Score aggregation: aggregation period

Peakon's aggregation settings control the visibility period of the submitted survey feedback over time for current employees, dictating how far the system should go when searching for the latest score per question, per employee. See Score aggregation: inclusion of former employees for the equivalent setting for leavers.

To learn about aggregation in general, see Score aggregation resources.

This article will cover:

• How it works
• Configuration

How it works

This setting sets a global visibility window for scores to be considered. Peakon calculates the end-point of the window automatically. This is not measured from today, instead this is set by the round close date of the most recent response, and is checked each survey round.

Peakon's recommendation is setting this to 1 year, to increase the amount of data points that make up each driver's current score when using the question rotation algorithm (see Understanding the question algorithm).

Example: Aggregation period is set to 1 year

During every survey round, the system will recalculate based on the 1 year visibility window and ensure that:

• Every single driver and question is made up of each employee's latest score to that driver or question from the last 1 year
• The participation rate contains employees who have been surveyed at least once in the last 1 year

Let's imagine it is now December and below are the different times a specific employee has answered the Autonomy driver questions.

First, the system goes back 1 year from the latest survey and throws out all data from outside of this window. 

From these, it only keeps the latest score on each question. If the latest score for a specific question is outside the 1 year visibility window, it does not use any score from that employee.

In the above example, for this employee the overall Autonomy driver score is the average of the scores she provided in September, October and November. The scores are also individually contributing to each of these questions.

Example: Aggregation period is set to 3 months (12 weeks exactly)

During every survey round, the system will recalculate based on the 3 month visibility window and ensure that:

• Every single driver and question is made up of each employee's latest score to that driver or question from the last 3 months 
• The participation rate contains employees who have been surveyed at least once in the last 3 months

Using the previous example, imagine it is now the 15th of December. With this decreased visibility window, the system only uses the latest score if it is from the last 3 months (12 weeks exactly).

This employee's overall Autonomy driver score will be the average of her October and November scores, because at this point the September score is outside the 12 week visibility window.

Increasing the survey frequency to weekly could increase the amount of data when using a very short visibility period. 

Configuration

System administrators can manage the aggregation settings in the administration panel. As with all other visibility settings, they are effective immediately and will affect the entire account.

To set the Aggregation period:

1. Go to Administration > Data settings
2. Click on the Data tab
3. Set the Aggregation period setting 

Once this is set, scores and participation will instantly be reflective of the chosen visibility period.