AP Statistics (Sampling and Surveys)

Cluster

A grouping within a population that is diverse (mini-populations).

Cluster Sample

A method of random selection where diverse groups (mini-populations) are numbered and then selected. Each member of the group(s) chosen are sampled.

Convenience Sample

A method of sampling in which individuals that are easy to access or close by are chosen. This is a bad sampling technique that is not random and should be avoided as it is prone to bias.

Voluntary Response Sample

A method of sampling in which the respondents choose themselves. This is a bad sampling technique that is not random and should be avoided as it is prone to bias.

Stratified Random Sample

A method that first separates the population into groups of similar characteristics then takes a SRS from each of those groups.

Simple Random Sample (SRS)

A method that gives every group of a certain size (n) an equal chance of being selected.

Undercoverage

A sampling error that occurs when entire groups of the population are left out of the sample.

Random Sampling Error

A type of sampling error that occurs through the chance process involved in taking a sample. SRS can sometime produce these when they create samples that over represent some individuals by a large degree.

Random Sampling

Any method that uses chance to select a sample from a population.

Sampling Errors

Errors that come from the act of choosing the sample itself. Undercoverage is a common example of this.

Nonsampling Error

Errors that have nothing to do with the method of sampling. Can be present even in a census (collection of entire population data).

Strata

Groups of individuals that all have a common characteristic or trait.

Response Bias

Incorrect answers from respondents of a survey. Often intentional because they are aware they are being observed or recorded.

Sample Survey

Selects a sample from a population.

Bias

Systematically favoring an outcome.

Sampling Frame

The list of individuals in the population that we are actually sampling from.

Nonresponse

The single biggest problem (type of nonsampling error) that occurs in sample surveys. This is when individuals chosen/contacted for a sample refuse to answer.

Wording Bias

This is when the way a prompt is presented causes answers to shift toward an outcome.