AP Stats 2.1-2.4

samples and biases

Simple Random Sample (SRS)

1) label everyone in the chosen population by assigning numbers or putting names in a hat

2) randomly generate unique numbers within range, or choose names from hat

3) select the corresponding individuals

Voluntary Response

can lead to bias; ex: sending a survey that people can choose to respond to

Convenience Samples

can lead to bias; ex: asking the first 50 people who buy something

Stratified Random Sample

1) split the population into homogeneous groups (strata)

2) take an SRS of each strata

Good Estimates

low bias and low variability

Cluster Random Sample

1) heterogeneous, mixed groups

2) choose an SRS of the groups

3) survey everyone included in the selected groups

Systematic Random Sample

1) SRS

2) use equal intervals until the desired number of participants is reached

Undercoverage

when some people are less likely to be chosen in a sample; ex: calling landlines

Nonresponse

people can’t be reached or refuse to answer; ex: someone doesn’t answer a phone call or hangs up

Response Bias

problems in the data gathering instrument or process; ex: wording of a question, people lying