How to identify learning gaps in students

What a learning gap is

A learning gap is the distance between what a student was taught and what they can do on their own, on a question they have not seen before. It is narrower than "behind": a student can be behind overall and have two gaps, or on track overall and have one gap that will matter later.

Gaps build on each other. A student who cannot identify the main claim of a paragraph will keep struggling with summaries, essays and source questions that depend on it, and every one of those will look like a separate problem unless someone traces it back.

Three kinds of gap, three different responses

Telling the kinds apart matters because the fix is different for each. Repeating the same explanation louder works for none of them.

Why the grade alone does not tell you

A student who scores 68% on a unit quiz could be missing one skill badly or five skills a little. The grade is the same; what you would teach next is not. That is why many teachers say testing shows how far a student is behind, but not where.

The information is already in the quiz. It sits in which questions each student got wrong, and in the wrong answers they chose. Nothing new needs to be tested; the existing answers need to be read differently.

Step 1: give every question one skill

Take the questions you already use and write next to each the single skill it tests. Five to ten skills per unit is enough to act on; fifty is too many to read. One skill per question matters: a question tagged with three skills tells you little when a student gets it wrong.

Step 2: read answers by skill, and wait for a pattern

For each student, count right and wrong answers per skill. One wrong answer is noise; the same skill wrong on fresh questions, again and again, is a gap. A practical rule is to call a skill a gap only after a student has seen it three or four times on different questions.

Read the wrong answers too, not only the count. A student who keeps choosing the option that repeats words from the passage is usually matching words rather than reading for meaning. That points to one skill, inference, and to one reteach: model how to answer from what the text implies, not from what it repeats.

A real example: one gap behind three questions

In Gaplight's own history course, three different questions about prehistoric Ukraine pulled students towards the same wrong answer: the Eneolithic, the copper-stone age.

Asked when herding peoples came to dominate the steppe, 43% chose the Eneolithic and only 41% the right answer, the Early Iron Age. Asked when tools began to be made from an alloy of copper and tin, 33% chose the Eneolithic instead of the Bronze Age. Asked when the first pottery appeared, 28% chose the Eneolithic instead of the Neolithic.

Looked at one question at a time, these are three unrelated mistakes. Looked at by skill, they are one gap: the order of prehistoric periods and what defines each one. Copper came before bronze; herding began long before nomadic herders dominated the steppe. One lesson on that sequence answers all three.

Step 3: group students by the skill they miss

Once you know each student's gaps, sort the class by skill rather than by score. Six students who match words instead of reading for meaning need the same short lesson; that is one lesson instead of six separate conversations. Two students with the same 60% may end up in different groups, which is the point.

Keep the groups temporary. A group exists for one skill and dissolves when its students get that skill right on new questions.

Common mistakes when looking for gaps

• Reading the total score and stopping there. The total is where the question starts.
• Tagging each question with several skills, so a wrong answer cannot be traced to any of them.
• Calling a skill a gap after one wrong answer. Wait for a pattern on different questions.
• Checking the gap closed on the same questions students practised. Use questions they have not seen.
• Turning every check into a graded test. Keep checks short and low-stakes so students treat them as practice.

Why timing matters

Research on formative assessment, where teachers check often and adjust teaching to what the checks show, reports typical effect sizes of 0.4 to 0.7, larger than most educational interventions (Black & Wiliam). A later re-analysis of feedback across 435 studies found an average effect of 0.48, strongest when the feedback tells students what to do next (Wisniewski, Zierer & Hattie). The value comes from finding the gap before the unit test, not from knowing about it after.

Doing this without extra work

By hand, this means going back through every paper and keeping a spreadsheet, which few teachers have an evening for. Gaplight does it from the quizzes a school already gives: it reads every answer, maps it to the skill, and gives each teacher one list of who needs them and on what, before the unit test.