AT Stats Chapter 11 (Sampling Bias & Sampling Methods)

Undercoverage Bias

happens when some portion of the population is not sampled at all or has smaller representation than the sample should

Voluntary Response Bias

happens when people choose to put themselves in a sample by responding to a general appeal such as a write in, call in, or internet poll (people who respond usually have really strong opinions, negative or positive)

Non Response Bias

happens when people who are chosen for a sample do not respond because they are unwilling to participate or cannot be contacted (people who do not respond usually differ in some important way from people who do respond)

Response Bias

happens when anything in the survey design influences the responses (ie. the wording of the question, the interviewer, illegal/socially undesirable behavior)

You cannot have ___ bias and ___ bias at the same time

Voluntary, Non Response

Simple Random Sampling (SRS)

every individual in the population has an equal chance of being selected AND every combination of individuals has an equal chance of being selected (drawing from a hat, random number generator, wheel, etc.)

Stratified Random Sampling

first divide the population into groups of similar individuals called strat, then take a random sample within each stratum

• strata are internally homogenous but externally heterogeneous

• every individual has an equal chance of being selected but every combination does NOT

• Advantage = guarantees fair representation from all groups, reduces variability

Cluster Sampling

first divide the population into representative clusters, then select one or a few clusters at random and perform a census within in each selected cluster

• clusters and internally heterogeneous and each resembles the overall population

• every individual has an equal chance of being selected but every combination does NOT

• Advantage = less time consuming, less expensive, easier

• Disadvantage = hard to find clusters that are truly representative of the population

Systematic Sampling

randomly select any individual to start with, then select every (N/n)th individual after that (N = population size, n = sample size)

Convenience Sampling

BAD SAMPLING METHOD, undercoverage bias

Voluntary Sampling

BAD SAMPLING METHOD, voluntary response bias