AP Statistics Unit 3 Test: Collecting Data

What Unit 3 Covers in AP Statistics

Unit 3 is about how data are gathered — and why the method of data collection determines what conclusions can legitimately be drawn. This unit is conceptually rich and frequently appears in AP Statistics FRQs that ask students to design a study or critique an existing one.

Observational Studies vs Experiments

In an observational study, researchers record data without imposing any treatment. In a well-designed experiment, researchers actively assign subjects to treatments. Only a properly designed experiment with random assignment allows researchers to draw cause-and-effect conclusions. Observational studies can reveal associations but cannot establish causation — this distinction is fundamental to AP Statistics reasoning.

Sampling Methods

AP Statistics covers four key probability-based sampling methods:

• Simple Random Sample (SRS): Every group of n individuals from the population has an equal chance of being selected.
• Stratified Random Sample: The population is divided into homogeneous groups (strata), and an SRS is taken from each stratum. Reduces variability compared to SRS when strata are internally similar.
• Cluster Sample: The population is divided into groups (clusters), and entire clusters are randomly selected. More practical when the population is geographically spread out.
• Systematic Sample: Every k-th individual on a list is selected after a random starting point.

Sources of Bias

AP Statistics identifies several forms of bias that undermine the validity of a sample: undercoverage (certain groups are excluded from the sampling frame), nonresponse bias (selected individuals do not respond), response bias (question wording or interviewer presence influences answers), and voluntary response bias (self-selected samples overrepresent strong opinions).

Experimental Design Principles

A well-designed experiment incorporates three core principles: random assignment of subjects to treatment groups (to create groups that are roughly equivalent at the start), replication (enough subjects in each group for results to be meaningful), and control (holding other variables constant or accounting for them). A placebo and blinding are additional features used to manage confounding and bias in experiments.

Key AP FRQ Skills for Study Design

• Distinguishing between an observational study and an experiment and explaining what conclusions each allows
• Describing a completely randomized design or a randomized block design with enough detail for someone to carry it out
• Identifying and explaining the type of bias present in a described sampling method
• Explaining why random assignment — not random selection — is what enables causal conclusions

Common Study Design FRQ Errors

Confusing Random Selection with Random Assignment

Random selection from a population allows generalization of results. Random assignment of subjects to treatments allows causal conclusions. These are different concepts with different implications, and AP exam questions frequently test whether students can distinguish between them.

Vague Experimental Descriptions

FRQ answers that say 'randomly assign subjects to groups' without explaining how the randomization is carried out — for example, using a random number generator, drawing names from a hat, or assigning based on randomly generated labels — typically do not receive full credit.