4.2 Representing context with statistics
Arguments around the transportation needs of people are typically supported by representations of associated statistics that describe the financial, environmental, or social context. As a method of communication, statistics can be very effective because they can be used to distil the essence of an argument and to make key points apparent and incontrovertible. They are also psychologically effective because they suggest formal, scientific analysis of the facts, based on raw data. There is an impression that statistics cannot lie. But, all statistical representations result from decisions about what data to represent, and how to represent it. These decisions are made based on consideration of the message that is to be communicated and on the target audience. For example, Figure 17 illustrates statistical representations presented to support arguments for and against the HS2 scheme. Here, the data is concerned with predictions of future trends in rail use. The first graph was composed by the ‘for’ campaign and is included in the government-produced ‘The Strategic Case for HS2’. The accompanying text describes this as a conservative prediction of future rail use, below expectations based on trends over the past 20 years (Department for Transport, 2003). The second graph was composed by the ‘against’ campaign and is a re-examination of the data presented in the first graph. It presents a prediction of future rail use based on trends over the past 50 years, and extrapolates the government’s prediction over the longer time frame. The accompanying text brings into question the government’s conservative estimates, and suggests that they are inflated (Stop HS2, 2011).
At first glance the two graphs seem to be quite different, and in both cases the statistics seem to be supporting the arguments presented. But the graphs are based on the same data, and differences arise because of decisions about how this data is represented. In the first graph, less historical data is represented, to give the impression of a continual growth. In the second, historical trend and government prediction are represented as exponential curves to emphasise the differences. This illustrates the important lesson that statistical representations should not be accepted at face value; instead it is beneficial to try to understand the source of the data being represented and the decisions that have been made in choosing how the data is portrayed.