A small group is useful only if the tutor can teach the same core idea without pretending that all students need the same amount of help. Primary 5 learners may be ready for the same passage while needing very different prompt levels for vocabulary, inference, evidence selection or writing transfer.
This preserved 2019 Yishun URL is rebuilt around one reader job: same text, different prompt levels. It is not a generic Yishun Primary 5 tuition landing page.
Historic Yishun address, tutor, telephone and enrolment claims from the old page are not carried forward. Families should verify current class location and schedule through current eduKate information. eduKate Singapore is an independent tuition provider and is not affiliated with MOE, SEAB or any school.
Shared Text Does Not Require Shared Support
Three Primary 5 students can read the same passage while operating at different levels of independence.
| Learner state | Possible prompt |
|---|---|
| Cannot enter the question | Restate the answer contract |
| Understands question, cannot localise evidence | Ask which paragraph job matches the question |
| Finds evidence but over-infers | Ask what the evidence does not prove |
| Independent | No prompt; add extension or transfer |
The lesson remains coherent because the text and underlying concept are shared. The support changes.
The Prompt Ladder
- Model: tutor demonstrates the process.
- Guided: several sequential prompts.
- Single discriminating question: one cue unlocks the next step.
- Location-only cue: tutor indicates where the problem is.
- Independent: student decides and executes alone.
- Extension: student defends, transfers or compares alternatives.
Progress should move toward lower support when the skill is becoming stable.
Why Identical Worksheets Can Hide Different Needs
If all three students complete the same worksheet with the same explanation, the final scores may hide the real learning states.
- One student may understand but work slowly.
- One may copy a peer’s evidence choice.
- One may know vocabulary but misunderstand the question.
- One may need no repair and should be challenged instead.
Prompt level is evidence. It tells the tutor how much of the process the learner currently owns.
Example: One Comprehension Passage, Three Jobs
Suppose the passage asks why a character changes a decision.
- Student A: tutor helps identify the decision-change paragraph.
- Student B: student localises independently but receives one question: “Which clue is strongest?”
- Student C: answers independently and compares two possible interpretations.
All three are learning from the same text without being forced into the same support state.
Example: Vocabulary
The group studies the same word in context.
- One learner chooses from two meanings.
- One explains why a near-synonym is wrong.
- One retrieves the word later in an original composition sentence.
The semantic object is shared; the cognitive demand differs.
Example: Writing
Three students can plan from the same prompt while receiving different scaffolds.
- Student A receives the causal-spine headings.
- Student B receives only the question “What causes the climax?”
- Student C plans independently and runs a deletion test on unnecessary events.
Prompts Should Target the Failed Layer
Good prompting is diagnostic. It does not simply make the task easier.
If the failure is vocabulary, pointing to the correct paragraph may not help. If the failure is evidence selection, reteaching grammar is irrelevant. The smallest useful prompt identifies the missing link.
Do Not Let the Strongest Student Become the Permanent Prompt
Peer explanations can help, but they can also hide dependence. Use individual first attempts, rotate speaking order and require independent rebuilding after discussion.
The final test should always ask whether each learner can perform without the peer present.
When Shared Material Stops Working
A shared text is no longer productive when:
- one learner cannot access the basic sentence meaning;
- another is chronically under-challenged;
- prompt levels become so different that the shared discussion disappears;
- one learner spends most of the lesson waiting.
At that point, regrouping or separate material is cleaner than forcing the group to remain together.
The Shared-Text Decision Matrix
| Condition | Decision |
|---|---|
| All can access text; prompt needs differ | Keep shared text, differentiate prompts |
| One needs major decoding support | Use alternative access/temporary separate task |
| One is consistently far ahead | Add extension or regroup |
| Peer discussion still adds reasoning value | Shared material remains useful |
Prompt Fading Is Part of the Lesson Plan
The tutor should know not only which prompt to give, but which prompt to remove next.
- full model → partial model;
- several questions → one question;
- one question → wait time;
- familiar example → fresh example;
- immediate success → delayed retest.
This makes independence measurable.
What Parents Can Ask About a Small Group
- Can the tutor explain my child’s current prompt level?
- What does my child still need help to notice?
- What support has already been removed?
- How is a fresh task used to test transfer?
- When would the group need to be changed?
Current Primary English Context
The MOE Primary English Language Syllabus 2020 develops students’ ability to apply language knowledge, skills and strategies across contexts. Differentiated prompt levels support that goal when scaffolds are used temporarily and then faded toward independent transfer.
Signs the 3-Pax Format Is Being Used Well
- Students share meaningful discussion without copying.
- Prompt levels differ visibly.
- Strong students receive extension, not waiting time.
- Weaker students receive targeted support, not simplified busywork.
- Fresh independent retests follow group discussion.
- Prompt levels decrease as competence grows.
Same Text, Different Prompts: Almost-Code Summary
SHARED_TEXT:
accessible_to_all?
FOR each_student:
diagnose_failed_layer()
choose_minimum_prompt()
test_fresh_example()
fade_prompt()
IF productive_ranges_diverge_too_far:
regroup_or_change_material()
OUTPUT:
real_small_group_differentiation
lower_prompt_dependence
stronger_transfer