Statistics are frequently quoted as evidence to support a claim that is being made. On the surface, they usually appear to be valid – they are important evidence for our mind-computer to process.
- How reliable is the source of these statistics?
If the figures are quoted by a team-member in a debate, there is a good chance that they may be ‘made up’. Certainly they need further checking! On the other hand, if the statistics are offered by a reputable opinion – poll group, or the Commonwealth Statistician, or a respected textbook, we can have considerably more confidence in the source.
- How representative is the sample from which the statistics have been drawn?
- large enough to reflect the whole population. The larger the sample, the more confidently can we accept conclusions drawn from it. It would be risky to draw conclusions about the academic standards of large high school, for example, by looking at the exam results of just ten students. On the other hand, we might feel quite confident of conclusions drawn from a sample of a hundred students.
- Random. By random we mean that the people or objects used as the sample have been chosen as though by chance. For example, choosing a thousand people from the wealthier areas of a large city, in order to assess the average income of the people of the city, would result in disorted conclusions being reached. The sample was not random. People in poorer areas were not given any chance of being represented in the sample.
- How relevant are the statistics to the conclusions?
Sometimes statistics are misused to impress us or confuse us – so that on further analysis they may well turn out to be irrelevant or, at the most, to have a tenuous relationship to the conclusions which are reached. For example, would statistics quoted about the increasing member of car accidents in Australia be a strong argument if used as support for further government spending on public transport?
