MindOS · Comparison State · Align → Notice → Explain → Discriminate → Generalise → Transfer
Wait, What? Two Questions Can Look Different and Be the Same Problem
Students often classify by surface appearance. Different numbers, different story, different diagram: different problem.
But expertise frequently depends on seeing through the surface. Two questions can look different while sharing the same underlying relationship. Two answers can look similar while depending on completely different reasoning. Two methods can both work, but only one may be efficient under the present conditions.
Comparison is therefore not merely “spot the difference.” It is a learner operation for discovering what stays the same, what changes, and which difference actually matters.
Quick Answer
The Comparison State asks whether the learner can place two or more cases into alignment, identify meaningful similarities and differences, explain why those features matter, and use the comparison to improve later selection or transfer.
CASE A + CASE B ↓ ALIGN THE RELEVANT PARTS ↓ WHAT IS THE SAME? ↓ WHAT IS DIFFERENT? ↓ WHICH DIFFERENCE CHANGES THE ROUTE? ↓ EXPLAIN WHY ↓ FORM A RULE OR DISTINCTION ↓ TEST ON CASE C
The Owned Job of This Page
This page owns comparison as an explicit learner operation. It does not own interleaving, which mixes practice to force selection, or Strategy Selection, which decides which known route applies. Comparison happens when the learner deliberately brings cases, explanations, examples or methods together and asks what relation between them is informative.
Why Comparison Can Change What the Learner Sees
A single example can be memorised as a story. Two carefully chosen examples can reveal a structure.
Research on contrasting cases has repeatedly explored this idea. A 2025 study in Instructional Science examined how the salience of contrasting cases affects learning before and after instruction, while earlier work has shown that comparison is most powerful when learners are prompted to explain the significance of similarities and differences rather than merely inspect them. See the 2025 ERIC record on contrasting cases and the study How do contrasting cases and self-explanation promote learning?
The educational point is modest but useful: comparison can direct attention toward relational structure that is easy to miss when examples are studied one at a time.
Observable Signs of a Fragile Comparison State
- The learner says two problems are different because the numbers or wording changed.
- They say two answers are “the same” because the final number matches, even though the reasoning differs.
- They can list differences but cannot explain which difference matters.
- They compare everything at once and become overloaded.
- They copy a teacher’s comparison categories but cannot generate their own.
- They can compare familiar examples but fail when one surface feature is deliberately misleading.
Discrimination Test 1: Surface or Structure?
Give two questions with different surface stories but the same underlying relation. Ask: “What is the same underneath the wording?”
Then reverse the test: give two questions that look almost identical but require different routes because one critical condition has changed. Ask: “What small difference changes what we should do?”
If the learner succeeds only when the teacher points to the relevant feature, comparison is still scaffolded rather than independent.
Discrimination Test 2: Comparison or Recall?
A learner may appear good at comparison because they memorised the teacher’s sentence: “Method A is faster when…”
Change the examples. Remove the labels. Ask the learner to generate the comparison from the cases themselves. If the distinction survives, the comparison operation is becoming portable.
Discrimination Test 3: Comparison or Strategy Selection?
Comparison asks: “How are these cases or methods alike and different?” Strategy Selection asks: “Given this new case, which route should I use?”
The operations support each other, but they are not identical. A student can make an excellent comparison and still hesitate when no paired example is present. That is where the Strategy Selection owner takes over.
Intervention 1: Align Before Comparing
Comparison fails when the cases are not aligned. Put corresponding parts side by side.
- problem condition beside problem condition;
- step beside step;
- claim beside claim;
- evidence beside evidence;
- diagram feature beside diagram feature;
- sentence function beside sentence function.
This reduces irrelevant search and makes the relation inspectable.
Intervention 2: Ask for the Consequence of the Difference
“What is different?” is often too shallow. Follow with: “So what?”
- What does this difference change?
- Does it change the method?
- Does it change the interpretation?
- Does it change the evidence strength?
- Does it change whether the rule applies?
Now comparison becomes causal and operational rather than decorative.
Intervention 3: Use Correct and Incorrect Examples
Contrasting a correct route with a plausible error can make a hidden distinction visible. A 2025 systematic review of erroneous and contrasting erroneous examples found that these formats can support learning, but their effectiveness depends on prompts, explanations, learner knowledge and task complexity. See Conditions for Effective Learning from Erroneous Examples.
The MindOS rule is not “show more mistakes.” It is: choose a mistake that exposes a useful distinction, then require the learner to locate, explain and repair it.
Intervention 4: Let the Learner Generate the Categories
At first, a tutor may provide comparison headings. Later, remove them.
TEACHER GIVES CATEGORIES ↓ LEARNER FILLS THEM ↓ TEACHER GIVES ONE CATEGORY ↓ LEARNER GENERATES THE REST ↓ LEARNER CHOOSES WHAT TO COMPARE ↓ LEARNER EXPLAINS WHICH DIFFERENCE MATTERS
The fade matters because real problems do not arrive with a comparison table already designed.
Comparison in English, Mathematics and Science
- English: compare two inference answers, two introductions, two tones, or two uses of evidence.
- Mathematics: compare two solution methods, two problem structures, or one correct and one attractive wrong route.
- Science: compare two mechanisms, experiments, variables, graphs, or explanations while identifying which relationship remains invariant.
Evidence Boundary
Comparison is not automatically beneficial. Poorly chosen cases can create confusion. Too many dimensions can overload the learner. Merely presenting two examples side by side does not guarantee that the relevant relation will be noticed. Prompts and explanation often matter.
So MindOS treats comparison as a designed operation, not a universal trick.
Transfer Test
- Remove the paired example and give a new case.
- Ask which earlier case it resembles structurally and why.
- Introduce a misleading surface similarity.
- Ask the learner to invent a contrasting case.
- Ask what single feature would change the correct route.
If the learner can generate and use the distinction without the original pair, comparison has become part of independent navigation.
Examination Implication
Examinations rarely say, “Compare this with the example you studied last Tuesday.” The learner must recognise relevant structure internally. Good comparison practice should therefore end with unlabelled, mixed and time-bounded cases.
Return Test: What Came Back?
- Can the learner identify a meaningful comparison without being told what to compare?
- Can they separate surface from structural differences?
- Can they explain why one difference changes the route?
- Can they invent a useful contrasting case?
- Can they use the distinction later during strategy selection?
Parent and Tutor Teaching Guide
When two examples are available, resist the urge to explain the distinction immediately. Ask the learner to align them first. Then ask three questions: What is the same? What is different? Which difference matters?
The final step is always another case. Otherwise we may have taught one comparison rather than built the ability to compare.
MindOS Direction Graph
COMPARISON STATE ├── Cannot align cases? → REPRESENTATION / LOAD ├── Notices only surface? → CONTRAST STRUCTURE ├── Lists differences but cannot explain? → SELF-EXPLANATION ├── Too many dimensions? → REDUCE COMPARISON FIELD ├── Knows distinction but cannot choose? → STRATEGY SELECTION ├── Can compare paired cases only? → FADE PAIRING └── Uses distinction on a new case? → TRANSFER
Continue Through MindOS
MindOS boundary: Comparison is an educational operation, not a diagnosis of intelligence or cognitive ability. Use observable performance and changed task conditions to decide what support the learner needs next.
