STAB22H3 Lecture Notes - Lecture 6: Survey Data Collection, Multistage Sampling, Simple Random Sample
38
STAB22H3 Full Course Notes
Verified Note
38 documents
Document Summary
Population: entire gp of individual that we want info abt. X predict which individuals @ sample -> large enough sample = approximately right proportion of diff factors. Size of sample x size of population matters except sample > 10% of population cuz population size = small. Survey: census - data collection from every unit @ population, sample survey - data collection from 1 pt of unit @ population -> 1 x always ok cuz too time-consuming & costly. [a] if needa update frequently = x feasible. [b] population ch while collecting data -> x accurate. That describes sample which can ch from sample to sample. Simple random sample: every individual has equal chance to be selected -> need sampling frame = whole population list. Stratified sampling: strata = homogeneous, non-overlapping gps -> select 1 individual srs @ each stratum -> obtain info from all strata & combine srss = full sample; (+ve) 1. ensure each stratum represented 2.
