Experiment

First taught in High school statistics

Definition

In statistics, an experiment is a study where the researcher deliberately applies a treatment to subjects and measures the effect. In probability, an experiment is any action with uncertain results, like rolling a die.

QuestionAssign groupsat randomApplytreatmentMeasureand compare
A well-designed experiment assigns subjects to groups at random.
Advertisement

In probability: each repetition of an experiment is a trial, and its possible results form the sample space. Example: flipping a coin is an experiment with outcomes heads and tails.

In statistics: a good experiment has

  • a treatment group and a control group for comparison;
  • random assignment of subjects to the groups;
  • enough subjects, so that chance differences average out (replication).

Example: To test a new study app, 60 students are randomly assigned to use the app or not. After a month, both groups take the same test.

Unlike an observational study, an experiment can show causation, because the researcher controls who gets the treatment.

Advertisement