Light

Completely Randomized Design

Definition

Completely randomized design is an experimental design where subjects or items are randomly assigned to different treatment groups. It helps eliminate bias and ensures that any differences observed between groups are due to the treatments applied.

Analogy

Imagine you're testing different fertilizers on plants in your garden. You label each plant with a number and then use dice to assign them randomly to different fertilizer treatments. This way, each plant has an equal chance of receiving any fertilizer, making it fair like completely randomized design.

Related terms

Matched Pairs Design: Matched pairs design is an experimental design where subjects are paired up based on similar characteristics, and one subject receives one treatment while the other receives another treatment.

Factorial Design: Factorial design is an experimental design where multiple factors (variables) are studied simultaneously by combining different levels of each factor.

Latin Square Design: Latin square design is an experimental design where subjects or items are arranged in a square grid, and each treatment appears once in each row and column.