A dashboard can help a teacher notice a pattern, but it cannot explain the student. The safest question is not “Who is weak?” but “Which learner may benefit from a closer look, and what evidence should I check next?”

Start with participation before performance

Separate students who have not started, started late, or stopped midway from students who completed the work and struggled. The first group may face access, time, instructions, login, device, or motivation barriers. A score cannot explain a task the learner did not meaningfully attempt.

Look for converging signals

One unusual number is rarely enough. Give attention to combinations: low completion across several assignments, repeated attempts without improvement, heavy hint use on the same concept, or one question missed by many students. A pattern across time is more useful than a single bad day.

  • Progress: where did work stop?
  • Accuracy: is difficulty broad or tied to one concept?
  • Attempts: do retries show improvement?
  • Hints: did support lead to an independent answer?
  • Context: what do classwork and conversation add?

Check the activity before judging the learner

Open the student view and inspect the exact question. Ambiguous wording, an incorrect accepted answer, inaccessible media, or an unexpected language level can create a class-wide problem. If many students fail at the same point, investigate the item and instruction first.

Choose the smallest useful response

A signal should lead to a proportionate action: a quiet check-in, one worked example, a shorter practice set, a peer explanation, or a corrected question. Avoid assigning large amounts of extra work simply because the dashboard is red. Record what you tried so the next review has context.

Protect dignity and teacher judgment

Use analytics privately and describe observations, not identities: “three attempts with the same fraction error” is actionable; “low student” is a label. Never let an automated summary make a high-stakes decision. The teacher, learner, and evidence together determine the next step.

The five-minute scan

Scan for non-starts; inspect stalled progress; compare attempts and hints; open one common missed item; choose one student or group for a small response; and schedule a later check. The goal is not perfect diagnosis—it is earlier, kinder attention.