Understanding Punggol Additional Mathematics, SEC 2027: reconstruct quadratic functions from roots, vertices, three points, symmetry, slope and models. These are national K232 G2 and K341 G3 methods, taught with original exact examples.
For the complete series guide to reconstructing quadratic functions, read the primary Punggol Additional Mathematics quadratic reconstruction article. This companion provides extra examples, data-consistency checks and practice tasks.
Start with the information the question actually gives
Two real roots r and s imply y=a(x−r)(x−s), for a≠0. A vertex (h,k) implies y=a(x−h)²+k. Three known values f(x) at distinct x inputs can be used to determine a,b,c in y=ax²+bx+c. These forms are mathematically equivalent once their parameters are determined.
The best form reduces the number of unknowns, but any valid form must reproduce all the original conditions. Write down whether a condition describes a point, a root, a turning point, a gradient or an interval before substituting. An examination command might request the equation, coefficients, an interval or proof of impossibility.
National 2027 scope and why this is a bridge
Both 2027 SEC G2 K232 and G3 K341 Additional Mathematics include quadratic A1 topics: maximum/minimum using completed squares, positivity/negativity conditions and quadratic models. G2 prerequisite knowledge explicitly includes simple quadratic graph forms. Reconstructing a curve uses those ideas together; it is not advertised here as a separate compulsory national examination paper.
Punggol follows Singapore’s national syllabus. A tutor should confirm the student’s registered subject code and school’s order of teaching. If a child is unable to factor x²−5x+6 or read an ordered pair, repair that earlier skill before presenting a three-variable coefficient system.
Two roots give two factors
Suppose the x-intercepts are one and four. Since y(1)=y(4)=0, write y=a(x−1)(x−4). The signs in each factor are dictated by its zero, so x+1 would be inappropriate for root one. The multiplier a controls vertical scale and opening direction.
One cannot set a=1 simply because two roots are given. For a=1 and a=2, both curves have the same zeros but different values at other x inputs. To determine a, use a further independent point, a vertex value or another suitable condition.
Worked example: roots and y-intercept
If the curve with roots one and four passes through (0,8), substitute 8=a(−1)(−4)=4a. Thus a=2 and y=2(x−1)(x−4)=2x²−10x+8. Checking x=0 yields eight; checking x=1 and four yields zero.
The axis is the average of the roots, x=5/2. At that x value, y=2(3/2)(−3/2)=−9/2, so the curve has a minimum at (5/2,−9/2). This third representation independently checks the root-and-scale calculation.
Axis symmetry is an average, not a guess
For a quadratic with distinct real roots r,s, the axis is x=(r+s)/2. Root locations −2 and four therefore give axis x=1, not x=0. This follows algebraically by expanding a(x−r)(x−s) and evaluating the vertex coordinate −b/(2a).
The scale a disappears from the axis calculation. Thus a positive and negative scaling of the same factorised expression have identical zeros and axis, but one opens upward and the other downward. The height of the vertex needs a separate scale condition.
A vertex calls for completed-square form
If the turning point is (h,k), begin with y=a(x−h)²+k. For vertex (2,−3) and point (0,5), 5=4a−3 gives a=2. The curve is y=2(x−2)²−3=2x²−8x+5.
Completed-square form makes the extremum immediate: when a>0 the vertex is a minimum, and when a<0 it is a maximum. A student writing (x+2)² for a vertex with x-coordinate two has misread the difference x−h. Substitute the vertex input to expose the error.
A repeated root is a vertex on the axis
If the graph touches the x-axis at x=3 with a repeated root, its form is y=a(x−3)² and its vertex is (3,0). Given y(0)=18, solve 18=9a and obtain y=2(x−3)².
A repeated root contributes multiplicity two but only one distinct x-intercept. Listing x=3 twice as two different values confuses algebraic multiplicity and location. A learner should state both the factor and the number of distinct roots if the question makes that distinction.
Three points yield simultaneous coefficient equations
Take points (0,1),(1,2),(2,7). In y=ax²+bx+c, the first gives c=1, the second gives a+b=1 and the third gives 2a+b=3. Solving gives a=2,b=−1,c=1, hence y=2x²−x+1.
Substitute each input into the recovered equation; all three outputs must match. Its discriminant is −7, so it has no real x-intercepts, demonstrating that three exact points can determine a quadratic even if no roots were given.
Negative inputs require complete brackets
For points (−1,6),(0,1),(2,3), c=1, a−b=5 and 2a+b=1. Therefore a=2,b=−3 and y=2x²−3x+1. It gives six at x=−1 because (−1)² is positive one.
A student who writes a(−1)²=−a has misapplied signs before even beginning simultaneous equations. Revisit squares of negative numbers and use explicit brackets. The function factors as (2x−1)(x−1), giving a further independent check.
Axis plus one root determines its partner
If the axis is x=2 and a root is x=5, the other real root must be x=−1 so their average is two. Hence y=a(x−5)(x+1). If the curve passes through (0,−10), then −10=−5a and a=2.
This is symmetry about x=2, not reflection about the y-axis. A number-line diagram marking equal distances left and right of two can repair the common mistake of choosing −5 as the second root.
A root and a nonzero value at that root are inconsistent
If x=1 is a root, f(1)=0 by definition. An additional assertion that (1,3) lies on the same curve is contradictory. For y=a(x−1)(x−4), substitution gives 3=0 for every a.
No value of the scale can repair the disagreement. A pupil should be willing to state that no such function exists. Detecting inconsistent data is a valid mathematical result, not a refusal to work.
A repeated condition may add no information
Two roots at one and four already imply points (1,0) and (4,0). Giving point (1,0) again does not determine a in y=a(x−1)(x−4). Infinitely many nonzero values of a remain feasible.
Compare a useful third point such as (0,8): it gives 8=4a and determines a. Ask whether each new condition actually eliminates possibilities, rather than counting how many statements appear on the page.
Three collinear points can force a=0
Three distinct x-values determine a unique polynomial of degree at most two, but that polynomial may be linear. Points (0,1),(1,3),(2,5) give y=2x+1, corresponding to a=0 in ax²+bx+c.
If the task insists on a genuine quadratic with a≠0, no such quadratic passes through those three points. Altering a to a small nonzero decimal would fail at least one given coordinate. Check the leading coefficient before claiming success.
A repeated x-value can contradict the function rule
A single-valued function cannot pass through both (1,2) and (1,5), because one input cannot produce two distinct outputs. This is true for linear, quadratic and more complicated functions.
The issue should be recognised before solving a coefficient system. It differs from insufficient information: there is no possible function satisfying these two given conditions as stated.
Negative scale reflects rather than moves roots
With roots −2 and four, y=a(x+2)(x−4). If a=−1, then y=−x²+2x+8, axis x=1 and maximum nine at that input. If a=1, the roots are the same but the graph opens upward.
A pupil who believes negative a creates negative roots is confusing vertical reflection with a change in factors. Root coordinates come from the factors; opening direction comes from the multiplier.
Maximum value versus specified vertex
If a quadratic has maximum nine at x=1 and passes through (0,8), use vertex (1,9): y=a(x−1)²+9. The point yields a=−1, giving y=−x²+2x+8.
Knowing only that the maximum output is nine is weaker; both y=−x²+9 and y=−2(x−3)²+9 have that maximum at different positions. The x-location of an extremum is independent information.
A tangent gradient belongs in the derivative
Let f=ax²+bx+c, with f(0)=1, f(1)=4 and tangent gradient at x=1 equal to four. The point conditions give c=1 and a+b=3. Since f'(x)=2ax+b, the slope condition is 2a+b=4.
Thus a=1,b=2 and f=x²+2x+1. If a student writes f(1)=4 twice, they have failed to use the gradient condition. The derivative measures slope, not curve height.
Discriminant information can leave a parameter interval
For y=x²+kx+9 to have no real x-intercepts, k²−36<0, giving −6<k<6. Tangency to the x-axis corresponds to k²−36=0, so k=−6 or six.
These conditions do not necessarily give one unique curve. An inequality restricts a family of parameters rather than fixing a single value. State the full allowed set instead of guessing a representative k.
One axis plus two points can determine the whole curve
Suppose the axis is x=2 and the points are (0,3),(1,0). Begin with y=a(x−2)²+k. Substitution gives 3=4a+k and 0=a+k. Subtraction yields a=1 and k=−1.
Thus y=(x−2)²−1, whose roots one and three have average two. Vertex form used the known axis and required just two equations for a and k, saving unnecessary simultaneous equations for a,b,c.
Models require permitted input values
A fictional rectangle of perimeter twenty units has sides x and 10−x and area A=x(10−x)=−x²+10x. Positive side lengths restrict 0<x<10, though the polynomial itself is defined for every real x.
Its maximum twenty-five at x=5 follows from the quadratic vertex and belongs to the feasible interval. This is original mathematics, not a measured Punggol property or actual garden design.
Exact interpolation is not proof of a physical law
A polynomial through three exact constructed points satisfies those points by design. Real-world observations may contain measurement error; fitting a quadratic to a few observations does not prove the process follows that law outside the observed range.
When using fictional data, label them as classroom coordinates. Real Punggol measurements would require authoritative sources, units, uncertainties and a validation argument before predictive claims.
Choose an independent verification route
A curve found from roots can be checked in expanded or vertex form. A curve found from a vertex can be checked by direct point substitution and expansion. A three-point solution should be tested at all three coordinates and checked for a nonzero a.
Repeating the same mistaken arithmetic twice is a weak check; another representation can reveal the error. Ask the learner why two forms agree rather than merely copying both.
A six-week teaching and delayed-retrieval plan
Week one checks signed substitution, factorisation and coordinate reading. Week two reconstructs from roots and a scale point. Week three uses vertices and completed squares. Week four solves three-point systems. Week five adds gradients, symmetry and parameter constraints. Week six mixes unlabelled questions and retests earlier errors.
The Clementi standard combines clear explanation, the earliest wrong-line diagnosis, controlled small-group discussion and an independent changed problem. A particular learner may need a slower prerequisite repair; six weeks is a review structure rather than a grade guarantee.
Original quadratic method and result checks
These are original teaching checks rather than national examination reproductions. A correct answer must satisfy every original point, zero, gradient or domain restriction; a method should be selected from the given information.
1. Two roots
Problem: Roots 1 and4. Exact result: a(x−1)(x−4). Check by direct substitution into the given conditions; use a changed example for independent transfer.
2. Find a
Problem: Roots1,4 and y(0)=8. Exact result: a=2. Check by direct substitution into the given conditions; use a changed example for independent transfer.
3. Expand
Problem: 2(x−1)(x−4). Exact result: 2x²−10x+8. Check by direct substitution into the given conditions; use a changed example for independent transfer.
4. Axis
Problem: Roots1,4. Exact result: x=5/2. Check by direct substitution into the given conditions; use a changed example for independent transfer.
5. Minimum
Problem: Same function. Exact result: (5/2,−9/2). Check by direct substitution into the given conditions; use a changed example for independent transfer.
6. Vertex form
Problem: Vertex(2,−3). Exact result: a(x−2)²−3. Check by direct substitution into the given conditions; use a changed example for independent transfer.
7. Vertex scale
Problem: Vertex(2,−3) and (0,5). Exact result: a=2. Check by direct substitution into the given conditions; use a changed example for independent transfer.
8. Expand vertex
Problem: 2(x−2)²−3. Exact result: 2x²−8x+5. Check by direct substitution into the given conditions; use a changed example for independent transfer.
9. Double root
Problem: Repeated root three. Exact result: a(x−3)². Check by direct substitution into the given conditions; use a changed example for independent transfer.
10. Double root scale
Problem: y(0)=18, root3 double. Exact result: a=2. Check by direct substitution into the given conditions; use a changed example for independent transfer.
11. Three points
Problem: (0,1),(1,2),(2,7). Exact result: 2x²−x+1. Check by direct substitution into the given conditions; use a changed example for independent transfer.
12. No real roots
Problem: 2x²−x+1. Exact result: D=−7. Check by direct substitution into the given conditions; use a changed example for independent transfer.
13. Negative input points
Problem: (−1,6),(0,1),(2,3). Exact result: 2x²−3x+1. Check by direct substitution into the given conditions; use a changed example for independent transfer.
14. Factor check
Problem: 2x²−3x+1. Exact result: (2x−1)(x−1). Check by direct substitution into the given conditions; use a changed example for independent transfer.
15. Axis given
Problem: Axis2,root5. Exact result: Other root −1. Check by direct substitution into the given conditions; use a changed example for independent transfer.
16. Axis and point
Problem: Roots−1,5, y(0)=−10. Exact result: a=2. Check by direct substitution into the given conditions; use a changed example for independent transfer.
17. Contradiction
Problem: Root1 but (1,3). Exact result: Impossible. Check by direct substitution into the given conditions; use a changed example for independent transfer.
18. Redundancy
Problem: Roots1,4 and (1,0). Exact result: a not fixed. Check by direct substitution into the given conditions; use a changed example for independent transfer.
19. Line example
Problem: (0,1),(1,3),(2,5). Exact result: a=0. Check by direct substitution into the given conditions; use a changed example for independent transfer.
20. Repeated input
Problem: (1,2),(1,5). Exact result: No function. Check by direct substitution into the given conditions; use a changed example for independent transfer.
21. Reflected roots
Problem: −2,4 with a=−1. Exact result: −x²+2x+8. Check by direct substitution into the given conditions; use a changed example for independent transfer.
22. Reflected vertex
Problem: Same function. Exact result: (1,9) max. Check by direct substitution into the given conditions; use a changed example for independent transfer.
23. Maximum and point
Problem: Max9 at x1 and (0,8). Exact result: −(x−1)²+9. Check by direct substitution into the given conditions; use a changed example for independent transfer.
24. Maximum alone
Problem: Max output nine only. Exact result: Not unique. Check by direct substitution into the given conditions; use a changed example for independent transfer.
25. Gradient equation
Problem: f’=2ax+b, slope4 at1. Exact result: 2a+b=4. Check by direct substitution into the given conditions; use a changed example for independent transfer.
26. Gradient full
Problem: f0=1,f1=4,f’1=4. Exact result: f=x²+2x+1. Check by direct substitution into the given conditions; use a changed example for independent transfer.
27. No intercept
Problem: x²+kx+9. Exact result: −6<k<6. Check by direct substitution into the given conditions; use a changed example for independent transfer.
28. Tangency parameter
Problem: x²+kx+9. Exact result: k=±6. Check by direct substitution into the given conditions; use a changed example for independent transfer.
29. Axis and points
Problem: Axis2,(0,3),(1,0). Exact result: (x−2)²−1. Check by direct substitution into the given conditions; use a changed example for independent transfer.
30. Rectangle model
Problem: Sides x,10−x. Exact result: A=−x²+10x. Check by direct substitution into the given conditions; use a changed example for independent transfer.
31. Rectangle optimum
Problem: Same. Exact result: 25 at x=5. Check by direct substitution into the given conditions; use a changed example for independent transfer.
32. Root/intercept
Problem: Roots2,−3,y0=−12. Exact result: a=2. Check by direct substitution into the given conditions; use a changed example for independent transfer.
33. Root0/4
Problem: Roots0,4 and y2=8. Exact result: a=−2. Check by direct substitution into the given conditions; use a changed example for independent transfer.
34. Axis −3/7
Problem: Roots−3 and7. Exact result: x=2. Check by direct substitution into the given conditions; use a changed example for independent transfer.
35. General axis
Problem: Roots r,s. Exact result: (r+s)/2. Check by direct substitution into the given conditions; use a changed example for independent transfer.
36. General vertex
Problem: a(x−h)²+k. Exact result: (h,k). Check by direct substitution into the given conditions; use a changed example for independent transfer.
37. Intercept
Problem: ax²+bx+c at x0. Exact result: c. Check by direct substitution into the given conditions; use a changed example for independent transfer.
38. Derivative
Problem: ax²+bx+c. Exact result: 2ax+b. Check by direct substitution into the given conditions; use a changed example for independent transfer.
39. Direction
Problem: a<0. Exact result: Opens down. Check by direct substitution into the given conditions; use a changed example for independent transfer.
40. Quadratic validity
Problem: a=0. Exact result: Not genuine quadratic. Check by direct substitution into the given conditions; use a changed example for independent transfer.
41. Evaluate
Problem: 2x²−x+1 at2. Exact result: 7. Check by direct substitution into the given conditions; use a changed example for independent transfer.
42. Root sign
Problem: Root −4. Exact result: Factor x+4. Check by direct substitution into the given conditions; use a changed example for independent transfer.
43. Origin
Problem: f0=0. Exact result: c=0. Check by direct substitution into the given conditions; use a changed example for independent transfer.
44. Roots0/5
Problem: f0=f5=0. Exact result: ax(x−5). Check by direct substitution into the given conditions; use a changed example for independent transfer.
45. Same roots axis
Problem: 0,5. Exact result: x=5/2. Check by direct substitution into the given conditions; use a changed example for independent transfer.
46. Point redundant
Problem: Roots0,5 and (0,0). Exact result: a unknown. Check by direct substitution into the given conditions; use a changed example for independent transfer.
47. Line alternate
Problem: (0,2),(1,4),(2,6). Exact result: y=2x+2. Check by direct substitution into the given conditions; use a changed example for independent transfer.
48. Negative vertex
Problem: Vertex(−1,4),point(0,2). Exact result: −2(x+1)²+4. Check by direct substitution into the given conditions; use a changed example for independent transfer.
49. Minimum square
Problem: (x−3)²+2. Exact result: 2 at x=3. Check by direct substitution into the given conditions; use a changed example for independent transfer.
Eight student error diagnostics
Root sign is reversed
The student turns root four into factor x+4. That factor is zero at −4, not four. Test the proposed root in the factor and repair the correspondence before solving another problem.
Multiplier is assumed to be one
Two roots specify a family, not its exact vertical scale. An additional point must fix the coefficient. Comparing a=1 and a=2 with the same roots exposes the missing information.
Negative input loses brackets
At x=−1 the student writes ax²=−a. The squared input is positive one, so the contribution is +a. Teach signed-square meaning before another three-point system.
Three collinear points are forced into a quadratic
The three values lie on a line and give a=0. Inventing a tiny nonzero a fails one of the coordinates. The correct conclusion is no genuine quadratic fits the exact data.
Maximum coordinates are swapped
The maximum nine at x=1 is written as a(x−9)²+1. The correct vertex (1,9) produces a(x−1)²+9. Sketch or label the ordered pair first.
A root point is wrongly counted as new evidence
Point (1,0) adds nothing when root one is already known. Ask whether the extra condition actually removes any values of a. Use an off-root point to determine scale.
A tangent gradient is entered as a function value
A slope of four means f'(1)=4, not necessarily f(1)=4. Write the derivative separately and distinguish the rate from the graph’s vertical value.
A model uses impossible side lengths
The unrestricted parabola permits negative x, but the rectangular design does not. State positive side-length constraints before optimizing a model and check feasibility.
Punggol parent FAQ
Do both K232/K341 include quadratics?
Yes, under A1.
Why use factor form for roots?
Each root gives a linear factor.
What form uses a known vertex?
a(x−h)²+k.
Do three points give a unique quadratic?
Three distinct-x points give a unique degree-at-most-two polynomial, which may be linear.
What if a=0?
It is not a genuine quadratic.
What if a point repeats an existing root?
It may provide no new information.
Can a gradient fix coefficients?
Use the derivative equation.
Can a given data set be impossible?
Yes; conflicting values or root statements may be inconsistent.
Are roots the same as the axis?
No; axis is their average if two distinct real roots exist.
Does a quadratic model apply to every real x?
Only if its context permits every real input.
How should it be checked?
Test every supplied condition and an independent form.
What is the parent progress test?
A changed question solved without method prompts.
Official source documents and related guides
Checked against official 2027 SEAB K232 and K341 A1 content. All numerical examples, models and diagnostic cases are original educational constructions; they do not represent measured local sites or guaranteed grades.
- Official SEAB 2027 K232 G2 Additional Mathematics
- Official SEAB 2027 K341 G3 Additional Mathematics
- Immutable Clementi first-principles small-group reference
- Punggol quadratic graph overview
- Punggol quadratic inequalities and discriminants
- Punggol simultaneous linear/nonlinear equations
- Punggol quadratic optimisation
- Punggol A-Math education hub
Extended reconstructions and independent parent checks
Extra worked case: two roots with a negative multiplier
A parabola has zeros −2 and four, and contains the point (0,8). The factor form is y=a(x+2)(x−4). At x=0 the product is −8a, so 8=−8a and a=−1. Thus y=−(x+2)(x−4)=−x²+2x+8.
The axis is x=1 and the vertex is (1,9), a maximum. All three original conditions are satisfied. This shows that a negative leading coefficient can coexist with both a negative and positive root; the sign of a should be determined from data, not guessed.
Extra worked case: one vertex and a root
The vertex is (3,−8) and a root is x=1. Write y=a(x−3)²−8. At the root, 0=4a−8, so a=2. Hence y=2(x−3)²−8=2(x−1)(x−5).
The second root at five follows from symmetry around x=3. This is a three-way check: vertex, stated root and inferred partner. A student could use factorised form with roots one and five, but vertex form is the shortest starting choice.
Extra worked case: a quadratic with no real roots
Suppose f has vertex (−2,3), opens upward with f(0)=11, and the task asks for its equation. Let f=a(x+2)²+3. Substitution gives 11=4a+3, so a=2 and f=2(x+2)²+3.
Because a is positive and the minimum is three, no real x can give f(x)=0. The graph has no real x-intercepts. A pupil who draws two crossings merely because a quadratic usually looks U-shaped is ignoring the specified minimum.
Extra worked case: reconstruct from two points and a derivative
Let f=ax²+bx+c with f(0)=2, f(2)=6 and gradient f'(2)=5. The first condition gives c=2. The second gives 4a+2b=4 or 2a+b=2. Differentiation gives 4a+b=5.
Subtract the two equations to obtain 2a=3, so a=3/2 and b=−1. The curve is (3/2)x²−x+2. At x=2 it equals six and its derivative 3x−1 equals five. Both height and slope must be verified.
Extra worked case: the roots and the leading coefficient
A quadratic has leading coefficient three and roots −1 and two. Since a is already known, no extra point is required: y=3(x+1)(x−2)=3x²−3x−6. Its y-intercept is −6.
The axis is x=1/2. The value there is 3(3/2)(−3/2)=−27/4. This case contrasts with the family where the leading coefficient was not given; a different piece of information can supply the same missing scale.
Extra worked case: an unknown constant from tangency
For y=x²−4x+k, suppose the graph touches the x-axis. Its discriminant is (−4)²−4k=16−4k. Tangency requires zero discriminant, so k=4 and the curve becomes (x−2)².
This is a unique parameter because the equation has a fixed coefficient on x². When the problem says only ‘no real roots’, it instead requires k>4. Distinguish an equality that fixes a value from an inequality that gives a range.
Extra worked case: the symmetry axis and y-intercept
Suppose a quadratic has axis x=−1, passes through (0,3) and through (1,11). Use y=a(x+1)²+k. At zero, a+k=3; at one, 4a+k=11. Subtracting yields 3a=8, hence a=8/3 and k=1/3.
The resulting curve is y=(8/3)(x+1)²+1/3. It opens upward and has no real roots since its minimum is positive. Fractions are legitimate exact coefficients and should not be rounded away simply because the numbers look inconvenient.
Extra worked case: a quadratic through the origin
If f(0)=0, the constant c in f=ax²+bx+c is zero. Suppose the graph also contains (1,2) and (2,6). Then a+b=2 and 4a+2b=6, equivalent to 2a+b=3.
Subtracting gives a=1,b=1 and f=x²+x=x(x+1). Its roots are zero and −1. The origin point removes one unknown; starting with c=0 is easier than treating it as a general unknown in all equations.
Extra worked case: repeated x-value as a consistency test
The stated data (−1,2), (−1,2), (2,5) contain one repeated identical point. They are not inconsistent, but there are only two distinct input-output conditions. Infinitely many quadratics can pass through them.
This is different from a repeated x-value paired with two different y-values, which contradicts being a function. Teach pupils to distinguish repeated information from incompatible information rather than simply counting printed coordinate pairs.
Extra worked case: a constant second difference
For equally spaced x-values zero, one, two, three, consider y-values two, five, twelve, twenty-three. The first differences are three, seven and eleven; the second differences are four and four. For a quadratic on unit-spaced inputs, the second difference equals 2a.
Thus a=2. Since f(0)=2, c=2; and f(1)=5 gives 2+b+2=5, so b=1. The curve f(x)=2x²+x+2 gives f(3)=18+3+2=23. This numerical pattern is a useful check, not a substitute for confirming all coordinates.
A disciplined one-page family progress record
Use six fields: actual SEC level, independently attempted question, first wrong line, mathematical reason for the correction, independently completed variation and delayed review date. For example, a pupil may misread roots −2 and four, choosing the wrong factor, then later explain why x+2 vanishes at −2.
The record should not invent an official school assessment or compare identifiable classmates. It should show what the learner could do without hints and what remained unstable after several days. This is the Clementi benchmark of diagnosing a specific mental move and checking retained transfer.
Further exact reconstruction and information-sufficiency tests
Choosing forms when the roots are irrational
Suppose a curve has roots 2−√3 and 2+√3, and passes through (0,1). Its factors combine as (x−2+√3)(x−2−√3)=(x−2)²−3. Thus y=a[(x−2)²−3]. At zero this bracket equals one, so a=1.
The result y=(x−2)²−3 has vertex (2,−3). The irrational roots and completed-square form are exact. Premature decimal rounding of √3 would introduce unnecessary approximations and make the symmetry harder to recognise.
A y-intercept alone cannot determine three coefficients
Knowing only f(0)=5 gives c=5 in ax²+bx+c but leaves a and b free, with a≠0 required for a genuine quadratic. Both f(x)=x²+5 and f(x)=2x²−3x+5 satisfy the same intercept.
When a child assigns missing coefficients zero without being told to, the issue is information sufficiency. Ask which further independent conditions could select a unique curve—two other point values, an axis and a point, or two suitable roots.
The gradient at the vertex must be zero
For f(x)=a(x−h)²+k, differentiation gives f'(x)=2a(x−h); thus f'(h)=0. If a purported vertex at x=2 is paired with a nonzero tangent gradient at x=2, the data conflict for any differentiable quadratic.
This is a strong consistency test that does not require solving for a. A student who has learned differentiation can independently detect an impossible combination of conditions before doing a long coefficient calculation.
A negative maximum does not imply a negative multiplier is wrong
Suppose a parabola has maximum −1 at x=3 and passes through (2,−4). Write f=a(x−3)²−1 and substitute −4=a−1, giving a=−3. Hence f=−3(x−3)²−1.
The maximum output is negative one, and every value is at most that. A negative a is correct because the parabola opens downward. Confusing the sign of the maximum value with the direction of opening is an error in interpreting extrema.
The coefficient from a fixed second derivative
If a quadratic f has f”(x)=6 everywhere and passes through (0,1) and (1,5), the general form is f=ax²+bx+c with 2a=6, so a=3. The points give c=1 and 3+b+1=5, so b=1.
The curve is f=3x²+x+1. Its second derivative is six and both point conditions hold. This method uses an additional calculus condition rather than a root or a vertex.
Why one condition can be dependent on the others
Suppose roots one and four are given, and the axis x=5/2 is also stated. The axis agrees with the roots but contributes no independent information about a. The family y=a(x−1)(x−4), a≠0, remains.
Another point such as (0,8) would determine a. The pupil should distinguish a useful consistency check from a constraint that reduces unknowns; both are valuable but have different roles.
A corrected solution should survive an unfamiliar representation
A child may solve a factor-form problem after watching the tutor, yet fail when the next task gives a vertex and a point. Ask the student to translate both into the same expanded quadratic and explain which fact each parameter represents.
This is the difference between recognising a template and transferring a mathematical concept. A delayed exercise should change the form, coefficient signs and point locations, while remaining within the student’s actual SEC syllabus and prerequisite readiness.
