AP®︎ Statistics (pre-2026): Unit 3 Flashcards

The extent to which results from a sample can be applied to the population.

Choosing individuals who are easiest to reach rather than selecting by chance.

A sample made of people who choose themselves to participate.

Distortion caused when participants decide for themselves whether to join a study.

All individuals or items of interest in a statistical study.

A subset of the population from which data are actually collected.

A study that records variables without imposing treatments on individuals.

Retrospective examines existing past data; prospective follows subjects forward while collecting future data.

An observational study that gathers data from a sample to learn about its population.

A specific experimental condition imposed on subjects or experimental units.

The individuals or items to which treatments are assigned in an experiment.

Association is a relationship between variables; causation means changes in one directly produce changes in another.

When two variables' effects on a response are mixed so their separate influences cannot be distinguished.

The population to which study results may be applied, limited to the population sampled.

A sample that accurately reflects the important characteristics of the population of interest.

With replacement allows repeats; without replacement allows each population member to be chosen only once.

A sample where every possible group of a given size has an equal chance.

Label all individuals, then use random numbers to choose the required distinct labels.

Stratified samples some from every homogeneous group; cluster samples all from randomly chosen heterogeneous groups.

Choose a random starting point, then select every kth individual from the sampling frame.

The fixed spacing between selected individuals in a systematic sampling process.

The list of individuals or units from which a sample is actually selected.

Each method balances representativeness, cost, efficiency, practicality, and possible bias differently.

Divides a population into homogeneous groups, then takes an SRS from each group.

Divides a population into heterogeneous groups, randomly selects groups, then samples everyone in them.

Using a chance process to choose members of a population for a sample so each population member has a known opportunity to be selected.

A study or data collection method that includes every individual in the population.

Bias caused when some population members have a smaller chance of being selected.

Bias caused when selected individuals do not respond and differ from those who do.

Bias caused by flaws in the survey instrument or data collection process.

Bias caused by confusing, leading, or ambiguous wording that influences responses.

Methods without chance selection, such as convenience or voluntary response, often produce unrepresentative samples.

A systematic tendency for a study or data collection process to favor certain outcomes or responses and miss the truth.

Individuals or objects assigned to treatments in an experiment; people are called subjects or participants.

The manipulated variable in an experiment; its levels or level combinations are the treatments.

The outcome measured on experimental units after treatments are applied.

A variable related to the treatment that also affects the response, obscuring the true effect.

Units receiving no treatment of interest or an inactive treatment for baseline comparison.

An inactive treatment; the resulting response to it is the placebo effect.

Single-blind: subjects or evaluators unaware; double-blind: neither subjects nor interacting evaluators aware.

Use a random number generator, random digits table, or drawing chips without replacement.

Uses comparison, random assignment, replication, and control to reduce bias and confounding.

A study that imposes treatments on subjects or experimental units to measure resulting responses.

Label all units, then use a random method to place them into treatment groups.

A group of similar units, or the characteristic used to create those similar groups.

Every block must receive all treatments so treatment comparisons are made within each block.

Choose the simplest appropriate design: randomize if units are similar, block known differences, and match pairs for two treatments.

An experimental design in which all experimental units are assigned to treatments entirely by chance, without forming groups first.

A blocked experimental design for two treatments that uses matched pairs of similar units or one subject receiving both treatments.

An experimental design that groups similar units into blocks based on a relevant characteristic, then randomly assigns treatments within each block.

Using sample or experimental data to draw conclusions about a population or distribution.

Different random samples from the same population produce different statistic values by chance.

The difference between a sample statistic and the true population parameter.

An observed difference so large that random chance alone is unlikely to explain it.

A causal conclusion is justified when a well-designed experiment uses random assignment.

Random selection supports generalization; random assignment supports cause-and-effect conclusions.

The extent to which study results can reasonably be applied to a larger population, especially when the sample is representative.

A sample chosen by chance so each population member has a known selection probability, increasing the likelihood that the sample is representative of the population.

Using chance to assign subjects or experimental units to treatment groups in an experiment.