Most principals know which teachers are strong and which are struggling. The trouble starts when that feeling has to become a report: for the owner, for a salary decision, or for the teacher who asks, "On what basis?" A fair evaluation gives you something to point at.
This guide covers what to measure, a sample rubric you can copy, the traps that make a score unfair, and a simple process to follow each term.
Use two kinds of evidence, and keep them apart
- Records. Things that are already written down: attendance, punctuality, marks, homework, lesson plans, syllabus progress.
- The principal's observation. What you see in class and in the staff room, rated against a clear rubric.
Show them separately. If they are mixed into one number, nobody can tell whether a low score came from the registers or from the principal's opinion, and the teacher will assume the worst.
What to take from records
- Attendance and punctuality. Days present, days late. Leave that the school approved should not count against the teacher. Absence without leave should.
- Marks entered on time. Decide the deadline first, for example 7 days after the last paper, and tell everyone.
- Class attendance marked. For class teachers, how many days they marked their class attendance.
- Homework. How much was given, and how much of it was checked.
- Lesson plans. Whether there is a plan for each subject the teacher takes.
- Syllabus progress. How many weeks of the Scheme of Study have actually been taught, compared with how many should have been by now.
- Student results. Average marks and pass rate, handled carefully (see the traps below).
A sample rubric (an example, change it for your school)
Rate each question from 1 to 5. The useful part is writing down what each number means, so two principals would give the same mark for the same lesson.
| Question | 1 | 3 | 5 |
|---|---|---|---|
| Teaching method and classroom management | Students are often off task and the lesson has no clear shape | Most of the lesson is planned, some students drift | Clear aim, steady pace, the room is calm and busy |
| Subject knowledge | Makes errors that students notice | Sound on the textbook, unsure beyond it | Explains in more than one way and answers questions well |
| Student engagement | Only the front row takes part | Some questions asked, same few students answer | Most students speak, write or show work in each lesson |
| Homework and assessment | Rarely set or checked | Set regularly, checked late or only ticked | Set, checked with comments and used in the next lesson |
| Professional conduct | Often late to meetings, resists feedback | Does the job and follows rules | Helps colleagues and acts on feedback |
Five questions are enough. A form with twenty questions gets filled in the same way for every question.
Five traps that make a score unfair
1. Judging by average marks alone
The teacher given the weakest section will always look worse. A better check is to look at the same students in their other subjects. If the class is low in every subject, the problem may be the class. If only one subject is low, that deserves a closer look. Also compare with the earlier exam for the same students.
2. Treating missing records as zero
If the school never marked staff attendance, no teacher was "absent 100% of the time". Leave that part out of the score and say so. Starting to record it is a school task, not a teacher's fault.
3. Small numbers
Marks for eight students in one exam tell you very little. Mark such results as a rough guide and wait for more data.
4. Changing the rules afterwards
Decide the weights, the deadline and the rubric before the term starts. If the rules change after the results are known, teachers will see it as a way to reach a decision already made.
5. Ranking everyone from 1 to 40
A full ranked list makes staff rooms tense and hides the real information. Three groups, such as strong, good and needs attention, are easier to act on. Rank only teachers who do a similar job.
How to weigh the parts
There is no correct split. As an example, a school might give 30 percent to teaching practice, 30 percent to student results, 20 percent to syllabus and lesson plans, 10 percent to attendance and punctuality, and 10 percent to the principal's rubric. A school that values observation more could raise the rubric. What matters is that the split is written down before the term and the teacher can see it.
A simple process for each term
- Before the term. Share the rubric, the weights and the deadlines with all teachers.
- During the term. Visit classes. Short visits are better than one long one. Write two or three lines after each visit while you still remember it.
- End of term. Pull the record numbers. Rate the rubric. Write what the teacher does well, what needs to improve, and one to three goals.
- Talk to the teacher. Start with strengths, then the numbers, then the goals. Let the teacher explain a low number. Sometimes there is a real reason.
- Lock the report. Once finalized it should not change quietly. If it has to be reopened, that should be written down.
- Next term. Begin by looking at last term's goals.
What an evaluation should not decide alone
A score is a starting point for a conversation. It should not be the only reason for ending a contract or cutting pay. Look at the reasons behind the numbers first.
Doing this in Skoo
Skoo works out the record numbers from what your school already enters, so nobody types them again. A part with no data says so instead of showing zero, a result based on few marks is flagged, and results are compared with the same students' other subjects. The principal rates the rubric, writes the comments and finalizes the report, which is then locked, and reopening it goes into the audit log. Your school sets the weights, the period, the rubric questions and whether teachers can see their own report. AI can draft the written comments, but it never sets a score.
See how it works on the teacher evaluation page. The records it uses come from attendance, exams and the Scheme of Study.
