3. Sampling Flashcards
What are the two directions or “sign” of a relationship
- Positive: When X goes up, Y goes up
- Negative: When X goes up, Y goes down
What are two types of Research Studies
- Observational Studies (collecting data while not directly impacting how the data arise)
- Experiments (researchers manipulate a variable and look for an effect: must have control group and experimental group)
Why are experiments great at finding causal relationships?
Experiments address spuriousness by having a control group and an experimental group - and these groups are split randomly.
Randomization kills any possible…
…spurious variable’s cause of X. Randomization ensures that any potential confounding variables (Z) are evenly distributed across the groups.
Sampling frame
the set of cases that could be selected into the sample
You draw from the population to get a sample and the sample is for
generalization to the public
how much your sample is like the population
Representativeness
the idea that your sample will tell us something about a larger group than just the sample
Generalizability
What are the two types of sampling?
- Probability Sampling (the chance of a unit’s selection is known, and has a non-zero chance)
- Non-Probability Sampling (the chance of a unit’s selection is unknown - not everyone has a chance to get selected)
What’s an example of a non-probability sample?
Convenience Sample
4 Types of Probability Sampling
- Simple Random Sample (truly random)
- Systematic Sample (every nth)
- Cluster Sample
- Stratified Sample
What probability sampling randomly samples existing clusters first then samples within those groups?
Cluster Sample
What probability sampling creates subgroups of the sample based on variables then samples from within those subgroups?
Stratified Sample
