Colleges and organisations
For institutions.
An institution runs one bounded session — 45 to 60 minutes with a single batch, verified by phone. Each participant sees their own Gap privately. The institution receives a behaviour report describing how the group answered. Once the public record holds sealed readings, that report also sits against national figures. No individual is named, scored or ranked in it, and that is not a setting anybody can change.
A Human Judgment Lab.
A session is a single sitting. A batch of students works through a set of situations in a browser — no app installation, no accounts to create afterwards, nothing running in the weeks that follow. Each person answers alone, locks a prediction about what everyone else chose, and meets the verified figure. That is the same instrument the public app runs, with the same rules.
What differs is the second output. Alongside each person's private reading, the session produces an aggregate picture of the batch: where it split, where it agreed, and where it collectively misread the wider public. That picture is what the institution receives.
What it is not
It is not an assessment, and nothing in it can be used as one. There are no correct answers to grade against, no per-student output, and no way to identify how a particular participant answered — not for a coordinator, not for a placement officer, and not for us. A session that could be graded would produce answers written for the grader, which would make the exercise worthless in both directions.
It is also not a recurring engagement. One session per batch, bounded and finished. There is no dashboard to log into afterwards, no ongoing student tracking, and no reason for anybody to keep using it after the session ends.
Every participant gets something of their own. The report is a by-product of the session, not the reason a student sits through it — which is why the answers are worth reading.
Both sides
What you get,
and what we get.
Stating this plainly is more useful than pretending the arrangement is one-directional. Both columns are real, and you should weigh the second before agreeing to the first.
| The institution receives | Democrez receives | |
|---|---|---|
| Immediately | A session your students find genuinely interesting | Verified, deduplicated responses from real people |
| Afterwards | A behaviour report for the batch as a group | Nothing further. The session ends |
| Comparison | Your batch's own picture — plus national comparison once the public record fills | A batch whose composition is known and bounded |
| About individuals | Nothing. No names, no per-student output | Nothing. No identity is stored with an answer |
| Ongoing | No dashboard, no tracking, no subscription to manage | No continuing access to your students |
What the report actually contains, field by field. What you get
Where the data goes
Your session does not
feed the public record.
This is the structural commitment that matters most, and it runs in both directions.
Institution cards stay institutional
Situations written for a particular institution feed that institution's report only. They never enter the public corpus, and a card cannot be moved from one pipeline to the other.
Your batch is not published
No reading identifying your institution appears on the public readings page. If a national figure ever draws on session responses, it does so anonymously and without naming any institution.
No student is retrievable
Verification confirms a distinct person; it is not stored against an answer. There is no lookup that returns what a participant chose, for anybody.
Small groups are not broken out
A subgroup too small to be reported without risking inference is not reported at all — however specific a breakdown a coordinator asks for.
Why we say no to the obvious requests
The requests that come up are reasonable ones: a per-student summary for counselling, a breakdown by section, a list of students whose answers stood out. Each of those would be useful to somebody with good intentions, and each is refused for the same reason — a session that can produce them is a session students should answer strategically in, and then it measures nothing.
If you need per-student assessment, this is the wrong instrument and we would rather say so before a session than after one.
Read before you book
Four things worth
knowing first.
In this order. The privacy page is the one to read most carefully, because it is the one your students would want you to have read.
The session
How a Human Judgment Lab actually runs
What happens in the room, what you need to provide, and what the 45 to 60 minutes are spent on.
The report
What the Student Behaviour Insight Report contains
Field by field, including what it deliberately leaves out and why.
Student privacy
What is collected, and what cannot be retrieved
Read this one properly. It is what a student would want you to have checked.
Talk to us
Booking a session, and what to ask us
Including the questions we would push back on if we were sitting on your side of the table.
Fit
Who this suits,
and who it does not.
A good fit
You want the batch to learn something about itself
A group that has just discovered it disagrees — about workplace credit, about family obligation, about what a stranger is owed — has something to discuss that nobody prepared. That conversation is the point, and the report gives it structure afterwards.
Not a fit
You want to evaluate individual students
Nothing here produces per-student output, and no arrangement can add it. If the requirement is assessment, screening or counselling input on named individuals, this instrument cannot serve it and would waste your session slot.
Questions
Straight answers.
How long does a session take, and what do we need?
One sitting of 45 to 60 minutes with a single batch. Participants need a browser and a phone for verification; there is no app to install and no account that persists afterwards. Everything else is on our side.
Can we see how a particular student answered?
No. Verification confirms that a participant is a distinct person; it is not stored against their answers. There is no lookup that returns an individual's choices — not for a coordinator, not for a placement officer, and not for us.
Can the report be broken down by section or department?
Only where each subgroup is large enough that no individual could be inferred from it. Below that size the breakdown is not produced, because a small enough group makes an anonymous report identifiable in practice even when it is anonymous on paper.
Does our students' data go into your public dataset?
Situations written specifically for your institution feed your report only and never enter the public record. The two pipelines are separated structurally rather than by policy, and a card cannot move between them.
Is this an assessment we can use for placement or counselling?
No, and we would rather say so before a session than after. There are no correct answers to grade against and no per-student output. A session that could be graded would collect answers written for the grader, which would make the whole exercise worthless.
What does it cost?
Pricing depends on batch size and is discussed directly rather than published here, because a figure on a page tends to become a figure nobody revisits. We would rather quote against what you actually need than have you plan around a number that may not apply to you.
Do students have to agree to this?
Yes. Participation is explained before the session and a participant can decline without consequence to them. A batch where people felt compelled to answer would produce compliant answers rather than honest ones, which is a problem for the data as well as for the student.
What happens after the session ends?
It ends. There is no dashboard to maintain, no student tracking, and no subscription. You receive the report; participants keep whatever they took from their own readings; nothing continues in the background.
Talk to us about your batch.
Tell us the size and the setting. We will tell you plainly whether this is the right instrument for it.