Consensus & statistics · Administrator
Feedback Review Rate (FRR)
Understand Surveylet’s live Feedback Review Rate, Reviewed Panelists, review cycles, and how these differ from consensus, COUNT, REVISIONS, and historical statistics.
What Feedback Review Rate means
Feedback Review Rate (FRR) is a live indication of participation in the latest feedback-review cycle for a question. It increases as respondents review the latest group results and keep their existing selections and comments unchanged.
It answers a different question from consensus: Have respondents revisited the current feedback without making further changes? The former Approval Index / Approval Rate labels referred to this same feature. FRR is not a vote approving the group’s conclusion, and it is not a measure of agreement.
How the current review cycle works
- A respondent changes an existing selection or comment. Surveylet starts a new review cycle and counts that respondent as the first participant in the cycle.
- Other respondents view the current question results. Each qualifying respondent who leaves their existing selection and comment unchanged is counted toward the current cycle, once per cycle.
- When another selection or comment is revised, the count for the new cycle starts again. Previous reviews do not carry forward as confirmation of the newer results.
- As additional respondents answer the question, the eligible respondent population can change, affecting the percentage.
Example: If the applicable respondent population is 50, each respondent counted in the cycle represents 2 percentage points. One counted respondent is 2%; 40 are 80%; 50 are 100%. With 51 applicable respondents, each one represents about 1.96 percentage points.
Calculation concept: FRR is the share of the applicable respondent population counted in the current feedback-review cycle. Do not assume a displayed Responses count always equals the FRR calculation denominator without checking the question’s response and eligibility rules.
Where to find FRR in Surveylet
| Location | Column | Meaning |
|---|---|---|
| Download Results → CURRENT: Consensus | Reviewed Panelists | Number of respondents counted in the current review cycle for the reported question or matrix cell. |
| Download Results → CURRENT: Consensus | Feedback Review Rate | Current percentage of the applicable respondent population counted in that review cycle. |
| View Responses | Feedback Review (%) | Respondent-level feedback-review summary in the selected round. Do not confuse this with the question-level FRR or the separate Progress column. |
Select the intended round before comparing results. The two views serve different purposes: CURRENT: Consensus is question-level; View Responses is organized by individual questionnaire.
Interpret FRR with COUNT and REVISIONS
Always interpret the live FRR alongside the size of the responding panel and revision activity. Surveylet already provides the COUNT and REVISIONS statistics; FRR does not replace them.
| Measure | What it shows |
|---|---|
| FRR | Current feedback-review coverage since the latest change. |
| COUNT / Responses | Number of submitted responses for the question, as defined by that statistic or report. This supplies essential participation context. |
| REVISIONS | Revisions recorded for the question, covering selection changes and comment changes together. |
| Consensus / Stability | Agreement and method-dependent strength or stability measures; they do not establish that the newest feedback has been reviewed. |
Interpretation example: 100% FRR with 2 respondents describes far less panel coverage than 95% FRR with 100 respondents. A high FRR can coexist with disagreement because respondents may examine the same feedback and retain different judgments. A low FRR may reflect a very recent change rather than disengagement.
FRR is current-state information, not a historical statistic
FRR is displayed as a current value only. It is not an additional selectable statistic in Survey Settings → Statistics or an FRR history series in Analyze Survey → Statistics.
Surveylet’s historical statistics are stored against an event_id when a new answer or a selection/comment revision is recorded. A review that makes no response change does not generate that event. FRR can increase without a new historical statistics event, so the existing event history cannot reconstruct a complete FRR time series.
If the research protocol requires change-over-time analysis, use the historical COUNT and REVISIONS measures where enabled and available, and describe the live FRR separately. A current FRR value should not be presented as a retrospective trend.
Deleting respondents is a separate issue. If a respondent with saved answers is removed, a new statistics event is not created; current and historical question-level results may need a full support-managed recalculation. See Deleting responses and recalculating statistics. This does not create historical FRR tracking.
Common questions
Does a 100% Feedback Review Rate mean everyone agrees?
No. It means the current cycle has reached the applicable respondent population without a further revision restarting the cycle. It does not mean everyone chose the same answer or approved the majority opinion.
Why did the Feedback Review Rate drop?
A selection or comment revision starts a new review cycle. Changes to the respondent population can also affect the percentage. Check the question’s activity and response counts before interpreting a lower rate.
Why does Feedback Review (%) in View Responses differ from a question’s FRR?
View Responses is organized by respondent and provides a review-activity summary for the selected round. CURRENT: Consensus reports review counts and percentages by question or matrix cell. The two percentages are not interchangeable.
Can I plot historical FRR in Analyze Survey?
No. Reviews without a changed selection or comment do not create a new statistics event, so FRR is shown as a current-state measure, not a historical statistic.