Producing Data Flashcards
Sample Survey
Participants provide the data/responses
Observational Study
Experimenter observes the participants and records the data or variables of interest
Experiment
Experimenter deliberately “imposes a treatment” on the units in order to observe a response
Population
Entire group of individuals that we are looking do describe
Population Parameter
Number used to model or describe a population (uses greek letters)
Census
Entire population is sampled
Sample
Smaller group of individuals selected from the population
Voluntary Response Sampling
Groups of individuals are invited to respond and data is collected
Convenience Sampling
Sample those who are convenient to us
Bias
Certain outcomes or views are systematically favored over others
Measurement Bias
Data or measuring of data is not accurate or doesn’t represent all views
Response Bias
Anything in survey design that influences responses (eg. wording)
Non Response Bias
Those selected do not respond
Undercoverage Bias
Some segments of the population are underrepresented or aren’t represented properly (result of bad sampling frame)
Sampling Frame
Group from which you’re selecting your sample
Simple Random Sample
Number everyone in sampling frame, use random digit table to select random units
Systematic Random Sampling
Select every nth unit - Where to start? (pop size / sample size) = n R p, select random 1-(n+p), start with that
Stratified Random Sampling
Break population into classes (strata), conduct SRS within each group
Quota Sampling
Similar to Stratified Random Sampling, but number of selected units from each strata is based on percentage of population that strata takes up
Cluster Sampling
Select everything geographically close
Multistage Cluster Sampling
Survey US pop -> Select 5 states -> 3 counties in each state -> 4 blocks in each county -> everyone on block surveyed
