Skip to content

Tool 3 — Mathematics

The mathematical toolkit the course assumes — arithmetic, algebra, units, uncertainties and graphing.

Editorial framing The content below is transcribed from the IB Chemistry guide (first assessment 2025), section “Skills in the study of chemistry”. These skills are cross-cutting: they are integrated into the teaching of the syllabus rather than taught as stand-alone topics.

Applying general mathematics

  • Arithmetic and algebra
  • Decimals, fractions, percentages, ratios, reciprocals, exponents
  • Logarithmic functions
  • Exponential functionsHL
  • Rates of change from tabulated data
  • Mean and range
  • Scientific notation
  • Approximation and estimation; knowing when effects can be ignored
  • Order-of-magnitude comparison
  • Direct and inverse proportionality and positive/negative correlation
  • Percentage change and percentage difference
  • Percentage error and percentage uncertainty
  • Continuous vs discrete variables

Using units, symbols and numerical values

  • SI prefixes and units
  • Symbols from the guide and data booklet
  • Appropriate significant figures / decimal places

Processing uncertainties

  • Significance of uncertainties in raw and processed data
  • Record uncertainties as a range (±) to appropriate precision
  • Propagate uncertainties through addition, subtraction, multiplication and division
  • Propagate uncertainties through exponentsHL
  • Propagate uncertainties through exponential functionsHL
  • Express absolute, fractional (relative) and percentage uncertainties appropriately
  • Apply the coefficient of determination R² to evaluate fit

Graphing

  • Sketch graphs with labelled unscaled axes to describe trends qualitatively
  • Construct and interpret tables, bar charts, histograms, scatter graphs, line and curve graphs
  • Plot linear and non-linear graphs with appropriate scales
  • Lines and curves of best fit
  • Interpret gradient, changes in gradient, intercepts, maxima, minima and areas
  • Draw and interpret uncertainty bars
  • Extrapolate and interpolate

What is HL-only in the mathematics

Exponentials HL

Exponential functions are HL-only. The Arrhenius equation (k = Ae^(−Ea/RT)) and [H⁺] = 10⁻ᵖᴴ both require interpreting a negative exponent.

Logarithms SL

Logarithmic functions are SL — they appear in pH, pOH, pKa and pKb calculations throughout the course, not only at HL.

Common misconceptions & bridge items for Tool 3

Common misconceptions (6)

  • M-78Repeating a measurement improves accuracy.

    Confidence: verified

    Why it’s wrong: Repetition reduces random error and improves precision.

    Correction: It does nothing about systematic error, which is what limits accuracy.

  • M-79'Human error' is a valid source of error.

    Confidence: verified

    Why it’s wrong: It is a non-answer.

    Correction: Name the specific measurement, its uncertainty, and whether the effect is random or systematic, and state the direction of the effect on the result.

  • M-80Write down all the digits the calculator gives.

    Confidence: verified

    Why it’s wrong: The number of significant figures is limited by the least precise measurement.

    Correction: For multiplication/division, match the input with the fewest sig figs.

  • M-81A single burette reading has uncertainty ±0.05 cm³, so the titre does too.

    Confidence: verified

    Why it’s wrong: A titre is a difference of two readings, so the absolute uncertainties add.

    Correction: The titre has uncertainty ±0.10 cm³.

  • M-82Percentage error and percentage uncertainty are the same.

    Confidence: verified

    Why it’s wrong: Percentage uncertainty comes from instrument precision (random); percentage error compares your result to an accepted literature value and reveals systematic error.

    Correction: Compare the two: if percentage error ≫ total percentage uncertainty, there is a systematic problem.

  • M-83For a graph, just join the dots.

    Confidence: verified

    Why it’s wrong: A best-fit line or curve is required.

    Correction: Draw a best-fit line/curve, and where the relationship is expected to be linear, use gradient and intercept quantitatively (with the linearised form of the equation where needed, e.g. ln k vs 1/T).

Personal data loaded.