Using samples to challenge assumptions about population parameters to standards or other populations, and making statements of statistical significance about the populations. It is a branch of statistics that specifically determines whether a particular value of intrest is contained within a calclated range (the confidence level).