Find the Gap: How to Identify Misunderstandings from Student Answers
"Knowledge gap finder" is a tool that analyzes student answers and helps see exactly where misunderstanding is hidden, then suggests targeted practice. Unlike checking "correct or incorrect," it looks at formulations, the reasoning process, and typical errors. In this article, three fictional learning scenarios are presented: mathematics, science, and language. For each, an example request is given, a useful result analysis is shown, and a way to check that the hint truly points to a gap, not a random slip. All examples are educational and do not replace teacher work.
Open tool: Knowledge gap finder ↗How the Tool Works and What to Input
"Knowledge gap finder" accepts two text fields: source material and goal or desired result. It is convenient to insert student answers, a fragment of their reasoning, or a list of solved problems into the first field. Into the second - what you want to understand: for example, "identify misunderstanding of percentages" or "select exercises on tense agreement". The tool returns one textual result that can be copied or downloaded as TXT.
It is useful to pre-format answers to be readable: remove unnecessary marks, preserve numbering of problems, and leave exactly those fragments where the thought process is visible. The more specific the source material, the more accurate the hint. If there are many answers, it is better to split them into meaning blocks and check one by one, rather than mixing topics in one request.
The tool does not grade and does not replace the teacher. It formulates hypotheses about possible gaps and suggests exercises. Any hypothesis should be verified in a live dialogue with the student: ask them to explain a step in their own words or solve a similar problem.
Source material: "Problem 1. 20% discount, then another 10%. I added the percentages and got 30%. Answer: 30%." Goal: "Understand if there is a misunderstanding of sequential percentages, and select practice."
Scenario 1: Percentages and Sequential Discounts
The student confidently adds percentages, perceiving them as ordinary numbers. This is a common misunderstanding: a percentage is a share of the current value, and the second discount is calculated from the new price. The tool can point out this gap and suggest tasks where it is important to count step by step, not as one action.
A good request contains not only the incorrect answer but also the student's explanation. Then it is visible that the error is not random: the student builds reasoning on a wrong model. As a result, it is useful to receive not a ready solution, but a series of short tasks: first with round numbers, then with a price that changes after the first discount.
Checking the result is simple: ask the student to solve a problem about 100 rubles with a 20% discount, and then another 10%, and verbalize each step. If they calculate the second discount from the original sum, the gap is confirmed. If the error disappeared, it was likely due to haste, not understanding.
Request: "Source material: student writes that 20% and 10% give 30%, because percentages are added. Goal: identify misunderstanding of sequential percentages and suggest 3 exercises with increasing difficulty." Useful result: explanation of the difference between sum of percentages and sequential application, then tasks on price 200, 500 and 1000 rubles.
Scenario 2: Science and Cause Instead of Consequence
In science, students often confuse cause and effect: for example, asserting that a plant grows because it receives more light, although in the specific experiment the light was the same, and the amount of water changed. The tool helps notice that a factor is substituted in the answer and suggest tasks for comparing conditions.
The source material should include the answer itself and a brief description of the experiment. Then the hint will be based on a specific discrepancy, not on general words. A useful result usually contains a question-prompt: which factor actually changed and how to check it.
Check the result by asking the student to formulate what was the same and what was different in the experiment. If they can name the variable and control, the misunderstanding is likely point-specific. If they again name the wrong factor, additional practice on planning observations is needed.
Request: "Source material: student writes that the plant grew faster because it stood closer to the window, although both plants stood on the windowsill, but one was watered more frequently. Goal: identify confusion of cause and effect, suggest practice on control variables."
Scenario 3: Language and Tense Agreement
In language, the gap often does not manifest in a separate word, but in the logic of the sentence: the student chooses a tense based on the meaning of the Russian translation, not on the agreement rule. The tool can show that errors repeat in similar constructions and suggest exercises on choosing a tense depending on the main part.
It is useful to provide not one sentence, but three or four similar ones, so the tool sees a pattern. In the goal, indicate that an explanation of the rule and short practice is needed, not a corrected text. This way the result will teach, not just give a ready answer.
Check: give the student a new sentence with the same construction and ask them to explain the choice of tense. If they refer to the rule, not a guess, the gap is closed. If they again orient themselves on translation, it is worth returning to basic cases.
Request: "Source material: three sentences, where the student puts past tense after "he said that...". Goal: identify misunderstanding of tense agreement and suggest 4 exercises with explanation." Useful result: brief rule, then sentences for choosing tense and a question on explanation.
How to Formulate a Request to Get Value
Start with a precise goal: not "help with mathematics," but "identify if the student understands that percentage is taken from the current value." Then add the source material exactly in the form the student wrote it, preserving their formulations. This is important because the tool searches for a gap between explanation and rule.
Indicate the desired result format: list of hypotheses, three exercises, questions for checking. The tool returns text, so the clearer you describe what is considered useful, the easier it is later to distinguish the hint from general words. Do not ask for grading or score forecast: this is not part of the task.
If the result seems too general, narrow the source material to one topic and one type of error. If, on the contrary, the hint points to several gaps, choose one and check it separately, before moving to the next.
Request template: "Source material: [student answers]. Goal: identify misunderstanding [topic] and suggest [number] exercises with increasing difficulty. Format: hypothesis, explanation, exercises, checking questions."
Checking the Result and Limitations
The "Knowledge gap finder" result is a hypothesis, not a diagnosis. Check it with a short dialogue: ask the student to explain a step, provide their own example, or solve a similar problem. If the explanation matches the hint, the gap is confirmed; if not, the tool may have seen a random error.
The tool works with text and does not analyze audio, images or video. It does not automatically save results, does not publish them and does not send by email. Model responses may contain errors, and queue availability may vary, so important conclusions are worth manually verifying.
Do not use hints as the sole source of study plan. They work well as a starting point for conversation and exercise selection, but the decision of what the student needs next remains with the teacher. For finance, health and rights, material should remain educational and not turn into personal recommendation.
Checking question after result: "Explain why the second discount is calculated from the new price, and solve a problem with price 400 rubles." If the student copes without the hint, the hypothesis about a gap is not confirmed.
"Knowledge gap finder" is useful where it is not enough just to correct an answer, but to understand what reasoning model led to the error. Three scenarios show the general technique: give a precise goal, preserve the student's formulations and ask for a hypothesis with exercises. Then check the hypothesis with a live question. Such a cycle - request, analysis, check - helps turn the tool's hint into real educational work, not a ready answer the student will just rewrite.
Open tool: Knowledge gap finder ↗Interface translations are machine-assisted and may need refinement.