Read this one properly
A session collects one thing from a participant: a phone verification confirming they are a distinct person. That confirmation is not attached to their answers. There is no record anywhere linking a student to what they chose, which is why neither the institution nor Democrez can produce one — it does not exist to be produced.
During the session
Every item in the second column is absent by construction rather than by policy. Nothing here is a setting that could be switched on for a particular institution.
| Collected | Never collected | |
|---|---|---|
| Identity | A phone verification, confirming one distinct person | Name, roll number, email, department, year |
| The answers | Which option was chosen, and the locked prediction | Any link from an answer back to a person |
| The device | Nothing beyond what a browser needs to load a page | Device identifiers, advertising IDs, location |
| Behaviour | Nothing | Typing patterns, time spent per option, hesitation, revisions |
| Afterwards | Nothing. The session ends | Follow-up contact, marketing, or any use of the number |
A student should be able to answer as themselves. Every item above exists to make that possible, and a session that collected more would collect worse answers.
The one thing we ask for
It is the only piece of information a session requires, and the reason is arithmetic rather than identity.
Without verification, a single participant could answer several times and shift the batch's split on their own. Every figure in your report would then measure persistence rather than judgment.
Verification happens before the answering and is not carried into the response record. The two are separated at the point of collection, not filtered apart afterwards.
No number appears in your report, in any export, or in any conversation with us. There is no list of participants that the institution receives.
No follow-up messages, no marketing, no app prompts, and no addition to any contact list. The session ends and nothing continues.
The retention period is set out in the privacy policy: Raw verification documents: up to 30 days after successful verification. Verification result / compliance logs: up to 1 year unless a longer period is legally required.
Three parties
These three views never overlap. There is no fourth view with more in it.
The student
Their choice, their locked prediction, the verified figure and the distance between them. Nothing about any classmate, and nothing anybody else can see.
The institution
How the batch split, where it agreed, and where it collectively misread the wider public. Every figure is an aggregate with nothing underneath it.
Democrez
The same aggregate, plus the operational record needed to run a session. No view that connects a verification to a response exists on our side either.
Before anybody answers
This is not only an ethical point. A batch that felt required to participate produces compliant answers, and compliant answers are indistinguishable from honest ones in an aggregate.
What we ask of you
Participants are told what the session is, what is collected, and that declining carries no consequence. We supply the wording. If declining would in practice cost a student something at your institution, tell us — that changes whether the session is worth running.
What a student can do
Before the session or partway through. A participant who stops answering is not recorded as having declined, is not listed anywhere, and their partial answers do not enter the aggregate.
Less than you might expect, and for a reason that works in their favour. There is no per-student record to request a copy of, and nothing linking their answers to them that could be deleted — the link was never created. What they can ask about is the verification, and that is covered by the same retention decision under review above.
Full detail sits on delete your data and in the privacy policy, both of which apply to session participants as well as app users.
Questions
No. Verification confirms a distinct person and is not carried into the response record, so there is no link between a student and their answers anywhere in the system. This is not a permission that could be granted on request — there is no record to hand over.
No, for the same reason. There is no internal view that connects a verification to a response, because the two are separated at the point of collection rather than filtered apart later.
Confirm that a participant is one distinct person, so the batch's split cannot be shifted by somebody answering repeatedly. It is never shown to the institution, never used for messages or marketing, and never added to any contact list.
The retention period is in the privacy policy, which is under legal review, and we are not stating a figure here until that is settled. A number published on a marketing page has to match the number in the legal document, and quoting a provisional one would be worse than saying this.
No. No device identifiers, no advertising IDs, no location, and nothing about typing, hesitation, time spent per option or changed answers. A session that recorded how somebody arrived at a choice would be recording a different thing entirely.
Yes. There is no list of who participated or who declined, so nothing exists for a coordinator to compare against a register. A participant can also stop partway through, and their partial answers do not enter the aggregate.
No. The session ends and nothing continues — no follow-up, no app prompt, no newsletter. There is deliberately no mechanism for it, which is a stronger guarantee than a policy against it.
Yes, without variation. Nothing on this page is a default that an institution can negotiate, and a request for per-student output is declined regardless of what the session cost.
If something on this page is not enough for your institution, we would rather hear it before a session than after.
