POLS 2503 Lecture Notes - Lecture 11: Observational Error, Statistic, Sample Size Determination
Document Summary
Three factors that determine how closely a sample statistic represents a population parameter: sample size, procedure used to choose the sample, amount of variation in population characteristic. Sampling: a valid sample is based on random selection. If the sample is not randomly selected, size doesn"t matter: eliminating bias does not eliminate error. Population parameter = sampling statistic + random sampling error: random sampling error = variation component/sample size component, large sample + low dispersion = less random error and greater confidence in sample statistic. Standard error of a sample mean = random sampling error of a sample mean. Small standard deviation values of individual cases are similar to the mean of the distribution: large standard deviation values of individual cases are not similar to the mean of the distribution.