Sampling and Sample Size + Ch9 Flashcards
Sample
Subset of population
Sample frame
Listing of population, from which you draw your sample
eg: McGill Students
Sampling error
Difference between population and sample due to chance variation
Can be reduced by increasing sample size
Accuracy depends on 2 components (RS)
Representativeness
Sample size
Census vs Sample
Census asks everyone - more accurate
Sample asks subset - accuracy varies
4 steps in a sampling process
- Define population of interest
- Identify sampling frame (if possible)
- Select sampling method
- Determine sample size
2 types of sampling methods (PN)
Probability method
Non-probability method
Probability method
Every element in the population has a known chance of being selected
More likely to be representative subset of the population
Allows you to calculate margin of error
Difficult + expensive
Non-probability method
Chances of selection for the various elements in the population are unknown
Unlikely to be representative subset of the population
Cannot calculate margin of error
Cheap + Easy
4 ways of probability sampling (SSSC)
Simple random
Systematic
Stratified
Cluster
Simple random sampling
Probability of selection is the same for all members
Systematic sampling
Order population randomly, select every nth member
Stratified sampling
Create mutually exclusive & collectively exhaustive subsets then sample randomly from each
Cluster sampling
Create mutually exclusive & collectively exhaustive subsets then randomly select subsets
4 ways of non-probability sampling (CPQS)
Convenience
Purposive
Quota
Snowball
Convenience sampling
Select participants who are easily accessible
Purposive sampling
Select participants using research’s best judgement about representativeness
Quota sampling
Select participants based on variables of interest
Snowball sampling
Select initial participants in some manner; ask initial participants for referrals for more participants
Skewed population
Distribution has a long tail on one end and a short tail on the other - leads to inaccurate results
Surrogate measure
Some easily determined characteristic of each population member is used to help partition or separate the population members into their various subgroupings