Inferential Statistics

Process of drawing information from sampled observations of a population and making conclusions about the population. Inferential statistics have a two-prong approach. First, sampling must be conducted to be representative of the underlying population. Second, the procedures must be capable of drawing correct conclusions about the population. See also Descriptive Statistics.

To infer is to draw a conclusion from only partial evidence, such as a sample, so inferential statistics are techniques which may be used to evaluate the value of such conclusions; they tell us the probability of a statement based on a sample being valid for the whole population. This probability is expressed as a percentage, and a 90% probability rating tells us that that deduction would be made from 90 out of every 100 samples. This calculation of validity should be included in any geographical enquiry based on sampling. See also descriptive statistics.

Complex statistical techniques used to infer cause and effect, and to determine the degree to which the findings of a sample can be generalized to a larger population. For example, in a study of the effect of a pre-match pep-talk on performance, a researcher may design a study with appropriate control groups, and use an appropriate statistical analysis to see if a significant difference occurs between the test and control groups. If a significant difference exists, the researcher would infer that the pre-match talks caused the effect. Compare descriptive statistics.