The Moderation within Adaptive Surveys empowers leaders with both manual and automated tools to oversee and manage the quality of participant feedback before it reaches their reports.
Moderation is performed manually and/or via machine.
Machine Moderation
ThoughtExchange’s machine moderation is enabled by default. You can manage these settings on the Manage page under the Moderate sub-section.
When active, open-ended responses are automatically reviewed as they are posted. This process is asynchronous, so please allow a few seconds to a few minutes for the review to complete.
There are two types of machine moderation:
-
Rude & Hurtful Language: This filter identifies responses that may be inflammatory, rude, or hurtful. The goal is to maintain a constructive environment and ensure that the quality of your engagement’s analysis and reporting remains high.
Important Note on Participant Wellbeing: This category may also flag comments that indicate self-harm. We recognize the sensitivity of these situations; ThoughtExchange is actively working on specialized functionality to better find and identify these responses.
- Personal Identifying Information (PII) This filter flags responses that contain private details (such as names, phone numbers, or addresses) to protect participant anonymity and data privacy.
Responses flagged by these filters are given a "Reported" status. These remain visible in your results until a moderator reviews them and chooses to either Hide or Display them in the Moderation page.
System Limitation: Only open-ended responses submitted after March 12, 2026, are eligible for machine moderation. Responses submitted before this date must be moderated manually.
Manual Moderation
Accessible via the moderate icon on the Results page, the Moderation page provides a comprehensive view of all text-based responses.
- Actionable Controls: Leaders can choose to Hide or Report individual responses, or perform these actions in bulk.
- Reporting vs. Hiding: A "Reported" response remains visible in the survey data until a leader takes further action, whereas a "Hidden" response is immediately removed from all summaries, detailed responses, transcripts, and exported documents.
- Advanced Sorting & Filtering: The dashboard includes a search bar and robust filters based on status (Reported, Hidden, Displayed), specific survey questions, and date (Newest to Oldest).
Impact of Moderating Responses
When a response is hidden, it is effectively removed from the "active" data pool to ensure your results accurately reflect the desired tone and scope of the conversation.
Visibility & Reporting
Once hidden, the response will no longer appear in:
- The "List" view on the Detailed Responses page.
- The participant’s specific entry on the Transcripts page.
- Any documents or reports generated after the moderation action has taken place.
AI Analysis & Summarization
To maintain the accuracy of your insights, hidden responses are excluded from all subsequent AI-generated content, including:
- The Summary component on the Summaries page.
- The survey question’s Summary, Categories and Common Ground components on the Detailed Responses page.
A Note on Versioning: Please be aware that AI-generated insights are not automatically retrofitted. If a summary was generated before a response was hidden, that response may still be reflected in the text.
We recommend evaluating whether the moderated content significantly alters the core themes of your discussion. If the moderation changes the "heart" of the conversation, refreshing your AI summaries will ensure your final reporting is both accurate and aligned with your moderation standards.