Tests and interviews workflow

What this covers and when you do it

This is the assessment middle of the admissions pipeline: it sits after staff review of an application and before an offer is issued. It exists so a school can select on evidence - a written entrance test, a panel interview, or both - and keep the decision trail attached to the application. Every scheduling and result action also emails the parent address on the application, so the family sees the same sequence the office does.

Work through it cycle by cycle, in batches rather than one applicant at a time. The test and interview records are one-to-one with the application: a second attempt to schedule a test for the same applicant is refused with the list of applications that already have one, so decide the batch before you schedule it. Reschedules and cancellations are edits to the existing record, not new ones.

Prerequisites

• A published admission cycle with applications in it, from /admissions/cycles and /admissions/setup.
• The Admissions committee configured at /committees, with a chair. Committee membership is what grants staff access, and the chair acts as admissions in-charge.
• SMTP working, checked at /settings/email-test. Test and interview scheduling, results, and shortlisting all send email to the parent address.
• Interviewers must be active staff users. Scheduling refuses a panel containing anyone who is not.

Test workflow

1. Open /admissions/tests and schedule a batch: pick the cycle and the applications, then set date, time, venue and optional instructions.
2. One test record is created per application with status SCHEDULED, each application moves to TEST_SCHEDULED, and each family is emailed the test details.
3. Applicants sharing a cycle, date, time and venue are grouped into one session, opened at /admissions/tests/sessions/[sessionKey], which is where you work through a room of candidates.
4. Record marks per applicant, or upload the whole batch keyed on application number. Percentage is computed from marks and maximum marks, the test moves to COMPLETED, and the application moves to TEST_COMPLETED.
5. Mark a no-show ABSENT, or CANCELLED if the test is called off. Both email the family, and cancelling reverts the application to its previous stage.

• Admit cards are a PDF per test, available only while the test is SCHEDULED or COMPLETED. Staff can download directly; a family can download by supplying the parent email and student date of birth as proof.
• The merit list is a read-only view over completed tests with a percentage, ordered by percentage then marks, optionally filtered by class and a cutoff percentage. It ranks candidates; it does not create shortlists.
• Bulk upload reports per-row errors rather than failing the batch. The usual one is a row whose application has no scheduled test.
• Internal notes recorded with a result stay internal - the result email carries marks, percentage and rank only.

Interview workflow

1. Schedule from /admissions/interviews: choose the cycle and applications, set date, time, venue, an optional meeting URL, and the interviewer panel.
2. The panel is stored on the interview and the first interviewer becomes the primary. Applications move to INTERVIEW_SCHEDULED and families are emailed.
3. Alternatively publish bookable slots with a date, time, venue and capacity, and let families choose. A family books by verifying parent email and student date of birth; a full slot is refused.
4. The panel records feedback: a score out of 100, remarks, a recommendation of STRONG_RECOMMEND, RECOMMEND, CONDITIONAL or NOT_RECOMMEND, and optional internal notes.
5. Saving feedback closes the interview as COMPLETED, stamps who conducted it, and moves the application to INTERVIEW_COMPLETED. The family receives the score, recommendation and remarks; notes are not included.

Shortlists

• Create a shortlist at /admissions/shortlists for a cycle, choosing a type of TEST, INTERVIEW or FINAL.
• Merit ranks continue from the highest existing rank for that cycle and type, assigned in the order you submit the applications - they are not recalculated from test or interview scores.
• Applications move to SHORTLISTED and each family is emailed with the shortlist type and rank.
• An application already shortlisted for the same type is reported as already shortlisted and skipped, so the response separates what was created from what was rejected.

Who owns each step

A committee member who is not the in-charge sees only the interviews they are assigned to in the interview list, which keeps a panel focused on its own candidates. Teachers who are not on the Admissions committee have no access to this module at all.

What breaks, and what the product does not do

• Scheduling a batch that includes an applicant who already has a test fails for the whole batch and names the offenders. Remove them and resubmit.
• Recording a result before the test exists is not possible; a bulk row without a scheduled test is returned as an error.
• Shortlisting before results are in produces a ranked list built on your submission order rather than on scores, because ranks are sequential, not computed.
• Nothing promotes candidates automatically. A cutoff on the merit list filters a view; someone still has to shortlist, and then issue offers at /admissions/offers.
• One test and one interview per application, one set of marks, no per-subject breakdown, and no seating plan or invigilator roster - that machinery exists for school exams, not admissions.