What are misleading statistics?

Answer 1

Almost all statistics can be misleading depending on how they are presented and reported.

There is always a margin of error, for example the sample size used to generate the statistics.

For example:

If I ask two random people on the street if they drink beer. I may get the following response: neither drink it, one person drinks, both drink it.

I cannot quite rightly say that this is a true reflection of the population, as I have only sampled two people. Also what about the sex of the people? Men are more inclined to drink beer than women. What about the age of the two people I sampled? If they were minors they may not drink at all (or possibly they do). If I were to only ask men, this would show biase in my sampling.

Therefore the figures need to be presented in context, also a representative sample size needs to be taken - the larger the sample the more accurate the results (or higher the confidence).

Another concept is that statistics are not necessarily mutually exclusive.

Again I ask 10 people on the street: Do you drink Whisky or Beer?

If 6 people answer positive for drinking whisky, this does not necessarily mean that the other 4 people automatically drink beer. They may not drink at all, or they may drink both.

Statistics are useful if interpreted correctly, but they should always be presented in context.