LectureResMethOzer.ppt

Stratification vs. Clustering Stratification Divide the population into groups that are different from each other Sample randomly from each group. Less error compared to simple random sampling. More expensive to obtain stratification information before sampling. Cluster Naturally divided population Randomly sample some of the groups. More error compared to simple random sampling. Reduces costs by sampling only some areas or organizations in manageable clusters. * Stratified Cluster Sampling Decide on the clusters. Group the clusters into strata of clusters, putting similar clusters together into a stratum. Randomly pick one (or more) cluster(s) from each of the strata of clusters. Sample the subjects within the sampled clusters (i.e., either all the subjects, or a simple random sample of them). * Stratified Cluster Sampling You can reduce the error in cluster sampling by creating strata of clusters. Sample one or more cluster(s) from each stratum. This combines the cost savings of clustering with the error reduction of stratification. * Systematic Sampling Systematic Sampling …a procedure by which the selection of the first sample member determines the entire sample. The population is in some type of organized order (e.g., alphabetical). Then, the sample size and sampling fraction determined (e.g., “There will be 300 people in this sample, and every 3rd person will be selected.”). * Systematic Sampling - Periodicity For example, if you selected the first person out of every group of four, then none of the checked samples are selected. Periodicity means that every kth member of the population has some unique characteristic(s) that are related to the DV. * Sample Size and Power What influences sample size? Practical Considerations Time Money Effort Statistical Procedures Considered Statistical Precision Larger samples yield greater statistical precision. Statistical Power Deals with whether a result is statistically significant (i.e., not due to chance). It is the