2.42.1.1 The Power and Sample Size (DOE) Dialog Box

Contents

Supporting Information

To open the Power and Sample Size dialog

Topics for Further Reading:


Factorial Type

Select a factorial type from the list below.

Design

2-Level Factorial

Number of Factors The number of factors. Together with the fraction, this determines the base number of corner-point runs.
Fraction The fraction of the full factorial to run.
Number of Blocks Blocks group runs performed under homogeneous conditions. Default is 1 (unblocked design).
Number of Terms Omitted from Model Terms deliberately left out of the fitted model; reducing model terms increases the available degrees of freedom for error and raises power. Default is 0.
Include Blocks in Model When ON, block effects are estimated and removed from the error term, protecting the analysis if block-to-block variation exists.
Include Term for Center Points in Model When ON, Origin fits a center-point (curvature) term, which slightly reduces power for factorial effects but allows a test for curvature.

Plackett-Burman

Number of Factors The number of factors.
Number of Corner Points The size of the Plackett-Burman design - must be a multiple of 4 and, as the note states, **no less than 12** (12 is the smallest standard Plackett-Burman design; the next sizes are 16, 20, 24, ...).
Number of Main Effect Terms Omitted from Model Main effects deliberately excluded from the fitted model (default 0). Omitting terms frees degrees of freedom for error, which raises power
Include Term for Center Points in Model When ON (as shown), Origin fits a center-point term so you can test for curvature across factors.

General Full Factorial

Number of Factors The number of factors.
Number of Levels for Each Factor One text box per factor - enter each factor's level count individually (e.g., 3, 2, 4 for a 3*2*4 design). The total base run count is the product of these values (3 * 2 * 4 = 24 runs before replication).
Model Order The highest-order interaction included in the model: 1 = main effects only, 2 = main effects + two-factor interactions, 3 = up to three-factor interactions, and so on.
Block on Replicates and Include Blocks in Model When ON, each replicate of the design is treated as a block, and block effects are estimated and removed from the error term.

Settings

Calculate Select the quantity to compute:
  • Replicates
  • Effects
  • Power
  • Number of Center Points per Block
Replicates Number of repetitions of the entire design. Disabled when it is the calculated quantity.
Effects The minimum effect size to detect, in the response units of the measurement (can be a list of values).
Power Desired power for detecting the specified effect(s); values between 0 and 1. Common targets: 0.8 to 0.9 (can be a list).
Number of Center Points per Block Center-point runs added per block to estimate pure error and test for curvature.
Standard Deviation The estimate of experimental noise (sigma), from historical data
Significance Level Alpha - the risk of declaring an inactive effect significant. Default 0.05.

Dialog Theme

Save and load the settings of the dialog.