Fieldwork data acceptance workshop / 6 MIN READ
Decide what happens to unfinished surveys before closing the field
Closing recruitment and freezing the analysis dataset are different boundaries.
AI-assisted research and analysis for the SNS Creative Fieldbook
Sources checked 2026-11-07
Survey field close must account for responses still in progress and the configured treatment of unfinished work. Establish what is saved, when pending responses become recorded or disappear, and which resulting records belong in the planned extract. Do not let an unnoticed default decide the evidence boundary.
What you will leave with: A field-close disposition matrix linking in-progress states, closure rules and the analysis freeze.
Use this when
- An online survey is approaching its fieldwork deadline.
- A recorded-response count changes after new survey starts have ended.
This concerns field-close disposition and dataset membership, not save-and-resume interface copy or statistical missing-data analysis. Qualtrics documentation checked on 7 November 2026. Closure, configuration changes and deletion require explicit authorization; this lesson describes preparation and acceptance only.
Separate the fieldwork clocks
Identify the deadline for new starts, the rule governing already-started responses and the analysis extraction cutoff. These need not coincide. State which decision each date supports and who owns it. A fieldwork announcement that no new invitations are being sent does not establish that all existing sessions have reached their final disposition.
Inventory the relevant unfinished states using permitted aggregate counts and restricted references where needed. Distinguish work with saved answers from an empty start, and completed responses from records merely recorded after partial closure. Keep these distinctions visible in the plan without exposing respondent answers in a scheduling document. The research owner should know what evidence may still arrive before approving a final analysis freeze.
Read the actual disposition settings and their history
Qualtrics distinguishes availability for new starts from the completion window for existing responses. Automatic survey closure can end in-progress sessions at the availability deadline. Its incomplete-response rules can record or delete unfinished work, and deadline timing and closure action have different update behavior. Review the deployed configuration rather than assuming one closing date governs everything.
Have the authorized platform owner record the relevant settings and effective history. Check whether a planned change would affect existing sessions as expected, and identify any irreversible consequence before seeking the necessary approval. Do not “clean up” unfinished records simply because they complicate the report. The treatment must follow the approved study plan, participant arrangements and retention requirements, with uncertainty resolved before action.
Test the resulting evidence boundary
Prepare synthetic cases that stop at different meaningful points, including before and after a normal saved page boundary. Specify what should remain in progress, what should become recorded and which answers should be available after the approved test closure. Inspect the actual output rather than treating a progress percentage as proof that particular answers were preserved.
Compare the observed disposition with the intended analysis inclusion rule. A recorded partial response is not automatically a completed questionnaire, and a usable answer to an early item does not establish a usable answer to a later one. Keep the technical acceptance narrow: it proves what the platform retains and exposes under the tested configuration, while the researcher decides which evidence can answer each study question.
Freeze and communicate the final population deliberately
Before extraction, reconcile completed, recorded-partial, still-pending and otherwise disposed states under the approved rule. Record the cutoff and explain whether any authorized later completions will create a revised dataset. Avoid quietly adding late arrivals to one chart while leaving other outputs on the earlier population. A stable release needs one explicit evidence boundary.
Hand the analyst the disposition record and material exceptions without turning it into an unsupported estimate of nonresponse bias. If the planned closure would destroy evidence needed for the study, pause and ask the authorized owner to resolve the conflict before acting. The outcome is a deliberate field-close handover, not a claim that closing a survey automatically produces a complete or unbiased analysis population.
Make the reasoning concrete.
A fictional study of business onboarding plans to stop new starts on Friday and issue its analysis on Monday. Several respondents have saved early answers but still hold active completion windows that extend beyond the recruitment deadline.
Freeze from the recruitment date alone
Opening: Assume Friday ends every response and label Monday’s download final.
Hypothesis: The calendar has one convenient boundary.
Risk: Later completions or partial dispositions can change the evidence set.
Reconcile the closure states
Opening: Document the existing windows, intended partial treatment and extraction rule before approving the freeze.
Hypothesis: The analysis population has an explicit operational boundary.
Risk: Any later update must be released consistently across outputs.
The team chooses a disposition-based close and does not delete unfinished work merely to simplify the count.
Take the questions into your next brief.
Use these prompts in your own notes. Nothing is submitted from this page.
Clocks
What are the new-start, existing-session and analysis deadlines?
Rules
Which effective settings govern unfinished-response disposition?
Fixtures
What survives each tested stopping and closure point?
Inclusion
Which completed and partial records belong in the extract?
Freeze
How will authorized late changes be communicated?
Watch for the wrong conclusion.
- Equating the end of new starts with the end of all sessions.
- Treating every recorded response as complete.
- Changing or deleting pending evidence to simplify reporting.
Your next action
Prepare a field-close matrix for three synthetic stopping points and reconcile the expected output with the planned analysis cutoff.
Sources & scope
Workflows, worksheets and fictional examples are original editorial analysis. Sources support only specifically attributed principles. Platform guidance can change; legal requirements depend on jurisdiction and context. No measured results or named human reviewer are claimed.
- Qualtrics: Incomplete Survey Responses →Primary survey-platform documentation · Supports: Supports separate new-start and partial-response windows and configuration-dependent closure. The field-close acceptance method is original analysis.