All modules
People and contact

Placement test

Somebody writes to the school in September: „I want to enrol, B1 I think”. Reception books a call to check. The placement test turns that call into a link: the candidate solves a set, sees their level at once, and the funnel gets an enquiry with the score.

submissions 37
enquiries from the test 30
active tests 2
CandidateScoreLevel
Kuba Nowicki [email protected] 48 %A2 in the funnel
Maja Wrona [email protected] 88 %C1 in the funnel
auto the link is waiting for candidates
Placement test

What this module does

5
starting thresholds: CEFR from A1 to C1
20
thresholds at most in one test
200
submissions an hour to one test
500
submissions on Pro, 2000 on the School plan

Questions from an exercise section

The test has no question editor of its own, and that is the whole decision: it takes a ready exercise section, so fifteen exercise types and grading from the key work here exactly as they do for homework.

A threshold turns a percentage into a level

Thresholds are a table of rows: from which percentage, which level. The CEFR scale from A1 to C1 by default, but the school writes its own names, because not every course is measured in letters.

No account, one link

The candidate page stands at its own address and needs no login: details, consent with a copy of the text, questions, result. A switched-off test returns 404 and tells a stranger nothing about what is behind it.

The result lands in the funnel

A submission creates an enquiry in the first open stage, with the note „Placement test: 72 %, B1”. The same address with an open enquiry gets an event instead of a second row.

Results stay with the test

The Results tab keeps every submission: when, who, address, phone, percentage, level and a link to the enquiry. The level distribution says who the school really attracts.

Who sees what

A module is a permission, and which records may be touched is settled on write, not in the menu.

RoleSeesChanges
OwnerEvery test, its submissions, the level distribution and the link to the enquiry.Everything, including archiving and deleting a test or a result. The methodologist has the same.
Front deskThe same as the owner: the module comes by default, because it is funnel work.Creates tests, sets thresholds, the look and the link. Archiving and deleting stay with management.
TeacherNothing by default. A placement test is a sales tool, not a lesson tool.Once granted, works like the front desk: creates and edits tests, but does not delete.
StudentNo access to this screen. A candidate comes in through the link, with no account.Nothing. The module cannot be granted to a student or a guardian.

The details nobody asks about during a demo

The things that only surface in the third week of working with the module.

What this module does not do
  • a QR code for the link - QR comes to every public page at once or to none
  • a time limit - a clock kept by the browser of a candidate with no account guards nothing
  • an email with the result to the candidate - they see it on screen, and the school gets in touch
  • an adaptive test - questions belong to Exercises, not to a question model of its own

No. The test lives at its own /p/ address and opens without logging in. Details first: full name, email, phone if the school asks for it, and consent. Then the questions: all exercises at once or one exercise at a time, with a progress bar. The page and the preview on the Look tab are drawn by the same component, so the school sees exactly what the candidate sees, in the school's colour and with its logo.

From the exercise section picked when the test is created. The test has no editor of its own: content, exercise types and the key belong to Exercises, and points come from the same function that grades a student's attempt, so a new exercise type shows up here by itself. An open answer scores zero, because a person grades it and there is nobody here to do so. A draft section yields no questions, and a test without questions will not open at its link.

Through thresholds: each row says from which percentage which level starts. A new test gets the CEFR scale: A1 from 0 %, A2 from 30, B1 from 50, B2 from 70, C1 from 85. The school writes its own names and thresholds, twenty at most, and the lowest starts at zero so that every score gets a level. The candidate gets the highest threshold they reached. A test without thresholds returns the percentage alone.

Nothing. The level FREEZES in the submission when it is sent: the candidate has already seen it on screen, and moved thresholds have no right to change that. New thresholds apply from the next submission. A result cannot be edited by hand either - the Results tab is read only, because it is a snapshot of what the candidate actually answered.

A submission creates an enquiry in the first open stage of the funnel, with the channel chosen in the test settings and a note like “Placement test “Spring”: 72 %, B1”. When the same email already has an open enquiry, that enquiry gets a note in its history instead of a second row. An exhausted enquiry limit in CRM stops the entry into the funnel, not the result: the candidate sees their level, and the school finds it on the Results tab.

Full name, email, phone if the test asks for it, the score, the level, the answers, the IP address and a COPY of the consent text the candidate ticked. Without consent the result is not saved. The result is the proof of that consent, so only the owner or the methodologist can delete it. A deleted enquiry in CRM does not take the result with it: the row keeps the name from the moment it was saved.

Pro: three active tests and 500 submissions; the School plan: ten tests and 2000 submissions. Both limits are hard. A test switched off still counts, because it keeps its link and comes back with one click; only the archive frees the slot. Submissions count for the whole school, all tests together. When they run out, the test page answers exactly as for a switched-off test - the candidate learns nothing about the school's limits.

The Sharing tab offers two ways. The switch turns the test off, and the page then answers 404, exactly like an address that does not exist. “New link” draws a different address, and the old one dies at once. The address is random rather than built from the name, so it cannot be guessed. One test accepts at most 200 submissions an hour, and a hidden trap field filters out bots without a captcha.

See Cerium from the inside.

14 days on your groups and instalments. No card, no strings.