Data with two categories
For data with two categories in which the values are partitions of a meaningful group total, we are usually most interested in the proportions in the two categories.
Examples are:
When there are only two categories, there is redundancy since each proportion is one minus the other. For example, the proportion of males in a city is one minus the proportion of females.
If there are two categories, only tabulate or graph one of the two proportions.
Small proportions
If the proportion of interest is always small (e.g. the proportion of children with AIDS/HIV), there is a second reason for only displaying this proportion and not the other (which will be always near 1.0). If both are displayed as bars on the same scale, there will be little visual difference between the values of the proportion of interest. By restricting attention to the smaller proportion, a more appropriate scale can be used.
Percentage aged over 65
The table below shows the numbers of New Zealanders who were over 65 in the 2001 census, in the four major ethnic groups.
Number (thousands) | |||
---|---|---|---|
Under 65 | Over 65 | Total | |
European | 2469.4 | 398.7 | 2868.0 |
Maori | 508.6 | 17.6 | 526.3 |
Pacific peoples | 224.2 | 7.6 | 237.5 |
Asian | 227.7 | 9.8 | 237.5 |
The diagram below shows stacked bar charts for proportion over and under 65 in the four ethnic groups.
Since the proportion aged over 65 is small for all ethnic groups, the differences are not displayed well. Choose Propns for Over 65 from the pop-up menu to hide the bars for 'Under 65' and expand the vertical scale. The resulting diagram looks like a simple bar chart of the proportion aged over 65 in the four groups.