Determining Normal Distribution of Sample Data

You can use Normality Tests to test whether a given set of observations is a sample from a normal distribution.

Graphical and statistical tests are available within the Normality tests node. You can also perform the Box-Cox transformation, which transforms the selected variables so that they are as close to normal as possible. The transformed variables are then available as a downstream document.

Prerequisites

You must add a source dataset file to the workspace to run the node successfully.

Procedure

  1. Open a workspace.
  2. To add a node to the workspace, go to Node Browser > Nodes > Import New Node...
  3. In the Open dialog box, browse to C:\Program Files\Statistica\Statistica 13\DataMiner\NormalityTests.dmi and click Open.
    The Normality tests node is displayed in the workspace.
  4. To open the Normality tests dialog box, double-click the node.
  5. Choose from any one of the available options to compute the test of normality.
  6. Select the Box-Cox transformation check box to specify values to transform the original variable so that the new variable has a distribution that is as close to normality as possible.
  7. After specifying the Normality test and Box-Cox transformation required, select a type of output on the Output tab and click OK.
  8. To view the Reporting Documents, run the node.