The estimate of a parameter, which is obtained by observation, and which in general is subject to sampling error.