Creating a study

Start a new study

Click New study on the dashboard (the same button starts a card sort, a pairwise study, or a Best–Worst study, pick Card sort here; pairwise and Best–Worst have their own articles, pairwise and Best–Worst). You'll set:

• Title (required): defaults to "Untitled study" if left blank.
• Client / folder (optional): files the study into a folder on your dashboard.
• Sort type (required): see below.

A new study starts as a draft. It gets a unique participant link automatically.

The study funnel

Once created, a study moves through six steps, left to right: Build (title, cards, categories) → Audience (who takes part) → Screener (qualifying questions) → Pre-launch (review reach, price and readiness, nothing is spent) → Launch (go live) → Results. The Build section on this page is the first step; the rest have their own articles.

Getting participants

On the Audience step you choose how you'll get participants. By default, Sorted Research finds and recruits a matched audience for you, you say who you want and how many, and we handle the recruiting (you pay per response before launch). If you'd rather recruit people yourself, switch to bring your own respondents and use the participant link from the Launch step. See Sharing & launching for the full walkthrough.

Sort types

Choose how participants group the cards. This is locked once the study goes live, so pick carefully.

Cards

Cards are the items participants sort. Manage them in the study's Cards section.

• Add one at a time, or use Bulk add to paste a list (one card per line).
• Each card has a label (required) and an optional description (shown to participants on hover, useful for clarifying an item).
• Reorder cards with the up/down arrows. Delete with the trash icon.
• Card order is automatically randomized for every participant to reduce order bias. There's no setting to change this.
• You need at least 2 cards to launch a study.
• Card sort cards are text only. Pictures on the items being judged are a pairwise feature today (see pairwise); they are not available on a card sort, a Best–Worst study or a findability study.

Categories (closed & hybrid only)

For closed and hybrid sorts, define the categories participants sort into.

• Add, rename (click to edit), reorder, and delete categories.
• Categories appear to participants in the order you set. They are not randomized.
• Closed and hybrid studies need at least 1 category to launch. Open sorts don't use predefined categories.

Study settings

All settings auto-save as you edit (you'll see a brief "Saving…" indicator). Key ones:

• Title and Sort type (sort type locks once responses arrive).
• Client / folder: move the study between folders, or blank for unfiled.
• Target responses (optional, set on the Audience step): your sample-size goal. Drives percentage quotas and triggers the "study full" message once reached. Leave blank for unlimited. When you're buying responses, this is also the number of responses you're paying for (price per response × target).
• Sorting interface: how participants sort. Auto (default) uses the one-card-at-a-time layout on phones and the drag-&-drop board on desktop; Mobile forces one card at a time on all devices; Traditional forces the drag-&-drop board on all devices.
• Description (internal): private notes, never shown to participants.
• Participant instructions: shown on the welcome screen before sorting.
• Let participants add their own categories (hybrid only): on by default.
• Thank-you message: shown when participants finish.
• Survey-full message: shown when the target is reached.
• Redirect URL on completion (optional): sends participants to this URL ~2.5 seconds after finishing (must be http/https). Handy for sending people back to a recruiting panel. For tools like User Interviews, paste their redirect link exactly, keeping the word TRACKINGID in it, and we swap in each participant's tracking code automatically (see Sharing & launching).
• Participant consent notice: a sensible default is provided; edit to match your needs. Shown on the welcome screen.
• Privacy policy URL (optional): adds a "Privacy details" link to the welcome screen.

Response quality

A group of settings that reduce low-effort responses, where someone dumps a big pile of cards into two or three groups:

• Minimum groups: if a participant tries to finish with fewer non-empty groups than this, they get a friendly nag asking them to split their cards into more groups. If they carry on anyway, the response is flagged for your review (it isn't blocked). Set to 0 to turn the check off. For a closed sort the minimum can't be higher than the number of categories you provide, so it's clamped automatically. Defaults to 3. For a larger deck Graham suggests a higher floor beside this field (roughly one group per dozen cards, so about 7 for 79 cards), with a one-click "Use" so you don't have to guess; it is advisory and only nudges participants, never forces a number on them. Separately, a participant who finishes a big deck (20+ cards) in just one or two groups gets a gentle "that's very few, take another look?" even when you set no minimum, again with no target number, so honest coarse sorts are never pushed into a shape that isn't theirs.
• Minimum time (seconds) (optional): responses finished faster than this are flagged for review. Leave blank to turn it off. Timing is invisible to the participant.
• Sort thoughtfully note: shows a short note on the welcome screen asking participants to take their time, and warning that rushed responses may be removed.

Open and hybrid-with-custom sorts also nag when one group holds more than half the cards, asking the participant to break the big pile into smaller groups. All of these feed the quality flag you see in the analysis view, so you can exclude weak responses with a click and reject them in your recruiting panel. Separately, a participant who keeps placing cards faster than they can be read gets up to three "slow down" warnings and is then removed from the study, and if you bought them they are rejected with the recruiting panel so a replacement is sent. The same guard runs on all four study types.

Not available yet

There's currently no "duplicate study" or "save as template" feature. Set up each study fresh.