This dialog defines the design information and meta data to generate a design. The design is required to fill a rectangular grid with no missing plots, and all varieties must be equally replicated. Thus the design has the requirement that the number of rows times the number of columns must equal the number of varieties times the number of replicates. If the number of entries in the trial does not allow a particular layout, you can often repeat some control entries to give a full grid.
The maximum number of rows, columns or treatments is 4000, with at most 8000 plots in the array, and the maximum number of replicates is 20. There must be at least two rows, columns and treatments, and at least three replicates in the design. The number of columns must be less than or equal to the number of treatments. The number of rows can be greater than the number of treatments, but it must then be a multiple of the number of treatments. Multiple replicates are stacked in the columns.
Trial Id
A trial code used to identify the trial. This would normally have letters to code the series, species, location and year for the trial. It must be unique as it is used as the default filename for the trial.
Description
A one line description of the trial.
Units name
This gives the name of the factor indexing the units in the trial, and is Plot by default.
Columns
Provides the number of columns in the trial. The columns are numbered from 1 at the left of the page. The factor defining the columns is always named Col.
Rows
Provides the number of rows in the trial. The rows are numbered from 1 at the top of the page downwards. The factor defining the rows is always named Row.
Replicates
Provides the factor name and number of replicates in the Name and No fields.
Treatments
Provides the factor name and number of treatments in the Name and No fields.
Treatment names
Click this button to open the Edit Factor Levels and Labels dialog to allow the treatment names to be entered. The Number of treatments must be entered before using this button. The treatment names can be pasted into this dialog from another source.
Starting date
The date when the trial was started/sown. This should be entered in the format d-mmm-yy, e.g. 12-Feb-23. Initially it will be the date that the trial was generated, but should be changed later to the date when the trial was started.
Plot size
The area of a plot in square metres.
Sowing rate
The sowing rate of seeds.
Column types
Click this button to open the Edit Column Types dialog to allow the column types to be defined. These can be edited later by using the Trials | Edit | Column types menu. The column types can be used to define columns whose repeated measures are to be summarized over seasons, or to define columns that are to be calculated automatically from other column types. For example, dry matter yield could be calculated automatically as the product of green yield and dry matter percentage.
Design
This specifies the type of design generated.
| Randomized block | Randomized block design using generated using the AGFACTORIAL procedure |
| Lattice | A square lattice design generated using the AGSQLATTICE procedure |
| Lattice square | A lattice square design generated using the AGSQLATTICE procedure |
| Latin square | Latin square design generated using the AGLATIN procedure |
| Flexible row-column design | A doubly resolvable row-column design generated using the AFRCRESOLVABLE procedure |
For the Flexible row-column design there is a Search time option (default 60 seconds) that specifies the maximum time in seconds to spend searching for an optimal design. For large designs, the time should be increased. For example, 1000 seconds is recommended for more than 100 treatments, and 4000 seconds for more than 200 treatments.
For the lattice designs the number of treatments must be a square of an integer (e.g. 4, 9, 16, 25, 36, 49 …). For a Latin design the number of replicates must equal the number of treatments, so this is useful only for a small number of treatments.
Randomization seed
This gives a seed to initialize the random number generator used for the design algorithms. Using zero initializes this from the computer’s clock, but specifying an non-zero value gives a repeatable design.
First plot
This option specifies where the numbering of units should start, as follows:
| Top left | Left-hand plot at the top of the design, i.e. in the first row and first column (default) |
| Top right | Right-hand plot at the top of the design, i.e. in the first row and final column |
| Bottom left | Left-hand plot at the bottom of the design, i.e. in the final row and first column |
| Bottom right | Right-hand plot at the bottom of the design, i.e. in the final row and final column |
Plot order
This option defines the order the plots are numbered from the first plot:
| Across rows | The plots are numbered within rows, column by column |
| Down columns | The plots are numbered within columns, row by row |
Plot numbering
This option defines the how the plots are numbered as you move to a new row or column plot:
| One side | The plots are numbered within rows or columns always from the same side of the design |
| Serpentine | The next row or column is from the end of the last numbered row or column, so the numbering zig-zags up and then down or left and then right |
Replicates
This option defines how the orientation of the replicates in the trial:
| Across trial | The replicates run across the trial, so the replicates are horizontal groups of plots |
| Down trial | The replicates run down the trial, so the replicates are vertical groups of plots |
Number of seasons
This option defines the number of seasons to which the repeated measurements across years are allocated:
| 4 | There are 4 seasons (default names Winter, Spring, Summer and Autumn) |
| 5 | There are 5 seasons (default names Winter, Early Spring, Late Spring, Summer and Autumn) |
The seasons used in the trial can later be edited using the Trials | Edit | Seasons menu which opens the Define Seasons dialog.
Comments
A multi-line text that records any comments about the trial
Action buttons
| Ok | Run the analysis. |
| Cancel | Close the window without further changes. |
See also
- Introduction to the Trials system
- Trials Menu
- Edit Factor Levels and Labels dialog
- Edit Column Types
- Convert Sheet to Trial
- Estimate Missing Values
- Define Seasons
- Add Seasonal Summaries
- Add/Edit Trial Column Information
- Add Calculated Column
- Sort Trial
- Analyse Trial Data
- Multiple Trial Analysis
- Trial Seasonal Analysis
- Select Entries for Summaries
- Change Units
- Generate a Doubly Resolvable Row-Column Designs menu
- Generate a Standard Design menu
- AFRCRESOLVABLE procedure
- AGFACTORIAL directive
- AGLATIN directive
- AGSQLATTICE directive