HMEC 2000 Lecture Notes - Lecture 1: Dependent And Independent Variables, Lybia
Document Summary
This creates a systematic bias by ensuring your sample will overrepresent opinions of those frequenting the library and underrepresents those that do not use the library v random error v has no systematic biasing effect, error results from chance v the larger the sample size, the smaller the sampling error v text pg 114 outlines the improved confidence inherent as the sample size increases v reduce margin of error by half if we quadruple the sample size (gray & Stratified random sampling: proportional or disproportional v proportional stratified random sampling v the same sampling ratio is used within each group (strata) v each group is representative of the population being studied v sampling ratio remains static, n value changes v disproportional stratified sampling v use when 1 or more of the subgroups is too small for adequate analysis v used when comparing results between groups v sampling ratio changes, n value remains static.