Home / DOE / Create Designs / Create Design

Create Design¶

QXL DOE New > Create Design > design type

Seven of the design types on the Create Design menu open the same builder dialog. The design type decides which pages it shows and in what order; everything else about the dialog is the same.

Three entries on that menu do not use this dialog at all. Create Custom Design (and Setup Historical Analysis, which is the same command), Create D-Optimal Design and Import From DOE PRO XL each open a dialog of their own. See Create Custom Design, Create Optimal Design and Import From DOE PRO XL.

Every design type is built in the same dialog. It shows one page at a time, names the current page in a heading at the top, and moves with Next > and < Back. Finish accepts the defaults for every page you have not reached yet.

The pages, by design type¶

Read the row for the design you are building. The names in the table are the headings the dialog puts at the top of each page.

Design type	Pages, in order
2-Level Factorial	Select 2-level factorial design. > Enter Factor Names and Coding > Define outputs (responses) > Enter number of replicates and number of blocks
3-Level Factorial	Enter factor names and factor levels > Define outputs (responses) > Enter number of replicates and number of blocks
N-Level Factorial	Enter factor names and factor levels > Define outputs (responses) > Enter number of replicates and number of blocks
CCD	Select the base design > Enter Factor Names and Coding > Define outputs (responses) > Enter number of replicates and number of blocks > Axial and center points
Box-Behnken	Select the Box-Behnken design > Enter Factor Names and Coding > Define outputs (responses) > Enter number of replicates and number of blocks > Center points
Plackett-Burman	Select Plackett-Burman design > Enter Factor Names and Coding > Define outputs (responses) > Enter number of replicates and number of blocks
Taguchi	Select Taguchi design > Factors > Define outputs (responses) > Enter number of replicates and number of blocks

Custom Design is not in the table because it is not built here. Its own dialog, titled Create Custom Design, has two pages, the factor and run counts and then the outputs, with Next becoming Finish on the second. That dialog shows no page headings at all, so it has nothing to contribute to this table. See Create Custom Design.

Three things follow from the table:

Only CCD and Box-Behnken ask about center points while the design is being built. For a 2-level factorial, center points are set on the replicates and blocks page instead. For every other type they are added afterwards, and for some types they cannot be added at all. See Modify Design.
The 3-level and N-level factorial designs have no separate factor page. Their first page does both jobs: it sets the number of factors and their names and levels together.
Only the CCD reaches the axial value, and its page is last in the sequence rather than first. The alpha values it offers are computed from the factor count, the number of factorial runs and the number of center points. Blocking and replication do not enter that calculation.

Factor names and coding¶

Enter Factor Names and Coding takes one row per factor. What the grid holds depends on the design type:

Design type	The grid
2-level factorial, Plackett-Burman	Name, Low (or Name), High (or Name) and Categorical
CCD, Box-Behnken	Name, Low and High. No Categorical column and no categorical hint line, so every factor is quantitative

The rest of this section describes the first case. The two value columns carry both readings in their headings because they hold different things depending on the factor:

Factor	What goes in the two value cells
quantitative, Categorical clear	the low value in one and the high value in the other
categorical, Categorical ticked	one level name in each cell, so the factor has exactly two levels

The page states the rule on screen as well, in two hint lines: For Quantitative Inputs enter the low and high values (e.g. low=10 high=20) and For Categorical Inputs enter the name of each level (e.g. Red,Blue). That last example means Red in one cell and Blue in the other. Nothing splits a cell on commas, so typing Red,Blue into one cell gives one level whose name is Red,Blue.

A factor with three or more levels cannot be described here, since there are only two cells. Those designs use their own factor page, with a # Levels column and one Level n column per level.

Full details, including what coding does to the numbers: Factor Names and Coding.

Define outputs¶

Define outputs (responses) asks for the number of outputs in Select number of outputs:, a list of 1 to 20, and then shows one block per output with Name:, Type: and a checkbox. Type: offers Quantitative, Binary and Nominal. The checkbox reads Has weights column for a quantitative or nominal output and Events / Sample size for a binary one, and it is available only when the design format is Stacked.

The same page carries the Design format choice, Stacked or Table, which decides how replicates are laid out on the sheet. While Table is selected the page shows a note beside the format choice: Weights and sample size columns are only available with stacked replicates. The note is hidden while Stacked is selected.

Full details: Define outputs.

Replicates and blocks¶

Enter number of replicates and number of blocks sets how many times the run block is repeated and whether the runs are divided into blocks. The two controls are labelled Select number of replicates: and Select number of blocks:. The blocking list offers No blocking first, then one entry per possible block count, each marked either (block in design), meaning the blocks are generated from the design and confounded with a chosen term, or (block with replicate), meaning the replicates are divided among the blocks.

On the design types that support it, the page also carries Help Me Choose Power/Sample size for Quantitative Outputs, which opens the power calculation with the design's current run count already filled in.

Full details: Define Replicates and Blocks.

What you get¶

Finish writes a new worksheet called DOE Design, or DOE Design (2), DOE Design (3) and so on when a sheet of that name already exists: one row per run, one column per factor, the columns for you to enter the responses into, and a description of the design in text. Nothing is calculated until you enter response data and run the regression.

How many columns an output gets depends on the replicate format, and it is rarely one:

Format	Columns per output
Stacked	one data column, plus the weights or events and sample size columns if you ticked them
Table, quantitative output	one column per replicate, plus Y-bar, plus S when there is more than one replicate. The per-replicate columns are headed Y1, Y2 and so on, or block1, block2 and so on when the design is blocked with replicate into more than one block. Y-bar and S arrive carrying `=AVERAGE` and `=STDEV` formulas, so they fill themselves in as you type. With a single replicate it is two columns
Table, binary or nominal output	one column per replicate, and never fewer than two

Whether the rows come out in standard order or shuffled is set once for all designs, by Run order for new design on the Create Design Settings page of the Options dialog. Randomize Run Order on the ribbon shuffles an existing design sheet afterwards.