Studio Panel Providers

Panel Providers

Sample providers such as Prolific, Respondent, Cint and Dynata send respondents to your survey with an id in the link, and expect them back on a URL that says how the interview ended — completed, screened out, or turned away because the sample was full. The Panel chip on an environment card sets this up and gives you the counts to reconcile with the provider's invoice.

---

How it works

 provider ──entry link──▶  study.siamang.org/<survey-id>/?RID=8f2…      (id arrives as a URL parameter)
                                   │
                         respondent answers
                                   │
        ┌──────────────────────────┼───────────────────────────┐
     completed                screened out       sample full (quota cell or cap)
        │                          │                           │
   Completed → return URL    Screened out → return URL    Quota full → return URL
        └──────────── each URL carries the id back: …&RID={url:RID} ────────┘

• The respondent's id is an ordinary URL parameter: it is stored with the response and reaches your flows as a url_<name> column (lower-cased: RID becomes url_rid, PROLIFIC_PID becomes url_prolific_pid).
• The three return URLs are part of the questionnaire — they are built into the published survey. Saving the panel setup creates a new Save; the environment uses it after you republish.
• The provider's name and the id parameter are project settings.

---

The Panel chip

Open an environment card's Panel chip. It is available while the environment serves a survey (live, paused, or a failed rebuild whose previous version still serves), not on previews: publish pilot and main before you set up a panel, then publish the panel's Save to both. The header reads Panel provider · not set until you save a setup, then the provider's name.

Field	What to enter
Provider	— none —, Prolific, Respondent, Cint, Dynata or Custom. Choosing a preset fills the next four fields
Respondent id parameter	the query parameter that carries the provider's id (letters, digits and _ only). Hint: "the query parameter in the entry link"
Entry link for the provider	read-only, with a copy button: your survey link plus the id parameter and the provider's macro. Paste it into the provider's project. With Prolific, the plain survey link (hint: "paste into Prolific's study link field"): Prolific adds its parameters itself. With Respondent, the plain survey link too (hint: "paste into Respondent's Link to Task field"). Shown once the environment has a link and the id parameter is set
Completed → return URL	where respondents go after completing
Screened out → return URL	where respondents go after reaching a screen-out page
Quota full → return URL	where respondents go when a full quota cell or a full response cap turns them away

Notes the panel may show:

• the preset's own advice (see the table below);
• "Replace the <…> placeholders with the codes or tokens from the provider's project page before saving." — while a URL still contains a placeholder such as <COMPLETION_CODE>;
• "Pages with their own redirect keep it: <page names>. Clear a page's redirect in the Builder if the provider's URL should apply there."
• with Prolific, under the entry link: "In Prolific, choose I'll use URL parameters: Prolific adds PROLIFIC_PID, STUDY_ID and SESSION_ID to this link. Then check that the study link in Prolific shows each of them once."; and above Save panel setup, a line about the screen after Save locally and finish (see Prolific);
• with Respondent, under the entry link: "In Respondent, paste this link into Link to Task (Custom URL). Respondent adds respondent_profile_id, respondent_user_id, respondent_project_id and respondent_screener_response_id itself."; and the same line above Save panel setup, for Respondent (see Respondent).

Save panel setup (enabled once something changed) saves a new version: "Saved as #18 — republish main to apply the return URLs". When the published Save is older, the panel reminds you: "Published Save #17 may differ from the current #18 — republish to apply."

---

Presets

Provider	Id parameter	Entry link adds	Completed	Screened out	Quota full
Prolific	PROLIFIC_PID	nothing — Prolific adds PROLIFIC_PID, STUDY_ID and SESSION_ID itself once I'll use URL parameters is chosen	https://app.prolific.com/submissions/complete?cc=<COMPLETION_CODE>	…?cc=<SCREENOUT_CODE>	…?cc=<QUOTA_FULL_CODE>
Respondent	respondent_user_id	nothing — Respondent adds respondent_profile_id, respondent_user_id, respondent_project_id and respondent_screener_response_id to the project's Link to Task itself	https://app.respondent.io/complete?cc=<projectCompletionId>	empty	empty
Cint	RID	?RID=[%RID%]	https://s.cint.com/Survey/Complete?ProjectToken=<PROJECT_TOKEN>&RID={url:RID}	https://s.cint.com/Survey/EarlyScreenOut?ProjectToken=<PROJECT_TOKEN>&RID={url:RID}	https://s.cint.com/Survey/QuotaFull?ProjectToken=<PROJECT_TOKEN>&RID={url:RID}
Dynata	psid	?psid=[%psid%]	https://dkr1.ssisurveys.com/projects/end?rst=1&psid={url:psid}	…?rst=2&psid={url:psid}	…?rst=3&psid={url:psid}
Custom	id (change it)	?id=<RESPONDENT_ID>	empty	empty	empty

The preset notes, as shown in the panel:

• Prolific: "Prolific identifies outcomes by completion codes: paste the codes from the study's page into the three URLs. What each code does to a submission is set in Prolific."
• Respondent: "A Respondent survey project has a single completion URL, its Completion Confirmation URL: paste it into Completed and leave Screened out and Quota full empty. Screening and quotas stay in Respondent, in its screener, before the invitation."
• Cint: "Cint passes the respondent as RID and expects it back on every return URL; the project token is on the project page."
• Dynata: "Dynata's return status: rst=1 complete, 2 screen-out, 3 quota full. Confirm the endpoint with your project manager."
• Custom: "Any provider: name the query parameter that carries their respondent id and put {url:} where the return URL needs it."

The presets are starting points: always check the URLs against the provider's current documentation and your project page.

---

Placeholders in return URLs

Placeholder	Replaced by
{url:NAME}	the value of URL parameter NAME the respondent arrived with, exactly as the provider named it (e.g. {url:RID}, {url:PROLIFIC_PID})
{answer:var}	the respondent's answer to variable var (codes; several answers joined with commas)
{label:var}	the label of that answer

Values are URL-encoded. A placeholder that has no value is removed rather than sent to the provider as literal braces. In the Quota full URL, answers are filled in when a full quota cell ends the interview; when a full response cap turns the respondent away (as the page opens, or at submit), only {url:NAME} is filled in and {answer:…} / {label:…} are removed.

---

Which URL applies when

How the interview ends	Where the respondent goes	After
presses Submit responses on the last page (a page with questions and no end page after it)	Completed → return URL	5 seconds ("You will be redirected in 5 seconds. Click here if not redirected.")
reaches a Final (thank you) page	the Completed URL	5 seconds
reaches a Screen-out page	the Screened out URL	5 seconds
reaches a Redirect page	the page's own Redirect URL (the Completed URL if the page has none)	the page's Delay (s), 5 by default
leaves a page with an answer whose quota cell is full	the Quota full URL	3 seconds after the "Thank you for your interest" screen, which adds "Redirecting you now. Continue if you are not redirected."
a response cap is full — the notice appears as the page opens, or at submit for someone already answering	the Quota full URL	3 seconds after the "Thank you for your interest" notice
the environment is closed or past its closing date	the environment's post-close redirect (redirect_after_close in studio/settings.json), if it has one; otherwise the respondent stays on the "This survey is closed" notice	3 seconds
the environment is paused	nowhere — the respondent stays on the "This survey is paused" notice	—
submitting fails and the respondent presses Save locally and finish	nowhere — the respondent stays on "Your answers are saved in this browser", with no end page	—
the survey page cannot reach Studio (its env.js did not load)	nowhere — every attempt to submit fails, with the Try again dialog; opening the link again or reloading the page, then Resume, sends the answers	—

A terminal page with its own redirect keeps it — the page wins over the survey-level URL; the Panel chip lists those pages. On terminal pages the redirect happens only after the response has been stored, with "Redirecting you now. Continue if you are not redirected."; the page's title and body appear only then too. A terminal page cannot be left: Esc, which goes back on a question page, does nothing there, so a respondent who reaches a Screen-out page goes to the Screened out URL (or the page's own redirect), never to the Completed one. When submitting fails, Save locally and finish keeps the answers in the respondent's browser, unsent, and sends them nowhere, so a provider's return URL is never followed, and a code in an end page's body never shown, without a stored response (see When submitting fails). Resume reopens the page they were on by its name, even if you republished in between, so someone saved on a Screen-out page is still stored as screened out and sent to the Screened out URL; if you removed that page, it opens the first page with their answers kept. Progress saved by a survey built before Resume kept the page's name never reopens an end page by its position: Resume opens the first page with the answers kept, and Next leads the respondent to the end page those answers reach today, and on to its URL. The survey page's source still holds every page and return URL, codes included, so check the provider's list against the completed outcomes below before you approve anyone (see Reconciling outcomes).

The quota-full rows depend on the survey page: a survey built before quota cells could stop respondents, or before the page-open check existed, uses the Quota full URL only when the response cap refuses a submission, until you republish it. How cells fill and when they stop someone is described under When a cell is full; the caps under Response caps.

Screen-outs are submitted responses: they are stored (with __status set to screened_out), but they are not completed interviews — they fill no quota cell and do not count toward a response cap. A screen-out is recorded even when the cap is full.

---

Recipes

Prolific

1. In Prolific, create the study. Keep the page with the completion codes open. Create a code for each outcome you use (table below) and set what each code does to a submission in Prolific, before you publish the study. Keep its places below the environment's response cap and, on the Free plan, below the project's 1,000 completed responses (below the table).
2. In the Builder, set Theme → Respondent experience → Estimated minutes to the study's estimated time, and save: the Builder's Prolific warnings read it.
3. In Distribute, publish the current Save to pilot and to main if they are not published yet: the Panel chip is only on the card of an environment that serves the survey. Each keeps its link when you publish again, and nobody reaches main before you paste its link into Prolific (step 9).
4. Open either environment's Panel and choose Prolific.
5. Replace <COMPLETION_CODE>, <SCREENOUT_CODE> and <QUOTA_FULL_CODE> with Prolific's codes. If you do not use a screen-out or quota-full code, clear that field. A Screen-out page that needs a code of its own gets its own redirect: redirect_url on the page, in the Builder's Source tab (the Redirect URL field is on Redirect pages only).
6. Save panel setup. With Prolific saved as the provider, the Builder now shows Prolific's rules where a question meets them (below); save any change you make there.
7. Publish the latest Save to pilot and to main. The return URLs and the saved-in-browser text are part of the questionnaire, so both environments send respondents to the same codes; the provider and the id parameter are project settings.
8. Test on pilot: open its link with a made-up id (?PROLIFIC_PID=test0000000000000000001&STUDY_ID=test&SESSION_ID=test), finish the survey, check that you land on Prolific's completion page with the right code, and find the row in Data. Repeat for each Screen-out page. A later Save has to be published to both again before you test it.
9. Copy main's Entry link for the provider — the plain survey link — into Prolific's study link field and choose I'll use URL parameters. Check that the study link Prolific shows has PROLIFIC_PID, STUDY_ID and SESSION_ID once each (and, if you preview the study from Prolific, the address of the survey page that opens). A study created through Prolific's API gets nothing appended: add ?PROLIFIC_PID={{%PROLIFIC_PID%}}&STUDY_ID={{%STUDY_ID%}}&SESSION_ID={{%SESSION_ID%}} yourself.

A study already live. What the survey page does after a failed submission, on an end page and when it cannot reach Studio is built into the page when an environment is published, and an environment keeps the build it was last published with. A survey published before the current Studio can still send a participant on as completed after Save locally and finish, let them go back from an end page, end as sent when its env.js did not load, and restart a duration at a reload, even one followed by Resume. Save in the Builder (a Save with no changes will do) and publish that Save to pilot and main (see Republishing).

The screen after Save locally and finish. Its default text asks the respondent to come back within 7 days; a Prolific submission times out in less than 24 hours, and Prolific asks participants who hit a technical problem to contact the researcher. So Save panel setup with Prolific sets Saved in this browser: text (Builder → Theme → Wording), in the same Save as the return URLs, to "They have not been sent yet. Please message the researcher through Prolific. While the study is still open on Prolific, you can also open its link again in this browser and choose Resume to send them." It never replaces a text you wrote yourself, and a Save that moves the provider away from Prolific takes it out again while it is still exactly that text. The line above Save panel setup says which applies ("Save panel setup also sets the screen after Save locally and finish to ask participants to message the researcher through Prolific. …").

Which code each outcome needs. Prolific allows screening inside a study only to re-check one of its prescreeners in its own wording, after a device check at the start when the study's description states the requirement ("Incompatible device", with a request to return), for a participant who does not consent, after failed attention checks that meet its policy, and after a comprehension check failed twice; any other criterion goes through Prolific's custom screening, with its fixed screen-out payment. Follow Prolific's current guidance in its help center.

Outcome in Studio	Studio sends the respondent to	Code in Prolific	Action set in Prolific
completed	Completed URL	the main completion code	approve, or manual review
a Screen-out page for those who do not consent	the page's own redirect, else Screened out	"No consent"	request a return
a Screen-out page for an answer that contradicts a Prolific prescreener	the page's own redirect, else Screened out	a Custom code, such as "Ineligible"	request a return
a Screen-out page for an incompatible device, after a device check at the start	the page's own redirect, else Screened out	"Incompatible device"	request a return
a Screen-out page for a criterion of your own	the page's own redirect, else Screened out	Prolific's custom screening code	custom screening's fixed payment
failed_attention_checks, the page of End the survey after N failed attention checks	the page's own redirect, else Screened out	"Failed attention checks"	as Prolific's attention-check policy allows
quota full as the page opens (a full response cap)	Quota full URL	a Custom code	request a return
quota full after answering (a full quota cell)	Quota full URL	Prolific's custom screening code	custom screening's fixed payment
quota full after completing (a response cap filled while they answered, and their submission was refused)	Quota full URL	the same code as the other quota cases	none fits: approve, and pay the full reward, as a bonus for the difference if the code auto-approved the submission at the screen-out reward

All three quota cases share the Quota full URL. Two response caps count completed responses: the environment's own, and on the Free plan the project's 1,000, counted over all its environments together, pilot tests included. Prolific's places limit the sample, so keep them below both caps: quota full then comes from quota cells only. A cap is also checked at submit, and someone turned away there has completed the study: approve them and pay the full reward, as a bonus for the difference if the Quota full code auto-approved them at the screen-out reward.

Attention checks. Prolific needs at least two failed checks to screen out or reject a participant in a study of 5 minutes or more, and checks used to screen out come at the start of the study. Keep Also end the survey for respondents who fail off; score the checks in the analysis, or use End the survey after 2 failed attention checks (see Attention checks). A nonsensical statement ("I swim across the Atlantic to get to work") is answered on a scale without a neutral option, so give its Likert scale an even number of points, and needs no prior knowledge; any answer on the correct side of the scale passes, and a statement to which an unexpected answer can be justified is unsuitable. With Prolific as the provider, the Builder warns about a check that screens out after one failure, notes a Likert check with an odd number of points, and warns about Email, Phone and Date questions (Prolific does not allow a phone number, nor a personal email address or a date of birth without its written approval; ask for age in years).

Reviewing submissions. Compare Prolific's submissions with the outcome CSVs below before you approve. In short, as Prolific's rules stand:

• A participant who completed is approved unless there is a valid reason to reject: required key questions unanswered, failed attention checks under its policy, clearly low effort, an exceptionally fast submission (more than 3 standard deviations below the mean time), a failed authenticity check.
• A submission with no stored response, or only a partial one, is not rejected for that. Check it: an exceptionally fast one may be rejected; otherwise message the participant, and approve when it looks like a technical problem. Consider a partial payment for the time spent: a partial row's duration_seconds runs to its last saved progress, a lower bound, so compare it with the time Prolific shows.
• A submission that came back with the Quota full code after the participant reached the end was refused by a full response cap: approve it and pay the full reward.
• A wrong code or NOCODE alone is not a reason to reject.
• Several responses with one id: keep one for the analysis and approve the submission.
• Speeders and quality flags are reasons to leave a response out of the analysis, not by themselves reasons to reject.

Respondent

A Respondent Survey project opens the survey from its Link to Task with Respondent's identifiers added, and at the end sends the participant to the project's Completion Confirmation URL, the link-based integration Respondent documents for survey tools. Studio does not use Respondent's API.

1. In Respondent, create a project with the Survey methodology. Set its audience, screener and incentive so that everyone Respondent invites qualifies for the whole survey: screening and quotas stay in Respondent, before the invitation. For the task, choose Custom URL with the Completion Confirmation URL option, and copy that URL (https://app.respondent.io/complete?cc=…).
2. In Distribute, publish the current Save to pilot and to main if they are not published yet: the Panel chip is only on the card of an environment that serves the survey. Each keeps its link when you publish again, and nobody reaches main before you paste its link into Respondent (step 7).
3. Open either environment's Panel and choose Respondent. The id parameter becomes respondent_user_id, Completed gets https://app.respondent.io/complete?cc=<projectCompletionId>, and Screened out and Quota full stay empty. Replace the Completed URL with the one you copied. If the panel lists pages with a redirect of their own, clear those redirects in the Builder unless they are meant to apply. Save panel setup.
4. In the Builder, make sure an invited participant can end the survey only by finishing it: by Submit responses, on a Final (thank you) page or on a Redirect page without a URL of its own. No branch leads to a Screen-out page, Also end the survey for respondents who fail and End the survey after N failed attention checks stay off, and there are no quota cells. With Respondent now saved as the provider, the Builder points out each of these (see Respondent's warnings in the Builder). A consent question whose decline leads to a Screen-out page may stay: write in that page's text that the participant should let the researcher know through Respondent's messages. Save any change.
5. Publish the latest Save to pilot and to main.
6. Test on pilot: open its link with made-up identifiers (?respondent_profile_id=test&respondent_project_id=test&respondent_screener_response_id=test&respondent_user_id=test0001), finish the survey, check that you land on app.respondent.io/complete, and that Data has the row with url_respondent_user_id test0001. For the whole path, invite a teammate with a Respondent participant account with the pilot link (Respondent lets you change the survey link when you invite someone), and check that they appear under Completes.
7. Copy main's Entry link for the provider — the plain survey link — into Link to Task as it stands. Add no parameters of your own: the survey keeps the first 8 parameters of its link, and Respondent appends its four after any already there, sometimes with participant_id and project_id.
8. Publish the Respondent project and invite participants; review each completion before Respondent's Auto-Pay pays the participant, 48 hours after task completion (below).

Each response stores url_respondent_profile_id, url_respondent_user_id, url_respondent_project_id and url_respondent_screener_response_id, on the partial row and the submitted one, so no Participant ID question is needed. Respondent recognizes the returning participant by their Respondent sign-in in the same browser, so the completion URL needs no placeholder.

Ending only at the completion URL. A Respondent survey project has no screen-out or quota-full return. A participant whom the survey stops on a page of its own stays there and never reaches the completion URL, and their task stays incomplete in Respondent until you resolve it by hand:

How the interview ends	Where the participant goes
submits on the last question page, or reaches a Final (thank you) page or a Redirect page without a URL of its own	the Completion Confirmation URL, once the response is stored
reaches a Screen-out page	the page's own redirect, if it has one; otherwise nowhere
turned away by a full quota cell or response cap	nowhere — "Thank you for your interest"
submitting fails, or Save locally and finish	nowhere, until a later attempt or Resume stores the response

Keep the number of participants you invite below the environment's response cap and, on the Free plan, below the project's 1,000 completed responses.

The screen after Save locally and finish. A participant whose answers were not sent has not reached the completion URL. Save panel setup with Respondent sets Saved in this browser: text, in the same Save as the return URL, to "They have not been sent yet. Please message the researcher through Respondent. You can also open this survey's link again in this browser within 7 days and choose Resume to send them." As with Prolific's, it never replaces a text you wrote yourself and goes again when the provider moves away from Respondent while it is still exactly that text; moving a setup between Prolific and Respondent replaces one provider's text with the other's.

Reviewing completions. In a project created in Respondent's web interface, as with this setup, reaching the completion URL marks the participant complete at once, and Auto-Pay, on by default, pays the incentive 48 hours after that. To undo a completion, change the participant's completion status in Respondent before Auto-Pay pays. In a project created through Respondent's API, completion is marked 24 hours after the participant reaches the URL. So review each completion within those 48 hours. Respondent's Completes tab lists the participants who reached the completion URL by their Participant ID, which Respondent documents as the same ID as respondent_user_id, the id column of the outcome CSVs below.

What you find in Studio	Action in Respondent
a completed response of acceptable quality	nothing more: the participant is already marked complete, and Auto-Pay pays them 48 hours after task completion
a completed response that does not meet the study's quality standards, such as failed attention checks	change the participant's completion status in Respondent before Auto-Pay pays; for a reduced incentive, or anything else, contact Respondent
a participant listed under Completes with no completed response in Studio	look for their partial row in Data and message them before Auto-Pay pays: they may have opened the completion URL without finishing, or hit a technical failure. If they did not complete the task, change their completion status
a completed response, or a partial row that reached the last page, and the participant is not under Completes	they did the whole survey but did not reach the completion URL: they may have closed the page before the redirect, or been turned away by a response cap that filled while they answered. Respondent does not mark them complete without it, so mark them as having attended in Respondent yourself
only a partial response, stopping before the last page, and the participant is not under Completes	message the participant through Respondent. Without the completion URL they are not marked complete, and Auto-Pay does not pay them
signs of fraud, such as one person answering under two different IDs	report the participant to Respondent

Studio's quality signals inform the decision; it stays with you, under Respondent's rules. The Median duration tile's "exceptionally fast" line is Prolific's rule and is not shown with Respondent.

Personal data. Respondent appends its identifiers so that a survey need not ask for an email address or other personal data, and prohibits collecting a participant’s phone number, physical address or date of birth, among other personal data, removing researchers who do from the platform. For a project created in its web interface, its help center describes an email visibility option, in which participants who opt in enter the email of their Respondent account in the survey tool; otherwise Respondent does not share participants' contact details with researchers. The Builder notes Email, Phone and Date questions, and the Email note names that option.

Respondent's warnings in the Builder

With Respondent saved as the provider, the Builder says where the questionnaire can end an interview short of the completion URL. The notes never stop a Save; Validation → Respondent lists them all.

Where	When	What it says
an attention check's Inspector	Also end the survey for respondents who fail, or End the survey after N failed attention checks, is on	"Respondent has one completion URL: a participant screened out on failed attention checks stays on Studio’s page and never reaches it; leave screening out off, score the checks in the analysis, and decide on the participant in Respondent before Auto-Pay pays them, 48 hours after task completion (24 hours for projects created through Respondent’s API)."
a Screen-out page's Inspector, under Kind	a participant can reach the page: a branch rule, a Default next or a Skip to leads there, or no page before it ends the survey for everyone (an attention check's own page is said at the check)	"Respondent has one completion URL: a participant who ends here stays on Studio’s page and never reaches it; screen in Respondent’s screener instead.", then "If you keep this page, as one for participants who decline consent, say in its text that they should let the researcher know through Respondent’s messages."
Validation only	the questionnaire has quota cells	"Quotas: Respondent has one completion URL: a participant a full quota cell turns away stays on Studio’s page and never reaches it; set quotas in Respondent instead."
Validation only	an environment has a response cap	"Response cap (pilot 50, main 1,200): Respondent has one completion URL: a participant the cap turns away stays on Studio’s page and never reaches it; keep the participants you invite in Respondent below it."
under Format of an Open text question	Email, Phone or Date	"Respondent recognizes participants by the IDs it adds to the survey link, not by an email address. Ask for an email only if the project uses Respondent’s email visibility option, in which participants who opt in enter the email of their Respondent account; for participants sourced through Respondent’s API, requesting personal data is a violation of its terms. Otherwise contact participants through Respondent’s messages." (Email), "Respondent prohibits collecting a participant’s phone number, and removes researchers found collecting personal data from the platform. Contact participants through Respondent’s messages." (Phone), and "If this asks for a date of birth: Respondent prohibits collecting it. Ask for age in years instead." (Date)

Follow Respondent's current guidance in its help center ("Auto-Complete & Auto-Pay for Research Studies", "Completion Confirmation URL: Surveys and Unmoderated Studies", "Unmoderated Research Methodology") and developer documentation ("URL Parameters for Project Link", "Completion URL").

Cint

1. Choose Cint. The id parameter is RID.
2. Replace <PROJECT_TOKEN> in all three URLs with the token from the Cint project page. Keep RID={url:RID} — Cint expects its id back.
3. Save panel setup, republish.
4. Paste the entry link (…?RID=[%RID%]) into the Cint project.

Dynata

1. Choose Dynata. The id parameter is psid; the three URLs differ only in rst=1 (complete), 2 (screen-out), 3 (quota full).
2. Confirm the end-point host with your Dynata project manager and adjust the URLs if needed.
3. Save panel setup, republish, and send Dynata the entry link (…?psid=[%psid%]).

Any other provider

1. Choose Custom.
2. Set Respondent id parameter to the parameter name the provider uses (for example uid).
3. Enter the three return URLs from the provider's documentation, putting {url:uid} wherever the provider needs its id back.
4. Save panel setup, republish, and give the provider the entry link, replacing <RESPONDENT_ID> with the provider's own macro for the id.

For all providers: mark the screener's disqualification ending as a Screen-out page in the Builder so screened-out respondents are recognized and sent to the right URL. See Logic and Branching.

---

Reconciling outcomes

While the environment is live, the Panel chip ends with Outcomes · reconcile with the provider:

Count	Which responses
completed	submitted responses that did not end on a screen-out page (Redirect and Final pages count as completed)
screened out	responses that ended on a Screen-out page
partial	interviews started and not submitted

Each count with at least one response has a CSV button. The file (<environment>-outcomes-<outcome>.csv, e.g. main-outcomes-completed.csv) has one row for every response of the environment with that outcome, with response_id, outcome, a column named after your id parameter (spelled as in Respondent id parameter, e.g. PROLIFIC_PID or respondent_user_id) holding each respondent's provider id, submitted_at, duration_seconds (whole seconds from the start of the interview to the response's last save; empty when unknown — the start is the opening of the survey page on which the respondent began, in that browser; a reload followed by Resume, or Save locally and finish followed by Resume, keeps it, and Start over, or answering after a reload without choosing Resume, starts the count again), duplicate_id (yes when the id is on another response of the environment too, whatever its outcome) and fast_outlier (yes on a completed, not screened-out response more than 3 standard deviations below the mean duration of those responses, each taken as the longer of duration_seconds and the time between the response's first and last save in Studio; with 10 or fewer timed responses, none is). If the download fails, the toast reads "Could not download the outcomes." followed by the reason. Each download is recorded in Settings → Activity as outcomes.export, since the provider ids leave Studio with it.

Under the counts, a line says how many responses carry an id — 972 of 1,184 responses carry a PROLIFIC_PID; (or "Set the id parameter to see provider ids;" before you set one) — followed by "quota-full returns are not counted here — the response is refused, the provider's count is the record." Respondents turned away by a full quota cell or cap leave no completed response, so the provider's own report is the reference for them. Someone who was turned away after answering a page or two may still appear under partial, with the progress the survey had saved.

More lines follow: "3 IDs appear on more than one response." when an id is on more than one response (with Prolific saved as the provider: "… Prolific's advice: keep one response for the analysis and approve the submission."); "Each CSV lists every response’s duration_seconds and marks duplicate_id on a response whose ID is on another response too, and fast_outlier on a completed response more than 3 SD below the mean time."; with Prolific: "On Prolific, check a submission without a stored response: an exceptionally fast one may be rejected; otherwise message the participant, and approve it when it looks like a technical problem."; and with Respondent: "On Respondent, a participant is marked complete as soon as they reach the completion URL, and Auto-Pay (on by default) pays the incentive 48 hours after that; for projects created through Respondent's API, completion is marked 24 hours after they reach it. Review each completion before Auto-Pay pays it. A participant listed under Completes without a completed response here: look for their partial row in Data and message them through Respondent before deciding. One who finished here but is not under Completes was not marked complete: mark them in Respondent yourself."

Partial rows arrive only from a survey built with the current runtime; for one published earlier, partial stays at 0 until you republish it.

Studio stores URL parameter names in lower case (url_prolific_pid), and the outcomes match your id parameter regardless of case, so the upper-case ids of the Prolific (PROLIFIC_PID) and Cint (RID) presets are found like any other.

Tip. To put the provider id next to the answers, export the responses table from Data: the export carries the id as a url_<parameter> column in lower case (url_prolific_pid, url_rid, url_psid), next to __status and partial (see Data Exports). A flow with an Export file node does the same after cleaning. See Analysis Flows.

The outcome block is not shown for paused or closed environments; for those, use a flow or the Data tab.

See also

• Links, QR Codes, Embeds and Access Control
• Publishing and Environments
• What Respondents See
• Quotas and Randomization
• Logic and Branching

---

← Links, QR Codes, Embeds and Access Control · Studio contents · Email Invitations →