Graphs and tables
There is a common saying that "a picture is worth a thousand words". This is particularly true in the context of understanding the information in data. Tables and text may display a wealth of detailed information, but a well-designed graph can be much more effective in highlighting the important features in the data.
This section introduces some graphical displays of the types of tabular data described in the previous section.
Bar charts
The most important graphical display of a simple table of values is a bar chart. Each category is represented by a bar whose length is proportional to its value.
The bars can be oriented either vertically or horizontally. Vertical bars are often used for small tables, but if there are many categories or if the category names are long, a horizontally oriented bar chart is usually clearer. A gridlines behind the bars can help to read off values, but they should be very faint so that the bars remain prominent.
Infant mortality rate in New Zealand
The indigenous people in New Zealand are Maori who form about 15% of the population. Most of the others are of European descent with a sizeable number of immigrants from adjacent Pacific Islands (about 7%).
The frequency table and bar chart below show the infant mortality rates per 1,000 live births for these three ethnic groups in 2001.
Click on any bar to see how its height equals the ethnic group's infant mortality rate in the frequency table.
The pop-up menu can be used to change the orientation of the bar chart and the two checkboxes can be used to add the numerical values of actual mortality rates. If possible, it is better to print values inside the bars to avoid giving the visual impression that the bars are longer than they are.
Proportions
If the values form a frequency table or a partition of some other total quantity, a second axis can be added to the bar chart, allowing both values and proportions to be read off the graph.
Tanzania drinking water
The bar chart below shows the sources of drinking water for rural households in Tanzania in the dry season of 2002.
In this example, a second axis shows the percentage of households in each category. Click any bar to read off the corresponding number and percentage.
Note that horizontal bars are better than vertical ones for this example since there are many categories with long names.