Relative Standard Error

SuperSERVER uses Data Control plugins to calculate RSE for weighted datasets.

Relative Standard Error Calculation Module

The RSE Calculation Module calculates the Relative Standard Error for display in SuperWEB2.

RSE Calculation Module Now Runs By Default

RSE is only available for weighted databases.

From version 9.7 onwards the RSE Calculation Module runs automatically when a database is configured to use SuperSERVER weightings. While you can still configure the module if you wish to change other RSE settings, or have legacy configuration from an earlier release, in most cases this will not be necessary:

• If you want to change the RSE formula, then you can do so by changing a setting in the weightings file. If both are configured, then the setting configured for the method in SuperADMIN takes precedence.

• If you want to configure the unreliable cell and table threshold, and message, you are recommended to use the RSE Poor Table Check module instead.

Other RSE Modules

A number of other RSE modules are supplied and can be applied to your data if necessary:

Example Configuration

Following is an example of configuring some of the RSE modules to work together. Refer to the module reference for more details on the available configuration settings.

1. Create a new method:

   CODE

   > method addmethod ExampleRSEMethod

   In this example the method ID has been set to ExampleRSEMethod.

2. Add a Data Control plugin to the method:

   CODE

   > method ExampleRSEMethod adddcplugin RSEPoorTableCheck rse_poor_table_check 1

   This example adds the Poor Table Check module, with its ID set to RSEPoorTableCheck and a priority of 1 (meaning it will be executed first before any other modules configured in this method).

3. Set the plugin properties. For example:

   CODE

   > method ExampleRSEMethod RSEPoorTableCheck addproperty UnreliableRseThreshold "0.75"
   > method ExampleRSEMethod RSEPoorTableCheck addproperty UnreliableTableThreshold "0.8"
   > method ExampleRSEMethod RSEPoorTableCheck addproperty UnreliableMessageSuccess "This table has passed the reliability check."
   > method ExampleRSEMethod RSEPoorTableCheck addproperty UnreliableMessageFail "This table is not considered reliable."

   This example sets the thresholds to 0.75 and 0.8 and defines the messages to display if the table passes or fails the poor table check.

4. Add a second Data Control plugin to the method:

   CODE

   > method ExampleRSEMethod adddcplugin RSESuppression rse_suppression 2

   This example adds the RSE Suppression module, with an ID of RSESuppression and a priority of 2 (meaning it will be executed after the poor table check).

5. Set the plugin properties. For example:

   CODE

   > method ExampleRSEMethod RSESuppression addproperty suppression_threshold "0.8"

   In this case the threshold for suppressing RSE has been set to 0.8.

6. Assign the method to a dataset:

   CODE

   > cat weightedsurvey2016 addmethod ExampleRSEMethod