Control Group
The control group is the set of subjects or samples in an experiment that does not receive the treatment, providing a baseline for comparison.
Definition
The control group is the part of an experiment that does not get the special treatment; it shows what happens normally so you have something to compare against. It experiences all the same conditions as the experimental group except the independent variable, which is left at its normal or zero level, and its results establish the baseline. In rigorous designs the control group receives no treatment (or a placebo or standard treatment) while being handled identically in timing, environment, and measurement, so the treatment effect is the difference between groups. Random assignment to control and experimental groups guards against selection bias.
Example
In a fertilizer test, the pot that gets plain water is the control group; if those plants grow $6\ \text{cm}$ and the fertilized plants grow $9\ \text{cm}$, the fertilizer made a $3\ \text{cm}$ difference. Testing whether salt lowers water's freezing point, the control is a cup of $100\ \text{mL}$ pure water in the same freezer, which freezes at $0^\circ\text{C}$. In a 12-seedling-per-group study, the control group's mean height after 21 days is $10.9\ \text{cm}$ with a range of $2.1\ \text{cm}$, and the fertilized mean of $14.6\ \text{cm}$ sits well outside that spread.
Key Insight
Without a control group you cannot tell whether your treatment did anything, because plants grow on their own too. The control group answers "compared to what?" and every other result is measured against it. It also catches problems with the experiment itself: if the control plants die, something in the shared conditions failed, and the treatment data cannot be trusted either.