stand a little above the axis and it appears that no women do running or jogging whereas this category is just not in the 'top ten activities for women so it has been omitted). However the main problem is illustrated by a comparison of Gardening and Swimming. The proportion of women doing Gardening is less than twice the proportion doing Swimming, but the difference seems much greater since the area of the Gardening woman is about 4 times greater and her volume is about 8 times greater.

Using pictures of objects instead of bars in a barchart is misleading and must be avoided.

Merit raises

The next example is not one from official statistics but is particularly misleading. As part of a study of how merit pay policies are tied to employee performance, data were collected about the merit raises (measured as a percentage of salary) for 3,990 employees in a large company. The diagram below was published to summarise the data.

Again, the visual impression is misleading because the widths of the carrots also change. In particular, the employees getting under 5% merit increase seem visually unimportant, but they comprise nearly 10% of the total employees.

If you are determined to use pictures...

Pictures in bar charts are really chartjunk and should be avoided — they do not make the information in the data clearer and have the potential to mislead. However if you are determined to use them, it is much less misleading to use stacks of pictures that are all of the same size.

Example

The bar chart below might be used to represent some aspect of car ownership or accidents.

Simple bars would be clearer, but the 'ink' used in each stack of cars is roughly proportional to the group's value, so the diagram does not badly mislead.