The true value of a statistic in an entire population. The population parameter is constant while the sample statistic will vary from sample to sample. When a population is large, it is not feasible to compute population parameters by taking measurements of all members of the population, so a sample is taken from the population to make an inference about (estimate) the population parameters. For example, in political polling, we are interested in the proportion of individuals that favor a candidate. This is a population parameter. A random sample is taken to estimate that proportion.