Sampling Methods

Probability Sampling

some form of random selection

simple random sampling

randomly draw n items from the population (ex. from a hat)

5% rule

to ensure a random sample, never take more than 5% of the population

advantage of simple random sampling

simple, cheap, easy

disadvantage of simple random sampling

minorities are at risk of underrepresentation

systematic random sampling

define stride and randomly select an offset; typically used in manufacturing

advantages of systematic random sampling

simple

disadvantages of systematic random sampling

risk of systematic errors being missed

stride

m=⌊N/n⌋

cluster sampling

population is divided into clusters and then are randomly selected; all items in the selected clusters are used as samples

stratified random sampiling

divide the population into strata, label items in each group, and randomly select items from each stratum.

non-probabilistic sampling

non random sampling

snowball sampling

participants are recruited by referral from previous study participants

convenience sampling

participants are recruited from easily accessible cohorts; sometimes only conveniently available or volunteered data is collected

quota sampling

The population is divided into groups based on order parameters and the experimenter chooses the items drawn from each group based on getting an equal number of items from each group or the presence of characteristics that were not used for stratification