Folio Surveyor takes a study from a blank instrument to analyzed responses without leaving Folio. It's on every plan, including the free Scholar plan. What the plan changes is how many responses a project can collect (50 on Scholar, 200 on Fellow, no limit on Chair, though each survey link stops at 1,000 responses) and whether the AI tools are available: the wording check, the Methodology assistant, AI item synthesis for Delphi rounds, AI theme analysis and AI translation need Fellow or Chair. The Methods and Results notes, which Surveyor builds from your data without AI, are on every plan.
Surveyor has its own help panel inside the product, with the statistics it runs and every error message explained. This article is the overview.
Building an instrument
Start a new project and add questions of any type: Likert scales, multiple choice, open text, ranking and more.
The fastest start is usually a validated scale. Choose Import scale in the builder, or open a scale's page in the validated scales library and choose Open Surveyor: the items, answer labels, subscales and reverse-scored items come in exactly as published. When a respondent answers a scale, its instruction (the stem, such as "Over the last 2 weeks...") repeats above every item, because Surveyor shows one question at a time. Some scales can't be imported: a licensed instrument needs permission from its copyright holder, and for a few others (Grit-S, WHO-5 and MSLQ among them) Folio holds the citation and methods details but not the item wording, so you add the items yourself from the published source.
On Fellow and Chair, the AI audit checks your wording for the problems that quietly bias results: leading phrasing, loaded terms, lopsided scales, and double-barreled items. It flags them before you distribute. (If you want the why behind those flags, see our post on survey questions that don't lead.)
If you're stuck on coverage, the Methodology assistant (Fellow and Chair) turns a research question into a study plan with one draft item per construct. Draft items are marked as AI-suggested until you edit them: a starting point, not a finished instrument.
Bring an existing Qualtrics study
On the Surveyor projects page, choose Import Qualtrics and upload the .qsf file exported by Qualtrics. Surveyor shows a preview before it creates anything. Multiple-choice, matrix, text-entry, slider, and descriptive items come across with their block order, answer choices, required status, and any single-condition display logic Surveyor can represent.
Read the Needs review list before importing. Unsupported question types and compound or platform-specific logic are never discarded silently: the preview names each one, and the same list is saved in the imported project's metadata for later reference. After import, open the builder and check the wording, scale anchors, branching, and block breaks before you distribute.
Delphi rounds
For expert-consensus studies, Surveyor supports Delphi rounds. Round 1 collects open-text answers from your panel; between rounds you turn those answers into items (with AI help on Fellow and Chair); each later round asks the panel to rate the items that have not reached consensus yet. Each round has its own invite-only link, and the consensus map shows how agreement forms (or doesn't) across rounds. Before each item, panelists see how many people rated it in the previous round; if your protocol calls for feedback on the group's ratings, add it yourself as an information block.
Translating a survey
Need the instrument in another language? On Fellow and Chair, creating a new version in another language can translate the whole thing (questions, options, and scale labels) while keeping scientific terms consistent, so the same concept is rendered the same way every time rather than drifting between synonyms. You review the translation side by side with the original and pick, question by question, which translations to accept. Accepted translations are baked into a new version of the survey, so respondents see a properly localized instrument, not an English survey with a language toggle.
Distributing
Create a distribution to get a shareable link: start with a Pilot to test, then a Main distribution for your real data collection (activating a main distribution locks the questions). Responses are pseudonymous by default: each one is linked to a random participant token, never to an email or an IP address, and you can choose full anonymity instead. Right-to-left languages read correctly for respondents, so Arabic and Hebrew instruments work as they should.
Turn on Save and continue later in the distribution settings if your instrument is long enough that people might not finish in one sitting. A respondent who closes the tab partway through can reopen the same link on the same device and pick up where they left off. It isn't available with full anonymity, since resuming needs a token tied to that respondent's progress.
Return Prolific participants after submission
For a Prolific study, open the distribution's After submission settings and paste the completion URL from Prolific. If you only have the alphanumeric completion code, enter that instead and Surveyor builds Prolific's standard completion URL. A saved full URL takes precedence when both are present.
After a valid response is recorded, Surveyor shows a manual Return to Prolific link and automatically sends the participant there after two seconds. This is the handoff Prolific uses to match the configured completion code and mark the submission complete. Test the participant link yourself before publishing the Prolific study, including the final redirect and code shown in Prolific. See Prolific's survey software guidance for its current completion-link requirements.
Analyzing results
The Results view summarizes responses per question (distributions, scale means, and open-text answers), with scale reliability, comparisons and regression. Speed outliers are flagged for your review. In Open text, a check for low-effort answers runs only when you click it, because it sends that question's answers to an AI service; theme analysis works the same way.
To take the data further:
- Export from the Results overview: a full data CSV, an SPSS-labeled CSV, an Excel workbook, a codebook, JSON, an APA text report, and more, on every plan.
- Google Sheets (live) creates a new Google Sheet with your data and gives you the link. Today the sheet is created in Folio's shared Google Drive and shared with the email you sign in with, so Folio keeps that copy. The option appears only where Sheets export is set up.
Writing it up
The Deliverables tab builds your Methods and Results notes from the data: sample, measures with their published and observed reliability, descriptive statistics, and a list of what the data can't tell you (how you recruited, how you chose the sample size). They are notes to copy from, not paragraphs, and no AI writes them. Your paper is yours to write.
When you're ready, Start the write-up in Results creates a document with an empty Method and Results outline and a table of your descriptive statistics, and opens it with the notes beside it.
Tips
- Pilot before you publish. Send the instrument to a handful of people outside your field and ask what each question is asking. It catches more than a solo re-read.
- Translate last. Get the source-language instrument right first. Every fix afterward has to be re-translated.
- Keep a neutral middle. Forcing a side manufactures opinions that don't exist; offer a genuine neutral and, where honest, a "not applicable."