IB Physics: Measurements and Uncertainties

Understanding the nature of measurement is the bedrock of all scientific inquiry. In IB Physics, you are not just learning to record numbers; you are learning to quantify the certainty of your knowledge. Mastering measurements and uncertainties equips you to design robust experiments, analyze data with a critical eye, and articulate the limits of your conclusions—a skill set that transcends physics and underpins all empirical research.

The Language of Science: SI Units and Significant Figures

Every measurement is a comparison to a standard. The International System of Units (SI) provides this universal language, built upon seven base units like the meter (m) for length and the kilogram (kg) for mass. Derived units, such as the newton (N = kg·m·s${}^{-2}$), are combinations of these base units. Always state units with your numerical values; a number without a unit is meaningless in physics.

The precision of a measured value is communicated through significant figures. These are all the digits in a number that are known reliably, plus the first uncertain digit. For example, a length recorded as 3.40 cm has three significant figures, indicating the measurement is certain to the tenths place (the '0') and estimated in the hundredths. Rules for counting are straightforward: all non-zero digits are significant, zeros between non-zero digits are significant, and leading zeros are not significant. Trailing zeros are significant only if a decimal point is present. In calculations, your final answer should reflect the least precise input. When multiplying or dividing, the answer has the same number of significant figures as the factor with the fewest. When adding or subtracting, the answer should be rounded to the least precise decimal place.

Systematic and Random Errors: Knowing Your Enemy

Not all uncertainties are created equal. Distinguishing between error types is crucial for improving experimental methodology. Systematic errors cause readings to differ from the true value by a consistent amount each time. These are often due to faulty calibration (e.g., a scale that reads 1 gram when nothing is on it), incorrect technique, or environmental factors. They shift all data points in one direction and are not reduced by repeating measurements.

In contrast, random errors cause readings to be scattered unpredictably around the true value. They arise from unpredictable fluctuations—in the measurement tool, the environment, or the observer's precision. Repeating measurements and averaging the results helps to reduce the effect of random errors. You can identify their presence by looking at the spread or scatter of data points around a line of best fit.

Quantifying Uncertainty: Absolute and Percentage

To communicate uncertainty meaningfully, you must quantify it. The absolute uncertainty is the raw margin of error, often denoted by $\pm$ symbol. If you measure a current as $I=2.3\pm 0.1$ A, the absolute uncertainty is 0.1 A. It carries the same units as the measurement itself.

Often, the scale of the uncertainty relative to the measurement matters more. The percentage uncertainty expresses the absolute uncertainty as a percentage of the measured value. It is calculated as: $(\text{absolute uncertainty}/\text{measured value})\times 100\%$. For our current measurement, the percentage uncertainty is $(0.1/2.3)\times 100\%\approx 4.3\%$. This allows you to compare the precision of measurements of different magnitudes.

Error Propagation: Carrying Uncertainty Through Calculations

When you use measured values in formulas, their uncertainties propagate to the final result. The rules for error propagation depend on the operation.

• Addition/Subtraction: Add the absolute uncertainties. If $T=A+B$, then the absolute uncertainty in $T$ is $\Delta T=\Delta A+\Delta B$.
• Multiplication/Division: Add the percentage uncertainties. If $P=A\times B$ or $P=A/B$, then the percentage uncertainty in $P$ is the sum of the percentage uncertainties in $A$ and $B$.
• Powers: Multiply the percentage uncertainty by the power. If $V=r^3$, then the percentage uncertainty in $V$ is three times the percentage uncertainty in $r$.

Example: You calculate kinetic energy using $E_k=\frac{1}{2}m{v}^2$. You measure mass $m=2.00\pm 0.01$ kg (0.5% uncertainty) and velocity $v=3.0\pm 0.2$ m/s (6.7% uncertainty). The percentage uncertainty in $v^2$ is $2\times 6.7\%=13.4\%$. The total percentage uncertainty in $E_k$ is $0.5\%+13.4\%=13.9\%$. You then convert this back to an absolute uncertainty for your final answer.

Graphical Analysis and Error Bars

Graphs are powerful tools for identifying relationships and extracting quantities. Error bars visually represent the absolute uncertainty in your measurements on a graph. When plotting a line of best fit, it should pass through as many error bars as possible. The uncertainty in the gradient or y-intercept can be found by drawing the worst acceptable lines—the steepest and shallowest lines that still pass through the rectangles formed by all error bars. The uncertainty in the gradient is then half the difference between the gradients of these worst lines.

For example, if your line of best fit has a gradient of 4.2 m/s${}^2$, and your worst acceptable lines have gradients of 4.5 and 3.9, the uncertainty is $(4.5-3.9)/2=0.3$ m/s${}^2$. You would report the gradient as $4.2\pm 0.3$ m/s${}^2$.

Experimental Design: Controlling Variables

A well-designed experiment tests one specific relationship. This requires you to identify three types of variables: the independent variable (what you change), the dependent variable (what you measure), and controlled variables (what you keep constant to ensure a fair test). Your procedure must detail not just what to do, but how you will measure each variable and the steps taken to control others. For instance, in an experiment to find the relationship between the length of a pendulum and its period, the length is independent, the period is dependent, and you must control the mass of the bob and the amplitude of swing. A strong design also includes a plan for collecting sufficient, precise data and for repeating measurements to address random error.

Common Pitfalls

1. Confusing precision with accuracy: A measurement can be very precise (repeatable to many decimal places) but inaccurate if it is far from the true value due to a systematic error. Always consider both.
2. Incorrect significant figures in final answers: A common mistake is to use all digits from your calculator display. You must round your final answer to reflect the uncertainty of the least precise input value used in the calculation.
3. Mixing uncertainty types in propagation: The most frequent error in error propagation is adding absolute uncertainties during multiplication. Remember: add absolute for +/-; add percentage for $\times$ / $÷$.
4. Ignoring uncertainty in graphical analysis: Simply drawing a single line of best fit without considering error bars or worst acceptable lines misses a critical part of analysis. The uncertainty in a gradient is often a key result of an experiment.

Summary

• All measurements have uncertainty, which must be quantified using absolute and percentage uncertainties to communicate the reliability of your data.
• Systematic errors shift data consistently and are not reduced by repetition, while random errors cause scatter and can be mitigated by taking repeated measurements and averaging.
• The rules of error propagation dictate how to combine uncertainties through calculations: add absolute uncertainties for addition/subtraction, and add percentage uncertainties for multiplication/division and powers.
• Error bars on graphs represent measurement uncertainty, and the worst acceptable lines method is used to determine the uncertainty in gradients and intercepts.
• Sound experimental design requires clear identification and control of independent, dependent, and controlled variables to validly investigate a specific physical relationship.