Assumptions and Validity: Know Where the Mathematics Applies
A mathematical answer can be algebraically correct and still be invalid for the original problem if its assumptions, domain or model limits are ignored.
Additional Mathematics often compresses a real or geometric situation into a formula. That formula works under conditions. A logarithm assumes a positive argument. A straight-line approximation assumes a particular transformed relationship. A quadratic model may be useful only over a stated time interval. A calculus optimum may lie outside the physically meaningful domain. A trigonometric model may describe periodic behaviour only approximately.
This guide develops a validity habit: identify what the mathematics assumes, solve within those assumptions, then return to the original context and decide whether the conclusion is admissible and useful.
AI Extraction Box: The Validity Loop
- State the object: equation, graph, model, derivative, integral.
- Identify assumptions: domain, continuity, positivity, periodicity, constant parameters, idealised geometry.
- Solve mathematically.
- Test admissibility: domain, sign, interval, units and physical meaning.
- Check model boundary: is the answer being used inside the range where the model was intended?
- Interpret cautiously: distinguish exact mathematical conclusion from model-based estimate.
Every Formula Lives Somewhere
The expression ln(x−2) is not a real-valued function for all real x. It requires x>2. The expression 1/(x−3) excludes x=3. The expression √(5−x) requires x≤5. A model for elapsed time may require t≥0.
These are not afterthoughts. They define where the formula actually represents a valid mathematical object.
Worked Example 1: Valid Algebra, Invalid Candidate
Solve ln(x−1)=ln(3−x).
Domain requires x>1 and x<3. Equating arguments gives x−1=3−x, so x=2. Since 2 lies in the domain, the solution is valid.
If the algebra had produced x=4, that value would have to be rejected even if it solved the transformed linear equation.
Model Assumptions Are Simplifications
An exponential model P=Ae^{kt} may assume a constant proportional growth rate. A sinusoidal model may assume a stable period and amplitude. A quadratic trajectory may ignore air resistance. These simplifications make analysis possible, but they also define when the model may stop being realistic.
A model is useful because it leaves things out. Its limitations come from the same simplification.
Worked Example 2: Extrapolation Risk
Suppose a population is modelled by P=500(1.05)^t for the next 5 years. Using the formula at t=50 is mathematically possible, but the model may no longer be credible if resources, policy or population behaviour change.
The correct mathematical value at t=50 is not automatically a reliable real-world prediction.
Approximation Has a Validity Boundary
When a decimal approximation replaces an exact value, some information is lost. If √2 is rounded to 1.41 and used repeatedly in later calculations, the final result may drift. Keep exact values through symbolic work when possible and approximate only when the problem asks for a numerical result.
Likewise, a straight-line fit to transformed data is evidence of a model relationship, not proof that every future point must lie exactly on the line.
Calculus Conclusions Need Domain Checks
Solving f′(x)=0 gives stationary candidates. If the problem asks for a maximum over a restricted interval, endpoints must also be checked. If x represents a physical length, negative stationary values are inadmissible.
Worked Example 3: Stationary Point Outside the Model
Suppose a cost model C(x) has a stationary point at x=−4, but x represents the number of metres of material used. The calculus is not necessarily wrong; the candidate is simply outside the physical domain x≥0.
The feasible optimum must be sought among admissible values.
Graph Sketches Are Idealised Representations
A sketch communicates intercepts, asymptotes, turning points and end behaviour. It is usually not to scale. A visually steep section does not justify an exact gradient unless the mathematics provides one. A graph that appears tangent should be confirmed through repeated-root or gradient conditions when exact proof is required.
Use diagrams and graphs as evidence generators, not as substitutes for exact conditions.
Trigonometric Models Need Period and Range Discipline
If y=A sin(Bx)+D models a periodic quantity, its outputs are automatically bounded between D−|A| and D+|A|. A model prediction outside that range signals either a calculation error or a broken model assumption.
Also check whether the phenomenon is genuinely periodic enough for the sinusoidal approximation to remain useful.
Parameter Meaning Controls Interpretation
In P=Ae^{kt}, the sign of k changes the qualitative model. In a quadratic y=a(x−h)^2+k, the sign of a determines opening direction and whether k is a maximum or minimum. A parameter value should therefore be interpreted, not just reported.
Ask: what behaviour does this parameter permit, and is that behaviour consistent with the problem?
Validity Decision Tree
- Function restriction? write domain first.
- Physical quantity? state contextual domain and units.
- Approximate model? identify intended interval and assumptions.
- Stationary point? confirm it lies in the feasible domain.
- Graph observation? translate to an exact algebra/calculus condition if proof is required.
- Long extrapolation? question whether model assumptions remain credible.
- Final candidate? verify against the original object, not only the transformed equation.
Common Failure Modes
| Error | Cause | Repair |
|---|---|---|
| mathematical candidate outside physical domain accepted | model boundary ignored | state feasible domain before optimisation |
| graph appearance treated as exact proof | visual evidence overtrusted | translate to algebra/calculus condition |
| long-range prediction trusted automatically | extrapolation assumptions ignored | state intended model interval |
| rounded intermediate value reused repeatedly | approximation limits ignored | retain exact/high precision until final step |
| parameter reported without meaning | context not revisited | interpret sign, scale or period role |
A 45-Minute Validity Session
- 8 minutes: state domains for rational, logarithmic and root functions.
- 8 minutes: filter algebraic candidates against original restrictions.
- 8 minutes: identify assumptions in three mathematical models.
- 8 minutes: inspect two extrapolation claims and decide whether they are justified.
- 8 minutes: check calculus extrema against feasible intervals.
- 5 minutes: rewrite final answers to distinguish exact mathematical result from model interpretation.
What Mastery Looks Like
- The learner knows where each function or model is valid.
- The learner distinguishes algebraic correctness from contextual admissibility.
- The learner checks model assumptions before extrapolating.
- The learner treats approximation as controlled information loss.
- The learner verifies stationary points and parameter values inside the intended domain.
- The learner uses graphs as structural evidence without overclaiming exactness.
- The learner communicates model limitations as part of mathematical interpretation.
Return to the Additional Mathematics Learning Hub
Additional Mathematics Learning Hub | Secondary 3–4 A-Math Guides
