Reduce Redundant Testing with an optional Random Selection Preview Resource

RandomWare's Random Selection Preview helps reduce redundant testing. The entire personnel roster is shuffled (random sort) and displayed in a new window with everyone's most recent random selection date displayed. Personnel names are not displayed to insure against biased outcomes.

For example, if you plan to select 5 people and you use the preview as the basis for a random selection, then the first five people represented in the previewed list will be the ones selected. Names are not displayed in the preview.

TPAs will find a similar resource to preview random selections from a consortium of companies. The shuffled list shows each person's respective company but not their name. This resource can help limit excessive testing of a single consortium member.

The preview gives users a chance to review a random sample before it's committed as an official random selection.  This random selection method is completely unbiased since the names of the participants are not displayed.  The preview displays the last selection date for everyone in the list so you can reduce redundant testing.

If a random sort represents a reasonable distribution of previously selected personnel, you can generate a random selection of 'n' people from the list (n=1,2,3,4,...) (identically the first 'n' people represented in the preview).

Read the example below to understand how the Random Selection Software Preview Works
When the preview is requested the personnel list is randomly sorted and displayed in a separate window.  Personnel names are not displayed, although the date of each person's most recent random selection is shown.  
For example, the person that corresponds to the first ID Card was last randomly selected on January 12, 2004 (01-12-2004).  The person that corresponds to the second ID Card has never been randomly selected or was last selected on or before a date you can specify - this is covered later.  The person that corresponds to the third ID Card was last selected April 16, 2004 (04-16-2004).   

Since names are not displayed the random selection preview is completely unbiased.

 
   
When almost everyone in a list has a previous random select date displayed, it's difficult to determine if a random sort represents an outcome that reduces redundancy.

For example, if you're generating a random selection for the second quarter you may want to suppress the display of dates for everyone selected last year.

Enter 12-31-2003 and click the Apply Date Restriction button:

If someone's last selection date is on or before the date specified, their date will display as a blank date: "../../...."

To resort, click the Re-Sort button.

If you know you're going to generate a random selection of 10 people from the list, simply click the Resort option until the first ten items show previous random selection dates that satisfy your requirements.

This resource may not completely eliminate redundant random selections, however, you can use it to help reduce redundancy. 

When the outcome of a preview sort is satisfactory, choose a random selection option from the menu. The random selection control you choose will duplicate the random selection as displayed in the preview. The first "n" people selected will correspond to the first "n" people whose records are represented by the ID cards in the preview.