| Scenario | Invited | Started | Screened Out | Completed | Eligible Attrition | Wave (Baseline→Follow-up) | Wave Attrition |
|---|---|---|---|---|---|---|---|
| Consumer pulse | 2,000 | 1,250 | 50 | 1,050 | 12.50% | — | — |
| Product UX test | 600 | 420 | 10 | 310 | 25.64% | — | — |
| Panel tracking study | — | — | — | — | — | 1,000 → 780 | 22.00% |
- Eligible starts = Started − Screened out
- Retention (eligible) = Completed ÷ Eligible starts × 100
- Attrition (eligible) = (Eligible starts − Completed) ÷ Eligible starts × 100
- Overall drop-off = (Invited − Completed) ÷ Invited × 100
- Wave-to-wave attrition = (Eligible baseline − Follow-up) ÷ Eligible baseline × 100
Confidence intervals use the Wilson method on the retention proportion, then converted to attrition by subtracting from 100.
- Choose Study mode (single survey or longitudinal).
- Enter the relevant counts from your dashboard or exports.
- Set an expected and minimum retention to flag risks (single survey mode).
- Click Calculate to see results above the form.
- Use Download CSV or Download PDF to share a clean summary.
Attrition definitions that match your denominator
Attrition is not a single number. This calculator reports eligible attrition (eligible starts minus completes, divided by eligible starts) and overall drop-off (invited minus completes, divided by invited). Use overall drop-off for outreach efficiency, but use eligible attrition for questionnaire experience. When quotas and screening are heavy, eligible attrition isolates true breakoff behavior instead of counting ineligible respondents as “dropouts.”
Using eligible starts to isolate breakoffs
Eligible starts equal started minus screened out. If 780 start and 30 screen out, eligible starts are 750. With 620 completes, eligible retention is 82.67% and eligible attrition is 17.33%. Recording partials helps separate abandonments from usable partial interviews. For instance, 70 partials imply that 9.33% of eligible starters produced partial data and 8.00% fully broke off, which changes how you plan recontacts.
Benchmarking retention with targets and thresholds
Set an expected retention to compare performance against plan and document assumptions. A deviation of −5.00 percentage points often signals friction such as long duration, confusing grids, or mobile formatting issues. The minimum retention threshold provides an operational alert for pausing invites, adjusting incentives, or simplifying routing. Pair these targets with a consistent time window, such as “first 48 hours,” so comparisons are internally fair.
Interpreting confidence intervals in fieldwork reporting
Because retention is a proportion, this tool uses Wilson intervals for more stable uncertainty bounds than the normal approximation, especially with smaller n or extreme rates. At 95% confidence, it converts the retention interval into an attrition interval by subtracting limits from 100. If your eligible starts are 80 and completes are 60, the point retention is 75%, but the interval may still be wide; report both to avoid overreacting to noise.
Operational levers that reduce attrition
Attrition typically concentrates at early screens, long open-ends, and late attention checks. Shorten median completion time, place sensitive items later, and add progress indicators. Use validations and save-and-return links for panelists. Segment results by device, region, and incentive to find where breakoffs spike. After a change, compare eligible attrition to the prior confidence interval to validate that the improvement is real.
1) What is the difference between attrition and overall drop-off?
Attrition focuses on eligible starters who abandon before completion; overall drop-off compares completes to invitations. Use attrition to improve survey design, and use drop-off to evaluate outreach efficiency.
2) How should I treat screened-out respondents?
Remove screened-out cases from the eligible denominator when assessing questionnaire breakoffs. Keep them visible for operational reporting, because high screen-out rates can indicate targeting or quota issues.
3) Do partial completes count as retention?
That depends on your analysis rules. If partial interviews are usable, treat them separately and report both ‘completes’ and ‘partials.’ This calculator shows breakoffs with and without partials so teams can agree on a standard.
4) Why does the calculator show a confidence interval?
The interval quantifies sampling uncertainty in the retention proportion. Wilson intervals behave well with small samples and extreme rates, giving more realistic bounds than simple normal approximations.
5) Which mode should I choose for panel studies?
Use longitudinal mode when the same respondents are invited to multiple waves. It computes wave-to-wave retention from an eligible baseline, and lets you account for known ineligible removals and optional replenishment.
6) How can I reduce attrition quickly?
Shorten survey length, simplify mobile layouts, move sensitive items later, and add progress feedback. Monitor attrition by device and entry source, then re-test after changes to confirm improvement.