4 Sampling Flashcards
basics of sampling
- populations
- representativeness
- probability
- sample frame
- sample selection method
- sample size
- sample mode
- weighting
- sample time frame
population
simply all the ppl or things that comprise your object of research; i.e., all the cases you’re interested in
population parameter
a known characteristic of a population; i.e., the basic knowledge about the makeup of that population
sampling
the process of taking a sample or a small amount of cases from the larger population
representativeness
to make generalizations about a population from a sample with confidence, research tries to have samples that mirror, as closely as possible, the larger population from which the sample was drawn; the closer your sample represents the population, the more certainty you can make a claim with
sampling is based on:
- probability theory
- central limits theorem
Probability Theory
- allows the researcher to estimate the likelihood that their sample provides a representative picture of the population they’re studying
- if you take a random sample of ppl from a population, as long as you get enough, you will have a representative sample
- this theory hinges on the idea that every case within the population has an equal chance of being chosen, so you have to create the conditions for this to be true
Central Limits Theorem
- states that regardless of a population distribution, as a sample size increases, the distributions of the sum and mean of random samples approach a normal distribution
- emphasizes random selection in samples
- using random samples allows you to determine the likelihood that a calculated statistic for your sample is a good estimation of a particular population parameter (margin of error)
what does a representative sample depend on?
- sampling frame
- sampling selection method
- sample size
- survey mode
sampling frame
a list of all the ppl in the target population and hopefully their contact info; systematically excluding some from the sample will result in inaccuracy
random digit dialing
computer generates all the numbers that exist within an area number and selects numbers out of this to ensure randomness
sample selection method
the process by which ppl are chosen to be included in the sample; can be random or non-random
sample size
- the number of ppl that are part of the population who are included in the sample
- to increase confidence level and decrease margin of error, you need a bigger sample size
- once you reach a certain point in the population, your sample size doesn’t need to be any bigger
survey mode
- how the individuals in the sample are contacted and how they’re administered the survey
- systematically, some don’t respond to surveys depending on their mode, leading to selection bias and difficulty in representative sampling
- sampling frame and sample mode are related bc the ability to have a randomized sample is dependent on the mode
weighting
- responses from respondents from overrepresented groups in the sample are weighted less in the analysis, while responses from respondents of underrepresented groups are weighted more
- this balances the statistical significance and the overrepresentation/underrepresentation disparity
- weights can change the actual numbers and can become heavy (to the point where they may subtract from the poll’s accuracy), so unweighted and weighted data can be provided
sampling time frame
the time from when the researcher started collecting data to when they stopped
cross-sectional data
allows conclusions about voters’ attitudes at one point in time
longitudinal data
allows conclusions about how voters’ attitudes evolve over time
