Evaluation

Q: What is the difference between accuracy, precision, and validity in an experimental evaluation?

Accuracy describes how close a measurement is to the true or accepted value; precision describes the closeness of repeated measurements to each other; validity describes whether the experiment actually measures what it intended to by controlling extraneous variables.

Q: What are random errors and how can their effect be minimised?

Random errors cause measurements to fluctuate unpredictably above and below the true value, and their effect is minimised by taking repeated measurements and calculating a mean.

Q: What are systematic errors and how can they be identified and corrected?

Systematic errors shift all readings consistently in one direction, identifiable on a graph as a shifted intercept, and are corrected by recalibrating instruments, subtracting zero errors, or modifying the method.

Q: How do you mathematically determine if an experimental result is consistent with an accepted theoretical value?

Calculate the percentage difference between the experimental and accepted values; if \(\text{Percentage Difference} \le \text{Total Percentage Uncertainty}\), the result is consistent within experimental error.

Q: How should anomalous data points be handled in data tables and on graphs?

Anomalies must be excluded when calculating mean values and should be circled or ignored when drawing a line or curve of best fit.

Q: What improvements should be suggested for timing experiments involving short intervals or oscillations?

Use light gates with a data logger or electromagnetic release mechanisms for short drops, and time multiple oscillations (e.g., 10 to 20) with a fiducial marker placed at the equilibrium position.

Q: What techniques improve the measurement of small lengths or wire diameters?

Use a micrometer screw gauge or digital vernier caliper, and measure at multiple positions and perpendicular orientations to calculate an average.

Q: How can heating effects be minimised in electrical circuit experiments?

Open the circuit switch between readings to allow components to cool, and use low currents by adding a current-limiting resistor or rheostat.

Physics 1210 · CCEA A Level · 8 min read · Updated: Aug 31, 2026

Welcome to Evaluation in AS Physics!

Hello and welcome to one of the most rewarding parts of your AS Physics practical journey! In your AS 3: Practical Techniques and Data Analysis unit, carrying out experiments and crunching numbers is only half the battle. A true physicist must be able to step back, look at their data critically, and ask: "How trustworthy are my results, and how could this experiment be improved?"

In this guide, we will break down the evaluation process step-by-step. Don't worry if experimental critique has felt daunting before—by following a few clear rules and logical steps, you will be able to score top marks on every evaluation question in your exam.

1. The Core Purpose of an Evaluation

Evaluating an experiment does not mean saying you did everything wrong! It means objectively assessing the quality of your data, identifying the sources of uncertainty, and suggesting realistic, scientifically sound ways to make the investigation better.

Key Terms to Know:

• Accuracy: How close your measured value or experimental result is to the true or accepted standard value.
• Precision: How close repeated measurements are to one another (the spread or repeatability of the data).
• Validity: Whether the experiment actually measures what it set out to measure by controlling all extraneous variables.

Analogy Time: Imagine throwing darts at a dartboard. If all your darts land tightly clustered in the top-left corner far from the bullseye, your throws are precise but inaccurate. If your darts are scattered evenly all around the bullseye, they are accurate on average but imprecise. The goal in physics is to be both accurate and precise: hitting the bullseye consistently!

Key Takeaway: Evaluation is about judging accuracy, precision, and reliability using numbers and physical reasoning, not guesswork.

2. Pinpointing Errors: Random vs. Systematic

To evaluate an experiment properly, you must identify what type of error is affecting your results. Every measurement in the lab has limitations.

A. Random Errors

Random errors cause readings to fluctuate unpredictably above and below the true value. They affect the precision of your results.

• Causes: Human reaction time when using a manual stopwatch, background temperature fluctuations, or small parallax errors when reading a scale from slightly different angles.
• How to spot on a graph: Data points are scattered randomly on both sides of the line of best fit.
• How to minimise: Take at least three repeat readings for each value of the independent variable and calculate a mean. Using instruments with higher resolution also reduces reading uncertainty.

B. Systematic Errors

Systematic errors shift all readings in one consistent direction (always too high or always too low) by the same amount or by a constant factor. They affect the accuracy of your results.

• Causes: A meter that is not calibrated properly, a zero error (e.g. a micrometer showing \(+0.02\text{ mm}\) when fully closed), or heat loss to the surroundings that was not accounted for in the theoretical model.
• How to spot on a graph: The line of best fit has the expected gradient, but the \(y\)-intercept is shifted away from where theoretical physics predicts it should pass (e.g. not passing through the origin in a direct proportionality relationship).
• How to minimise: Recalibrate instruments, subtract zero errors before taking data, or redesign the apparatus/technique to remove the underlying physical bias.

Did you know? A zero error is a classic type of systematic error. If a top-pan balance reads \(0.5\text{ g}\) when nothing is placed on it, every single measurement you take will be \(0.5\text{ g}\) too large!

Key Takeaway: Random errors cause scatter (fix with repeats and averages); systematic errors cause a fixed shift (fix by calibration, zero adjustments, or technique changes).

3. Comparing Results: Percentage Difference vs. Percentage Uncertainty

Examiners love testing whether your final calculated value agrees with an established theoretical or reference value (such as the acceleration due to gravity, \(g = 9.81\text{ m s}^{-2}\)). You must be able to compare your experimental result with the accepted value mathematically.

Step 1: Calculate the Percentage Uncertainty

Find the total percentage uncertainty in your final calculated result (obtained by combining the percentage uncertainties of all measured quantities):

\(\text{Percentage Uncertainty} = \left( \frac{\text{Absolute Uncertainty}}{\text{Calculated Value}} \right) \times 100\%\)

Step 2: Calculate the Percentage Difference

Find how far your experimental result is from the standard accepted value:

\(\text{Percentage Difference} = \frac{|\text{Experimental Value} - \text{Accepted Value}|}{\text{Accepted Value}} \times 100\%\)

Step 3: Make a Scientific Judgement

• If \(\text{Percentage Difference} \le \text{Total Percentage Uncertainty}\):
The difference between your result and the true value can be entirely explained by experimental uncertainties. Your experiment is valid and consistent with the accepted theory.

• If \(\text{Percentage Difference} > \text{Total Percentage Uncertainty}\):
Experimental uncertainties alone cannot explain the difference. There is a significant systematic error or an invalid theoretical assumption present in the method.

Worked Example:

A student determines the acceleration of free fall to be \(g_{\text{exp}} = 9.54\text{ m s}^{-2}\) with an estimated total percentage uncertainty of \(\pm 4.0\%\). The accepted value is \(g_{\text{acc}} = 9.81\text{ m s}^{-2}\). Is the student's result consistent with the accepted value?

Solution:
\(\text{Percentage Difference} = \frac{|9.54 - 9.81|}{9.81} \times 100\% = \frac{0.27}{9.81} \times 100\% = 2.75\%\)
Since \(2.75\% < 4.0\%\), the percentage difference is smaller than the experimental percentage uncertainty. Therefore, the result is consistent with the accepted value within experimental error.

Key Takeaway: Always compute both percentages and compare them directly before writing your conclusion.

4. Dealing with Anomalous Data

An anomaly (or outlier) is a data point that clearly does not follow the general trend shown by the rest of the data set.

• In a table: If three repeats for the time of a falling ball are \(0.42\text{ s}\), \(0.41\text{ s}\), and \(0.68\text{ s}\), the reading \(0.68\text{ s}\) is clearly anomalous.
• In calculations: Never include an obvious anomaly when calculating the mean. Discard it, and if possible, take another repeat reading.
• On a graph: An anomaly sits noticeably far away from the smooth line or curve of best fit. Do not force your line to pass through it; circle it or ignore it when positioning your ruler.

Common Mistake to Avoid: Blindly adding up all numbers in your calculator to find an average! Always scan your table first and cross out any obvious rogue values before computing the mean.

5. Suggesting Realistic Experimental Improvements

In the AS 3 exam, you will frequently be asked to suggest realistic improvements to a given procedure. Vague answers like "be more careful", "use a better ruler", or "do it in a dark room" will score zero marks.

Your suggestions must state a specific piece of apparatus or a specific procedural technique, along with the physical reason why it improves the measurement.

High-Scoring Improvements for Common AS Physics Experiments:

A. Experiments Involving Time (e.g. Freefall, Oscillations)

• Problem: Human reaction time introduces significant random uncertainty when using a manual stopwatch for short time intervals.
• Improvement: Use light gates connected to a digital timer / data logger, or release the object using an electromagnet.
• Oscillations Improvement: When timing a simple pendulum or mass-spring system, time \(10\) or \(20\) complete oscillations and divide by \(N\). Always place a fiducial marker at the centre of the oscillation (equilibrium position) where the object moves at maximum speed, making the transit sharp and easiest to judge.

B. Experiments Involving Length and Thickness (e.g. Wire Diameter, Stretching)

• Problem: A standard metre ruler has a resolution of only \(\pm 1\text{ mm}\), which gives a huge percentage uncertainty for thin objects like wires or glass slides.
• Improvement: Use a micrometer screw gauge (resolution \(\pm 0.01\text{ mm}\)) or a digital vernier caliper (resolution \(\pm 0.01\text{ mm}\)).
• Technique: Measure the diameter at several different positions along the wire and at right angles (perpendicular orientations) to account for any non-uniform cross-section, then calculate a mean.

C. Electrical Experiments (e.g. Resistivity, Internal Resistance)

• Problem: Current causes the wire or resistor to heat up, which changes its resistance and ruins the validity of Ohm's law / constant resistivity.
• Improvement: Open the switch between readings to allow the components to cool, or use small currents by placing a current-limiting resistor or rheostat in series.
• Contact Resistance: Use firmly screwed terminals or crocodile clips cleaned of oxide layers to ensure minimal contact resistance.

D. Optics and Light Experiments (e.g. Refractive Index)

• Problem: Thick rays from ray boxes make locating the centre of the beam imprecise on protractor scales.
• Improvement: Use a narrow single slit on the ray box, use a pin-marking method with pins spaced far apart (at least \(5\text{ cm}\)), or use a monochromatic laser source in a darkened room.

Key Takeaway: Always pair your improvement with a "why": [Specific Apparatus/Method] + [Reason/Reduction in Percentage Uncertainty].

6. Summary Evaluation Checklist

When reviewing any practical setup or answering an evaluation question in your AS 3 paper, run through this quick mental checklist:

1. Range: Did the experiment cover a wide enough range of values (at least 5 to 6 distinct values across the domain)?
2. Repeats: Were repeats taken at every value to identify anomalies and reduce random errors?
3. Resolution: Were the measuring tools appropriate for the size of the quantity being measured?
4. Controlled Variables: Were all other variables kept strictly constant (e.g. temperature in electrical circuits)?
5. Comparison: Is the percentage difference within the calculated percentage uncertainty bounds?

Quick check

Can you answer these now?

Open each question to check the key ideas from this chapter.

What is the difference between accuracy, precision, and validity in an experimental evaluation?

Accuracy describes how close a measurement is to the true or accepted value; precision describes the closeness of repeated measurements to each other; validity describes whether the experiment actually measures what it intended to by controlling extraneous variables.

What are random errors and how can their effect be minimised?

Random errors cause measurements to fluctuate unpredictably above and below the true value, and their effect is minimised by taking repeated measurements and calculating a mean.

What are systematic errors and how can they be identified and corrected?

Systematic errors shift all readings consistently in one direction, identifiable on a graph as a shifted intercept, and are corrected by recalibrating instruments, subtracting zero errors, or modifying the method.

How do you mathematically determine if an experimental result is consistent with an accepted theoretical value?

Calculate the percentage difference between the experimental and accepted values; if \(\text{Percentage Difference} \le \text{Total Percentage Uncertainty}\), the result is consistent within experimental error.

How should anomalous data points be handled in data tables and on graphs?

Anomalies must be excluded when calculating mean values and should be circled or ignored when drawing a line or curve of best fit.

What improvements should be suggested for timing experiments involving short intervals or oscillations?

Use light gates with a data logger or electromagnetic release mechanisms for short drops, and time multiple oscillations (e.g., 10 to 20) with a fiducial marker placed at the equilibrium position.

What techniques improve the measurement of small lengths or wire diameters?

Use a micrometer screw gauge or digital vernier caliper, and measure at multiple positions and perpendicular orientations to calculate an average.

How can heating effects be minimised in electrical circuit experiments?

Open the circuit switch between readings to allow components to cool, and use low currents by adding a current-limiting resistor or rheostat.

Ready to test yourself?

Turn these notes into exam-style practice. Get unlimited AI questions on this topic with instant marking and explanations.

Practice This Topic

More Physics 1210 chapters

* The content provided by thinka is generated by AI and may not always be accurate or up-to-date. Please use it as a supplementary resource and verify with official materials.

Put These Notes into Practice

Reading the notes is a great start. Now practice with unlimited AI-generated questions and get instant feedback. 100,000+ students are already improving their grades.

Start Practicing Now View Pricing

Done reading? Test yourself with AI practice questions

Practice This Topic Now