You can use the Simulator to simulate random noise in responses. Often your factors explain part of the variation in the response with the rest of the variation of the response attributed to random noise. You can specify the random noise using the following options:
No Noise
Generates the response from the model based on the factor settings. No random noise added to the response.
Add Random Noise
Adds a normal random value with mean zero and specified standard deviation to the generated response. If the response column contains a Predicting column property that includes a Std Dev value, the field used to specify the standard deviation in the Simulator is automatically populated with the value from the column property.
Add Random Weighted Noise
Adds an error term to the generated response based on a weighted sampling scheme used to simulate rare events from the tails of the distribution. If the response column contains a Predicting column property that includes a Std Dev value, the field used to specify the standard deviation in the Simulator is automatically populated with the value from the column property.
Add Multivariate Noise
Adds an error term to the generated response based on a multivariate normal distribution. Specify the standard deviation for the response. Specify the correlation structure between responses in the Y Correlation Specification report. If the response column contains a Predicting column property that includes a Std Dev value, the field used to specify the standard deviation in the Simulator is automatically populated with the value from the column property.
Random by Model
(Available only within the Standard Least Squares and Generalized Regression personalities of the Fit Model platform.) Adds an error term to the generated response based on the model distribution specified in the Fit Model launch window.
Tip: To specify the same noise setting for all of the responses, press Ctrl and select the desired noise setting.