SOC221H5 Lecture Notes - Lecture 6: Stratified Sampling, Random Number Table, Simple Random Sample
Document Summary
In 2010 the census decided to get rid of the long form which contained 40 personal questions (from ethnicity to amount of time spent doing unpaid work) The conservative government argued that these data could be collected through a voluntary household survey. Researchers were concerned because research shows that voluntary surveys are less likely to be returned by ethnic, linguistic and economic minorities- the very groups that social policy researchers who use these data are trying to help. The data would not be representative of the canadian population because poor response rates. The result of nonresponse bias is that a sample is not representative of its target population, making generalizations from the data produced from such samples questionable. Sample: a smaller set of cases a researcher selects from a larger pool and generalizes to the population. Population: large general group of many cases from which sample will be drawn.
