Overview
This lecture explains three main types of experimental designs: completely randomized design, randomized block design, and matched pairs design, using practical examples for each.
Completely Randomized Design
- Each experimental unit is randomly assigned to a group, and each group receives a different treatment.
- All units in a group get the same treatment; results from each treatment group are compared at the end.
- Example: 30 students randomly divided into 3 groups (library, own room, outside); each group has 10 students.
Randomized Block Design
- Experimental units are first divided into blocks based on a characteristic expected to affect the outcome (blocking variable).
- Units within each block are then randomly assigned to groups to receive different treatments.
- Example: 30 students split into blocks by gender (18 females, 12 males), then each block divided into 3 groups for treatments.
Matched Pairs Design
- Used for comparing two treatments using the same or similar experimental units.
- If using the same units, each receives both treatments in random order; results are compared for each unit.
- Example: Three cars each tested with gasoline type A and B, order randomized.
- If using similar units, pair up units with similar traits, randomly assign one of each pair to each treatment.
- Example: Students paired by similar GPA; within each pair, one is sleep-deprived and one has normal sleep, then compare test scores.
Key Terms & Definitions
- Experimental Unit — The individual or object to which a treatment is applied.
- Treatment — A specific condition applied to experimental units.
- Blocking Variable — A characteristic used to create groups (blocks) expected to affect responses.
- Matched Pairs Design — An experimental design comparing two treatments using either the same or closely matched experimental units.
Action Items / Next Steps
- Review the diagrams for each experimental design in your textbook or lecture materials.
- Practice creating experimental designs for different research scenarios.