Exclusion Rule Filtering

Every listening event has a minimum response threshold. If a group has fewer responses than that number, its results stay hidden so nobody can be singled out. That works well on its own, but there's a gap: if you can see the total for a department and you can see the results for everyone in it except one small group, simple math gets you the small group's results anyway.

Exclusion rule filtering closes that gap. Along with checking that a group is big enough to show, it also checks that whatever is left over is big enough too. When the leftover piece is too small, the rule hides one more group so the numbers can't be backed out.

You'll end up with a few fewer breakouts in your reports. That's the trade, and it's what keeps the people answering your survey protected.

How the Rule Works

The rule boils down to one idea: if what's left over after a filter or breakout is too small to show on its own, something else gets hidden too.

It shows up in three places in Advanced Reporting:

• Filter Data, when you narrow a report down to a specific group
• The Demographic Crosstab Report, when you break results out by a demographic like region or tenure
• The manager hierarchy, which follows the same rules in both of the places above, and in other breakout views like the Hot Spot Report, dashboards, and exports

A few things the rule always does:

• It counts blanks. People who left a demographic empty still count toward the leftover group, because their results could be worked out just as easily as anyone else's.
• It hides the smallest group it can. When the rule needs to hide one more group, it picks the smallest one that's at or above the threshold, so you lose as little as possible. If two groups are tied, it goes alphabetically by name.
• It respects your filters. If you've filtered a report, the rule runs on the filtered numbers, not the full population.

Exclusion Rule Filtering in the Filter Data Tool

When you apply a filter, the groups you picked need to meet the threshold, and so does everyone you didn't pick. If the unselected remainder is below the threshold, the filter can't be applied.

Here's why. Say a team has 6 responses: 5 men and 1 woman. Without the rule, you could filter to the 5 men, compare their results to the team overall, and work out exactly how that one woman answered.

Example 1: Gender

It isn't always that obvious. In the next two examples, the filtered groups are well above the threshold. The only difference is what gets left behind.

Example 2: Region, filtering on one group

Example 3: Region, filtering on two groups

Exclusion Rule Filtering in the Demographic Crosstab Report

The Demographic Crosstab Report shows a column for each group in a demographic. Any group below the threshold is hidden, which you'd expect. The rule then adds up all the hidden groups, including blanks. If that total is also below the threshold, the smallest group that does meet the threshold gets hidden as well.

That's why you'll sometimes see a column missing even though it clearly has enough responses.

Example 4: Region

When only one group would be left

Some events are set up to show nothing for a demographic if only one group would make it into the crosstab. A single column on its own is basically the overall result again, and next to the overall it can give the hidden groups away.

In Example 4, West is the only group left standing. With that setting on, West is hidden too, and no breakouts are available for the region demographic.

Example 4b: Region, with the one-group setting on

Exclusion Rule Filtering in the Manager Hierarchy

The manager hierarchy works the same way, whether you're filtering or looking at a breakout. The rule checks every manager against the managers who report to them, at every level of the organization.

For each manager, it adds up two things:

• Any managers below them who are under the threshold
• The manager's own direct team, the people who report straight to them rather than through another manager, if that team is under the threshold

If that total is more than zero but below the threshold, the smallest manager at or above the threshold under that same parent is hidden too.

Example 5

Example 6

Example 7

Choosing a Privacy Setup

There are two ways to set the rule up, depending on how much protection you need.

Base setup

This stops anyone from working out a small group's results inside a single report. People can still combine filters freely, including picking small groups as part of a bigger selection. It's a good fit for most organizations.

Strict setup

This does everything the base setup does, and also stops people from piecing results together across several reports with different filters. Small groups, and groups the rule has hidden, can't be picked in the filter at all, so there's no way to fold them into a different combination and compare.

Strict is worth a look if your organization works under tighter data privacy requirements, like rules set by a works council or regional privacy law.

Turning On Exclusion Rule Filtering

• Perceptyx-led events: reach out to your Perceptyx team and they'll set it up for you.
• Self-led events: platform administrators can turn it on from the Admin Portal, under Listening Events and then System Settings. Work with System Settings (Platform Admins Only) walks through that screen.

Want to know more about how confidentiality works in general? Understanding Confidential vs. Anonymous Listening Events covers the difference.