Office Discipline Referral: Form Template and Major/Minor Definitions

Two rows do most of the work in reducing referral volume. Dress code and tardies belong in the attendance and dress-code tracks respectively, not in the discipline pipeline, and moving them out typically removes a large share of referrals at a stroke. The other high-leverage decision is defining “refusal” as escalating rather than binary — a referral on the first refusal removes the teacher’s own repertoire from the equation and produces a referral pattern that looks like a student problem when it is a support problem.

The form

Twelve fields, one side of paper or one screen. Longer forms produce shorter answers.

1. Student, grade, date, time.
2. Referring staff member and role.
3. Location — from a fixed list, because free text destroys the analysis.
4. Behaviour — one primary category from a fixed list, with the definitions printed on the form itself.
5. Description — two or three observable sentences with a direct quotation.
6. Antecedent — from a fixed list, plus free text. What was happening immediately before.
7. Possible motivation — from a fixed list: obtain attention, obtain an item, avoid a task, avoid a person, unclear. Guessing is fine here; the aggregate is what matters.
8. What the teacher already tried, from a fixed list. This field turns a complaint into a request for support and changes how the office responds.
9. Others involved.
10. IEP, 504 or behaviour plan flag, and whether the plan was followed.
11. Family contacted by the teacher? Yes with date, or no.
12. Administrator section: action taken, consequence, family contact with channel and outcome, date closed.

Fields six through eight are what separate a referral system that generates interventions from one that generates suspensions. Without them, the office sees an event; with them, it sees a pattern with a function.

Definitions on the form, not in a binder

The most common cause of unusable discipline data is that categories are undefined, so “disruption” means one thing to a first-year teacher and another to a twenty-year veteran, and the difference shows up in a state report as a disparity.

Print one-line definitions on the form itself. For example: Defiance — refusal to follow a direction after it has been given twice, with a stated choice offered. Disruption — behaviour that stops instruction for the class, not behaviour that requires redirection. Physical aggression — physical contact intended to harm; distinguish from rough play and from contact during a separation. Harassment — conduct based on a protected characteristic; this triggers a separate policy route, not an ordinary referral.

Then calibrate. Once a term, take ten anonymised scenarios to a staff meeting and have everybody classify them independently. Where the answers scatter, the definition needs work. Districts that do this see referral rates converge across teachers within a year, and the equity data becomes worth reading.

What the office does with it

Three responses, and the referral form should make the choice easy.

Consequence and return. Appropriate for a genuine major with no pattern. Includes same-day family contact and a note back to the referring teacher, which is the courtesy most often skipped and the one that determines whether teachers keep filling the form in properly.

Support and return. Where the pattern in fields six through eight suggests a function — task avoidance in one subject, a peer conflict at one time of day — the response is a change in conditions rather than a consequence, and the teacher needs to be told what changed and why.

Escalate. Where the count is climbing, the team convenes. A useful hard rule: at the fourth referral in a term, or the second involving physical contact, or any referral for a student with an IEP where the plan was not followed, a meeting happens rather than another consequence.

Feed the aggregate back to staff monthly: referrals by location, by time of day, by category, by referring staff member, and by student group. The two findings that recur everywhere are that most referrals come from a small number of classrooms, and that a large share cluster in unstructured time in one or two locations. Both are fixable without touching a single student’s file.

Family contact after a referral

A referral that does not produce contact with the family produces nothing. Three practices.

Same day for majors, before the student gets home. A family hearing it from the school first responds differently from a family hearing it from their child.

A short factual message, then a call. The message says who is calling and why; the call does the work. Kastr sends SMS, email and voice, with terminal SMS failures failing over to voice and row-level idempotency so a retry does not call twice. Selecting a student expands to that student’s guardians.

Balance the ledger. A school where the only contact families receive is a referral call has a communications problem, not a discipline problem. Track the ratio of positive to corrective contacts per student; anything worse than one to one for the students who receive the most referrals is worth fixing directly.

What Kastr will not do: no rule fires a message when a referral is entered, there is no SIS or behaviour-system connector, and there is no grade, class or school targeting — audiences resolve to specific people or everyone. Referral follow-up is composed by a person and sent to a named list. Cost-and-reach preview shows what it will cost before it goes.

Questions people actually ask

Related