Split Plot Design
A split plot design with randomised complete blocks. Main plot treatments are randomised within each block, and sub-plot treatments are randomised within each main plot. This two-stage randomisation reflects the hierarchical experimental structure.
Design Reference — what this design is, parameters explained, output guide
Split Plot Design
A two-stage randomised design: main plot treatments are randomised within each block, then sub-plot treatments are randomised within each main plot. This reflects a hierarchical experimental structure where one factor is harder to change than another.
Use when: One treatment factor is applied to large units (main plots) and another is applied to subdivisions of those units (sub-plots). Common in field experiments where tillage or irrigation is applied to whole plots, while varieties or fertiliser rates are applied to sub-plots.
Parameters
| Parameter | Type | Default | Min | Max | Description |
|---|---|---|---|---|---|
experiment_name |
text | "" |
— | — | Name of the experiment |
block_factor |
text | "Block" |
— | — | Label for the blocking factor |
block_count |
integer | 2 |
1 | 100 | Number of blocks |
main_unit_label |
text | "Main plot" |
— | — | Label for the main plot experimental unit |
main_treatment_factor |
text | "MainTreat" |
— | — | Label for the main plot treatment factor |
main_treatment_count |
integer | 2 |
2 | 100 | Number of main plot treatments |
sub_unit_label |
text | "Sub-plot" |
— | — | Label for the sub-plot unit |
sub_treatment_factor |
text | "SubTreat" |
— | — | Label for the sub-plot treatment factor |
sub_treatment_count |
integer | 2 |
2 | 100 | Number of sub-plot treatments |
main_treatment_names |
list | None |
— | — | Optional custom main plot treatment names |
sub_treatment_names |
list | None |
— | — | Optional custom sub-plot treatment names |
seed |
integer | 0 |
— | — | Random seed |
Constraints
- Total units =
block_count × main_treatment_count × sub_treatment_countmust not exceed 5,000
Output
List View
| Column | Description |
|---|---|
| Unit | Sequential unit number |
| block_factor | Block this unit belongs to |
| main_unit_label | Main plot number within the block |
| main_treatment_factor | Main plot treatment |
| sub_unit_label | Sub-plot number within the main plot |
| sub_treatment_factor | Sub-plot treatment |
Layout View
One section per block: - Columns = main plots (and their treatment) - Rows = sub-plots within each main plot (and their treatment)
Total units: block_count × main_treatment_count × sub_treatment_count
Key Statistical Consideration
In a split-plot design there are two error strata:
- Main plot error — used to test main plot treatment effects (less precise, because main plots are larger)
- Sub-plot error — used to test sub-plot treatment effects and the interaction (more precise, because sub-plots are smaller)
This means sub-plot treatment comparisons and the interaction between main plot and sub-plot treatments are estimated more precisely than main plot treatment comparisons alone.
Algorithm
- For each block:
- Randomly assign main plot treatments to main plots
- For each main plot, randomly assign sub-plot treatments to sub-plots
- Assign sequential unit numbers across all blocks and main plots