CMN 102 Chapter Notes - Chapter 8: Simple Random Sample, Confidence Interval, Sampling Error
26 Oct 2016
School
Department
Course
Professor
Document Summary
When sample size increases, there is better distribution of mean estimates. If selected independent random samples from population, its sample statistics distribution around the parameter is known. Probability theory provides a formula to estimate the closeness sample stats are clustered around true value. Sampling error: expected degree of error for a specific sample design. Standard error is an inverse function of the sample size. As sample size then standard error . Confidence level: est probability that population parameter in within confidence interval. Confidence interval: range of values where population parameter should be in. List that probability sample is chosen from. Must be consonant w/ population under study. Simple random sampling: assigns number to elements. Stratified sampling: grouping the units in homogeneous groups before sampling, obtains higher degree of representativeness, sampling error is decreased, function: to organize into homogeneous subsets and select appropriate # of elements from each, will be representative on variables. When it"s impossible/impractical to create exhaustive list.
