Skip to main content
Go to the Anonymity tab ↗ · Opens the webapp — please log in first · Requires admin rights In this article you will learn more about how to adjust the anonymity for open text feedback.
As you know, anonymity is very important to us, generally you create an anonymity level which applies to your survey, in the standard case this is always 5, but now it may be the case that you survey 5 employees, but only 2 of them answered the open text field, so an evaluation would not be possible.
With our newly developed solution, you can still evaluate them.
Navigate to your account settings and select the tab “Anonymity”.
In the upper part you will see the settings for your surveys, in the lower part you can set the settings for your open text answers.
Screenshot 2023-07-20 at 10.53.07 Example 1
  • Manager with 1 team member
  • Text anonymity threshold = 1
  • Survey anonymity threshold = 5
Without filtering this manager would see the response from this one employee although other charts do not display any data. Since the above rule is in place and the total number of participating employees is less than 5, the text threshold of 1 is not applied and the text is not visible. Example 2
  • Manager with 5 team member
  • Text anonymity threshold = 1
  • Survey anonymity threshold = 5
Manager receives in total 5 responses but only one textual response, then this response would still show up (text anonymity threshold applies, because survey anonymity threshold as a prerequisite is given). Warning! If you have not set any setting before starting the survey, this setting is not possible afterwards!

Why open-text answers may still be missing

  • Too few answers for that question. Both thresholds have to be met: the survey’s Anonymity threshold on the total number of responses, and the Text response anonymity threshold on that question. Results appear once the count is at least the threshold, not once it exceeds it.
  • The question was skipped. If the open question is not mandatory, respondents can pass over it, and skipped answers do not count towards the threshold.
  • The filter would single out a small group. When you filter by attributes such as country or department, Honestly checks not only your selection but also the people the filter leaves out. The rules are explained under Open-text answers with attribute filters.
  • The survey has no dashboard. If the dashboard was deleted, there is nothing to display the data until you create a new one. No responses are lost — see What is a dashboard and how do I create a new one?
The Topics area is not a way around this. It applies exactly the same two thresholds as the dashboard, so it hides exactly the same answers.

Open-text answers with attribute filters

Without these rules you could compare two filters, say “Germany” and “Germany + Austria”, and isolate the answers of a group that is too small to be shown on its own. So a filtered open-text chart only shows answers when every one of the following rules is met. What counts is the number of people who answered the open question, not everyone who took part. The threshold is the survey’s Text response anonymity value. Each open question is checked on its own, so one question can show answers while another does not.
  • Your selection reaches the threshold. The filtered group needs at least as many people with a text answer as the threshold requires.
  • The group the filter leaves out reaches it too. The people your filter excludes must either be nobody at all or also reach the threshold.
  • Small values only count together. For each attribute, Honestly combines every value that does not reach the threshold on its own into one combined group. If that combined group is still below the threshold, the smallest remaining value is added to it. You can then select single values that reach the threshold on their own, and the combined group only as a whole, meaning all of its values together.
  • People without a value form a group of their own. If some people have no value for the attribute and that group is below the threshold, none of the values in the combined group can be selected.
If you combine several attributes, each attribute is checked within the other filters. If you filter by a department and then by country, for example, only the people in that department count towards the countries. An example with a Text response anonymity of 3: 10 people from Germany answered an open question, 4 from France, 2 from Austria and 1 from Switzerland. Austria and Switzerland form the combined group, with 3 people. If a rule is not met, the open-text chart shows the notice Selected filters do not satisfy anonymity constraints. Try to widen filters. Select fewer attributes or combine small values. The same rules apply to the Bubble Chart, including its summaries, and to the Topic Trend Chart, but there the answers are left out without this notice.
The filter bar still offers every value. It is based on everyone who took part in the survey, not on who answered an open question. So a value can be selectable even though the open-text chart shows nothing for it.

What the raw data export does and does not show

A raw data export from the Exports section lists one row per submission, without any respondent identity. It ignores the per-question Text response anonymity threshold, so open-text answers can appear there that the dashboard leaves out.
The export never bypasses the survey’s Anonymity threshold. If the survey has fewer responses than that threshold, the export is created but comes out empty.
Anonymity and data masking are separate settings. The anonymity threshold decides whether an open response is shown; data masking removes personal data from inside a response that is shown. See Data masking.