Sampling Flashcards
What are the four types of probability sampling?
Simple random
Systematic random
Stratified random
Cluster
What are the four types of non-probability sampling?
Convenience
Purposive
Quota
Snowball
What is simple random sampling?
“names out of a hat”
PRO: each person has an equal chance of selection
CON: non-proportional sample = pure chance
What is systematic random sampling?
Take every nth name on the list
PRO: easy to administer
CON: prone to bias if there is a relationship
What is stratified random sampling?
Divide the population into strata based on a characteristic and then randomly select a proportion from each stratum.
PRO: decreases sampling error.
CON: resource-heavy
What is cluster sampling?
Divide the population into strata, randomly select two or more strata and sample all.
PRO: cost-effective
CON: reduces representation
What is convenience sampling?
Easy to access (e.g. pick everyone at a grocery store)
PRO: cost-effective
CON: biased sample
What is purposive sampling?
All those available who fulfil a certain criteria
PRO: cost-effective
CON: biased sample
What is quota sampling?
Reach a predefined quota of a characteristic (e.g. first 20 men and women)
PRO: cost-effective and researcher can control demographic differences
CON: biased sample
What is snowball sampling?
First recruit recruits others they know that fit the demographic
PRO: good for hard to reach populations
CON: limits external validity
What is type 1 error?
False-positive. There is a perceived difference between the two groups. Wrongly rejecting the null hypothesis. Too many people in the sampling frame
What is type 2 error?
False-negative. Perceived no group difference. Wrongly accepting the null hypothesis. Caused by a sample that is too small.
Describe the 68, 95, 99 rule
68% of values fall within the 1st standard deviation of the mean.
95% of values fall within the 2nd standard deviation of the mean.
99% of values fall within the 3rd standard deviation of the mean.
What does a negatively skewed distribution look like?
Right is higher than the left
What does a positively skewed distribution look like?
Left is higher than the right