A number describing data.
A sample statistic summarises some aspect of sample data, it is generally also treated as an estimate, perhaps a least worst estimate, of a “population parameter”. For example if the mean baseline CORE-OM score from a year of data collection in a service is 24.3 that is the sample mean: it is the total of all the scores divided by the number of scores, it is an indicator of central location and be compared with that from another service. That is “descriptive statistics”. At the same time it may be regarded as a best guess of the mean for that service next year, or for all similar services: the population of interest to which perhaps we are interested in generalising this particular finding.
Try also #
Confidence interval (CI)
Mainly chapter 4 but the idea runs through all the subsequent chapters.