Showing changes to a histogram

Many kinds of data from a single time point can be displayed as bar or pie charts. However a histogram is needed to graphically display a frequency table for a continuous measurement such as the age distribution of a population.

The next example shows how a histogram varies over time. It would be difficult to show the same information in a static display.

Ages of registered cars in New Zealand

The histogram below initially shows the age distribution of cars that were registered in New Zealand in 1998.

Drag the slider to see how the age distribution changed since 1998.

'Births' and 'deaths' of cars

In recent years, a high proportion of cars sold for the first time in New Zealand have been second-hand cars that were imported from Japan. Unlike most other countries, 'new' cars are therefore introduced at ages 1-10 in addition to brand new cars at age 0.

The histogram below shows in blue how many cars of each age were more than would have been expected if the previous year's cars had just aged by one year (after 1998). This describes the ages of most cars that were new to New Zealand roads in the year.

In a similar way, if the number of cars of one age were less than would have been expected if the previous year's cars had aged by one year, the difference will be cars that 'died' during the year and these are displayed in red.

Drag the slider to see why the large number of Japanese imports at ages 1-10 have resulted in a bulge in the age distribution of cars around 10 years old.

(Most imported cars are less than 10 years old and most cars are over 10 years old when they are taken off the road. However there will be some overlap so the blue and red numbers actually reflect a mixture of the two effects.)