In AP Statistics, a control group is the group in an experiment that receives no treatment (or a placebo or standard treatment) so researchers have a baseline to compare against, letting them measure whether the treatment actually caused a change in the response variable.
A control group is your baseline. In a well-designed experiment, subjects are randomly assigned to groups, and the control group is the one that does NOT get the treatment being tested. It might get nothing, a placebo, or the current standard treatment. Everything else about the control group should be identical to the treatment group, so any difference in the response variable can be attributed to the treatment itself.
Here's the intuition. If you give 50 people a new sleep supplement and they average 7 hours of sleep, you've learned almost nothing. Seven hours compared to what? The control group answers that question. Without a baseline, you can't separate the treatment's effect from things that would have happened anyway, like the placebo effect, natural improvement over time, or just normal variability. The control group turns 'people who took the pill slept 7 hours' into 'people who took the pill slept 45 minutes more than comparable people who didn't.' That comparison is the whole point of an experiment.
Control groups live in Unit 3 (Collecting Data), where you learn the principles of experimental design: comparison, random assignment, control of other variables, and replication. The control group is how you achieve 'comparison.' But the concept follows you all the way to Unit 8. Topic 8.7 (Skills Focus: Selecting an Appropriate Inference Procedure for Categorical Data) asks you to pick the right test for data that often comes from exactly this setup, like comparing the proportion of successes in a treatment group versus a control group. Design and inference are two halves of the same story. Random assignment to treatment and control groups is also what earns you the right to make a causal conclusion, which is one of the most heavily tested scope-of-conclusions ideas on the entire exam.
Keep studying AP Statistics Unit 8
Experimental Group (Unit 3)
The experimental (treatment) group and the control group are two sides of one comparison. The treatment group gets the intervention, the control group doesn't, and the difference between their responses is your estimate of the treatment effect.
Randomization (Unit 3)
A control group only works if subjects are randomly assigned to it. Random assignment makes the two groups roughly equivalent before the treatment starts, so the control group is a fair baseline rather than a biased one. The 2017 FRQ tested exactly this, asking about randomly assigning two men and two women to treatment and control groups.
Confounding Variables (Unit 3)
The control group plus random assignment is your defense against confounding. If the groups differ only by the treatment, a lurking variable can't sneak in and explain the difference for you.
Null Hypothesis (Units 6-8)
When you run inference on experimental data, the null hypothesis usually says the treatment and control groups have the same true mean or proportion. The control group is literally built into H₀. Topic 8.7 asks you to choose the right categorical-data procedure for comparisons like this.
Placebo Effect (Unit 3)
People often improve just because they believe they're being treated. Giving the control group a placebo means both groups experience that psychological boost equally, so it cancels out of the comparison.
Control groups show up constantly in experimental design FRQs. The 2017 FRQ Q6 asked about randomly assigning subjects to a treatment group or a control group. The 2019 FRQ Q2 involved testing fungus concentrations against a control to see what actually kills tree-destroying insects. The 2021 FRQ Q2 set up a year-long experiment on walking and cholesterol. In these questions you're asked to do things like describe a method for random assignment (and a coin flip or random number generator must give every subject an equal shot at each group), explain WHY a control group is included, or evaluate whether a causal conclusion is justified. In multiple choice, expect stems asking you to identify the purpose of the control group or spot a design flaw when one is missing. The magic phrase to have ready is that the control group 'provides a baseline for comparison so the effect of the treatment can be measured,' and that random assignment to groups is what permits a cause-and-effect conclusion.
A placebo is a fake treatment (like a sugar pill); a control group is a group of subjects. The control group often receives the placebo, but not always. Sometimes the control group gets the current standard treatment, or nothing at all. The placebo exists to keep subjects blind to their group; the control group exists to give you a baseline for comparison. You can have a control group without a placebo, but a placebo is pointless without a comparison group to give it to.
A control group is the baseline group in an experiment that does not receive the treatment being tested, allowing researchers to measure the treatment's effect by comparison.
Subjects must be randomly assigned to the control and treatment groups, because random assignment is what makes the groups comparable and what justifies a cause-and-effect conclusion.
The control group can receive nothing, a placebo, or the existing standard treatment; using a placebo specifically controls for the placebo effect.
Without a control group, you cannot tell whether a change in the response variable came from the treatment, the placebo effect, or something that would have happened anyway.
Data from treatment-versus-control experiments often becomes the input for inference procedures, like the two-sample comparisons covered in Topic 8.7 for categorical data.
A control group is the group in an experiment that does not receive the treatment being studied. It serves as a baseline, so any difference in the response between the treatment and control groups can be attributed to the treatment.
Not necessarily, and this is a common trap. An experiment needs comparison between groups, but those groups can be two different treatments (like the four fungus concentrations in the 2019 FRQ, which also included a control). A no-treatment control group is one way to create comparison, not the only way.
The control group is a set of subjects; the placebo is the fake treatment those subjects might receive. A control group taking a sugar pill is receiving a placebo, but a control group could also receive the standard treatment or nothing at all.
No. The control group is the comparison group of subjects, while 'controlling variables' means keeping other conditions (like time of day or dosage schedule) identical across all groups. Both reduce confounding, but they're separate design principles.
Random assignment spreads lurking variables roughly evenly between groups, so the control and treatment groups start out comparable. The 2017 FRQ Q6 tested this directly, requiring a random assignment method that gives every subject an equal chance of landing in either group.