Field reports

The field report shows how respondents are progressing through a survey, where interviews stop, and how results differ by demographic group, quota, source, device and browser. Its sections share the same respondent base and report controls.

Opening the field report

Open the survey's Summary tab and find Respondent Sources.

1. For one source, select Report on that source's row. The dialog title names the respondent source.
2. For the survey's live sources together, select Field report in the Respondent Sources section header. The dialog says All live respondent sources. This includes sources in the Live dataset, rather than test or simulated sources.

Report actions depend on your access and whether a source has reporting available. Before any respondents start, the report explains that it will become available once responses begin.

Overview

Overview shows progress toward the target, participation, incidence, completion rate, respondent statuses, daily completes and the distribution of interview lengths.

Use these definitions when comparing sources:

Measure	Definition
Starts	Respondents with recorded survey activity and a valid start timestamp in the selected period. Metadata and status records alone do not establish a start. In a quota-term row, starts means respondents evaluated for that term.
Completes	Respondents whose latest source status is Complete.
Participation	(Complete + Terminated + Quota Screenout + Quota Overquota) divided by starts in that group.
Incidence rate	(Complete + Quota Overquota) divided by starts after excluding Poor Quality and derived Drops. Other statuses, including active In Progress, remain in the denominator.
Completion rate	Completes divided by all starts with a recorded current status, including In Progress.
Progress to target	Completes divided by the configured target.
Median LOI	Median valid interview length for Complete respondents with recorded start and completion times, displayed in minutes. Interview length sums active time between recorded survey events, rounding each gap up to whole seconds and capping each gap at 60 seconds. It is not simply elapsed time between opening and finishing the survey.

Rates with a zero denominator are unavailable, rather than 0%. The status chart retains the latest source statuses: a respondent classified as a Drop in Drop Review still appears under In Progress here. Fielding pace groups completes by completion date in the report timezone for the selected group of survey starters.

Date window and report controls

Select the settings icon, Report controls, beside Download.

1. Enter From and To, or leave either blank for an open-ended window. The bounds are inclusive and To must not precede From. Enter times in the local timezone named beneath the date fields.
2. Set Report timezone to a valid timezone such as UTC or Europe/London. The default is UTC. This controls daily grouping; it does not change the local timezone used to enter From and To.
3. Set Minimum question base to a whole number of zero or more. The default is 10. This filters Drop Review question rows by their Last recorded base; it does not remove those respondents from the Overview.
4. Select Apply to refresh the report using these controls.

The last applied minimum question base is saved for your signed-in user, separately for each survey or respondent-source report, including zero. Reopening that report on another device restores it. Editing the field without selecting Apply does not save it. From, To and Report timezone are not saved as user preferences. The visible controls, Drop Review and Excel download use the same applied value.

Read the period, timezone and Filter basis: respondent start time line above each section. Dates select people by their first recorded survey activity, rather than selecting individual answers or people who completed within the window. Their latest outcomes can therefore change after the selected end time. Daily completion charts are also bounded to the selected calendar window, so their visible daily totals need not equal all completes in that start-time group. Check Generated below the section when assessing how current the snapshot is.

Breakout

Breakout compares starts, completes, participation, incidence and median LOI for the available demographic or breakout questions. Use it to identify groups with different screening outcomes or interview lengths.

Each table uses the respondents represented by that question and category. Categories can overlap; use the displayed Total, which counts each respondent once within the group, instead of adding category rows.

Quotas

Quotas shows the same metrics for quota terms, grouped by quota question. A term's base is the respondents evaluated for that term. Use the results to investigate where qualification or quota availability differs; the table is not a list of everyone who started the survey.

Use the survey's quota settings to inspect or change targets. See Setting up quotas.

Drop Review

Drop Review attributes respondents to their last recorded answered question. It separates Active In Progress, Drops, Terminated and Poor Quality.

A Drop is a respondent whose latest source status is In Progress and whose latest recorded survey activity was at least one hour before report generation. An In Progress respondent with activity less than one hour old is Active In Progress. New, terminated, complete and other statuses do not become Drops merely because time has passed. Metadata or status-only updates do not reset survey activity time.

For a respondent to appear against a question here, they must have an answered, reportable human-response question with a valid timestamp and a positive survey-code line number. Calculated variables, quota records and completion records are not used as that last answered question. Respondents without an eligible question can affect overall metrics without appearing in these diagnostics.

Last recorded counts respondents whose last eligible answered question is that row. Each status share uses this base, not everyone who saw the question or all survey starts. The minimum question base can hide rows with too few respondents.

Select a linked Line number to open that location in the survey code. A positive line number and survey context are needed for the link. Investigate the question and the route after it: a last recorded answer identifies where recorded activity stopped, not proof that the question itself caused the dropout.

Technical

Technical compares exact source statuses by Device and Browser. Each status share uses the respondents in that device or browser group. A high loss rate on one browser is a reason to inspect that experience alongside the survey route and sample composition.

Unknown device or browser values mean the information was absent or could not be identified. N/A means unavailable; an explicit zero remains zero.

Quality

Quality shows duplicate results and the 0–5 MX8 suspect-score distribution, together with Respondents, Actual terminations and Unattributed terminations. Use the combined view or an available individual-source view to compare quality across the selected live sources.

The section uses the same source scope and date-filtered respondent snapshot as the rest of the report. Read the selected period, timezone, filter basis and generation time before comparing two reports.

• A score of 0 is a recorded clean result, and False is a recorded negative duplicate finding. Unavailable, Not recorded and Invalid results are separate categories. N/A is not zero.
• Actual terminations come from recorded quality-policy exclusions, rather than being inferred from the score or today's source settings. Terminations without an attributed quality dimension remain Unattributed.
• Available evidence is retained even when duplicate or suspect termination is disabled. A respondent with a positive finding is therefore not necessarily excluded.
• Distributions use the latest recorded scores and duplicate findings. Use the saved termination reason to check why a respondent was excluded.

Do not add duplicate and suspect termination counts to obtain an overall total; use Actual terminations. For the settings behind these results, see Fraud prevention measures. For respondent-level fields, see Downloading data and reports.

Respondent Sources

The survey-level report includes a Respondent Sources tab. Comparison data is available when the report includes more than one live respondent source. A single-source report does not show this tab; a survey report with only one source explains why comparison is unavailable.

Compare each source's starts, completes, participation, incidence and median LOI against the combined total. Respondent identities are counted within their source: matching IDs from different sources are not automatically merged into one person.

Moving through sections and checking report quality

Tables paginate independently. Moving to another page in one demographic, quota, device or browser table does not page every other table. Pagination changes which rows you see, not the underlying report base.

A Partial report warning names unavailable sections. Data quality warnings identify relevant issues such as unknown statuses, missing start times, invalid interview lengths or unknown device/browser values. Missing start times can exclude respondents from the start-time base. Inspect the affected section and its definitions before using those results; unavailable data is not evidence of zero respondents.

A pending report shows generation progress. An error state offers Retry. A section with no matching data explains that no data is available for the current controls.

Share a source's read-only field report

Open an individual respondent source's field report and select the share control. In Read only share links:

1. Enter a Name that identifies the intended recipient or purpose.
2. Choose Expires in: 7, 30 or 90 days. The default is 30 days.
3. Select Create link, then use Copy link to share the URL.

Recipients can open the read-only report without signing in. The link gives access to that source's field report, including aggregate Quality distributions, rather than the survey's combined source report or the rest of the account. Anyone with the URL can use it until it expires or is deleted, so share it only with intended recipients.

Return to this panel to manage each named link. Refresh validity extends access using the same URL. For an expired link, Refresh and reactivate link asks for confirmation before making that existing URL work again. Delete revokes the link immediately. Create separate named links when you need to revoke one recipient's access independently.

Excel downloads

Select the download icon and choose Excel. Export permission is required; the survey-level download also requires included sources and access to export their data.

The workbook includes Overview, Breakout, Quotas, Drop Review, Technical and Quality, plus Respondent Sources for a multi-source report. It includes definitions, charts and data-quality information; rate-cell comments show numerators and denominators. Interview lengths in the workbook are labeled in seconds, while the interface displays minutes.

The export uses the source scope and applied From, To, Report timezone and Minimum question base. A survey-level download includes all live sources. It exports the full report, not just the selected tab or current table page. If generation takes time, follow the job in Downloads; see Downloading data and reports.

Dispatch CSV for voice and text sources

For a supported Twilio Voice, Twilio Text or Twilio Text-to-Caller source, open that individual source's field report and select Download → Dispatch CSV. This choice requires destination-reporting support and export permission. It is in the download menu, not a separate button above the tabs, and is not offered in the survey-level comparison.

Dispatch CSV is the latest destination-level snapshot. The field report's date, timezone and minimum-base controls do not filter this CSV. Large exports may prepare in the background; follow progress in Downloads.

There is one row per dispatch attempt, so retries create additional rows. Use the destination and caller identity, campaign run, attempted sender, delivery/provider status, timestamps and failure details to investigate delivery. For Twilio text attempts, the sender snapshot records the configured phone number or Messaging Service SID, not the physical number selected by a Messaging Service. Twilio text error 30034 appears as an unregistered-sender failure with registration guidance. Filter to is_latest = True when reconciling current destination progress; retain all rows for the attempt history.

The lifecycle is Planned → Scheduled → Starting → Delivered or Failed. It measures outbound delivery, not survey participation. For voice, initiated, ringing, answered and in-progress events remain Starting; a completed provider event becomes Delivered. Busy, no-answer, canceled and failed events become Failed. Machine, voicemail and fax detections appear as no-answer; inspect the failure details for the reason.