Experimental Unit

In AP Statistics, an experimental unit is the individual (a person, animal, plant, or object) to which a treatment is assigned in an experiment. When the units are people, they're called participants or subjects. The unit is whatever gets the treatment, not necessarily whatever gets measured.

Verified for the 2027 AP Statistics examLast updated June 2026

What is Experimental Unit?

An experimental unit is the smallest thing in an experiment that could receive its own treatment. If treatments get assigned to people, the people are the units. If treatments get assigned to whole classrooms, fields, or driveways, then those bigger things are the units, even if you end up measuring the individual students, corn plants, or cracks inside them.

That distinction is the whole game. Ask one question: what got randomly assigned to a treatment? Whatever answers that question is your experimental unit. A researcher might spray one of three irrigation methods on entire fields of corn. You measure every corn plant, but you can't give two plants in the same field different irrigation methods, so the field is the experimental unit. The CED (under learning objective 3.5.A) defines units as the individuals assigned treatments, and that wording is precise on purpose.

Why Experimental Unit matters in AP Statistics

Experimental units live in Topic 3.5 (Introduction to Experimental Design) in Unit 3: Collecting Data. They're the first component you have to identify under 3.5.A, and every other element of a well-designed experiment under 3.5.B is defined in terms of them. Random assignment means assigning treatments to experimental units. Replication means having more than one experimental unit in each treatment group. If you misidentify the unit, you'll get randomization and replication wrong too, and that's exactly the chain of errors the AP exam is built to catch. Units also matter later for scope of inference. Random assignment of treatments to units is what lets you claim cause and effect, which comes back in significance tests in Units 6 and 7.

How Experimental Unit connects across the course

Treatment (Unit 3)

Treatments and experimental units are two halves of one definition. A treatment is the level of the explanatory variable you impose, and the experimental unit is the thing you impose it on. You can't correctly identify one without the other.

Random Assignment (Unit 3)

Random assignment happens at the level of the experimental unit. In a completely randomized design, you might draw chips or use a random number generator to decide which unit gets which treatment. If classrooms are randomized instead of students, your units are classrooms, period.

Replication (Unit 3)

Replication means more than one experimental unit per treatment group. Here's why unit identification matters so much. Thirty students in one classroom assigned to one teaching method is NOT 30 replicates. It's one unit, so there's no replication for that treatment.

Randomized Block Design (Unit 3)

Blocking means grouping similar experimental units before randomizing within each group. The 2022 FRQ used pairs of identical twins as blocks, with each twin as a unit. Matched pairs is just blocking with units that come in twos.

Is Experimental Unit on the AP Statistics exam?

Question 2 on the AP Stats free-response section is traditionally the data collection question, and experimental units show up there constantly. The 2019 FRQ (fungus concentrations on trees), 2022 FRQ (acne drug on identical twins), 2023 FRQ (fiber concrete on driveways), and 2026 FRQ (coffee grounds on rosebushes) all required you to work with units when describing or evaluating a design. You're typically asked to identify the units, explain how to randomly assign treatments to them, or critique a flawed design.

Multiple-choice questions love the mismatch trap. A stem describes an experiment where each classroom of 30 students gets one teaching method, then asks what the experimental units are. The answer is the classrooms, because that's what received the treatments. Another common stem applies irrigation methods to whole fields of corn and asks the same thing (answer: the fields). When you write FRQ answers, name the units explicitly and describe randomization at the unit level, like "assign each rosebush a number and use a random number generator to select bushes for the coffee-grounds treatment."

Experimental Unit vs The individuals you measure

The experimental unit is what receives the treatment, which is not always what you measure. If three irrigation methods are applied to three fields, you measure individual corn plants but the fields are the units, because two plants in the same field can never receive different treatments. Same with classrooms versus students. Always ask "what was assigned a treatment?" not "what produced a number?"

Key things to remember about Experimental Unit

  • An experimental unit is the individual (person, animal, plant, or object) that is assigned a treatment, and human units are called participants or subjects.

  • The unit is the smallest thing that could receive its own treatment, so if whole classrooms or fields are assigned treatments, the classroom or field is the unit, not the students or plants inside it.

  • Random assignment and replication are both defined in terms of experimental units, so misidentifying the unit usually breaks your whole design description.

  • A treatment group with only one experimental unit has no replication, even if that one unit contains many individuals you can measure.

  • On FRQ 2, name the experimental units explicitly and describe how treatments are randomly assigned to those units, like numbering the units and using a random number generator.

Frequently asked questions about Experimental Unit

What is an experimental unit in AP Stats?

It's the individual that gets assigned a treatment in an experiment, defined in Topic 3.5 of Unit 3. Units can be people (called participants or subjects), animals, plants, or objects like driveways and fields.

Is the experimental unit always a person?

No. Units are whatever receives the treatment. Released FRQs have used trees (2019), identical twins (2022), concrete driveways (2023), and rosebushes (2026) as experimental units.

How is an experimental unit different from the individuals being measured?

The unit is what's assigned a treatment; the measurement can happen at a smaller level. If a teaching method is assigned to a whole classroom of 30 students, the classroom is the experimental unit even though each student takes the test.

If 30 students are in one classroom that gets a treatment, is that replication?

No. Replication requires more than one experimental unit per treatment group, and that classroom counts as a single unit. You'd need multiple classrooms per teaching method to have replication.

What's the difference between an experimental unit and a subject?

A subject (or participant) is just the name for an experimental unit when the units are people. The CED uses 'experimental unit' as the general term covering people, animals, plants, and objects.