BIO Stat Wk 8

Spatial or temporal subdivisions within an experiment

For an experiment that includes physical blocks, a within-block factor is a factor for which all levels are represented within each block, such that the factor is crossed with "block"

This is an experimental design with physical blocks, a single within-block factor, and no replication for each factor-block combination. Experimental units must be completely randomized within each block.

• 1. Increase sample size
• 2. Eliminate as many confounding variable as possible
• 3. Re-randomize frequently over spatial domain
• 4. Include a confounding variable as a blocking variable
• 5. Balanced design

A mixed-model two-way design with no replication within groups