VCE Physics practical investigation structure: The 2026 guide

Study guide16 min readUpdated 7 Oct 2026

Quick answer

The VCE Physics Unit 4 student-designed practical investigation is a poster-formatted report on an original experiment, with sections covering research question, hypothesis, methodology, data, analysis (with uncertainty propagation and linearisation), discussion (uncertainty sources, limitations, improvements, comparison to theory) and conclusion; Band 6 posters demonstrate sophisticated uncertainty handling (random vs systematic, gradient uncertainty from min/max lines), explicit linearisation of non-linear relationships, and a discussion that names specific uncertainty sources tied to specific experimental steps.

Jump to a section

What this guide is for
Poster structure
Research question
Hypothesis
Variables and method
Uncertainty
Linearisation
Gradient and uncertainty
Discussion
Conclusion
Marking criteria
Check your knowledge

What this guide is for

The VCE Physics Unit 4 student-designed scientific investigation (Area of Study 2) is one of the major SACs. Students design, conduct, and report on an original investigation. The poster format requires disciplined attention to all aspects of scientific method. This guide covers the structure, marking criteria, and the moves that secure Band 6.

Poster structure

The study design sets the poster layout (no more than 600 words, no set paper size). The section list below is a planning checklist that you then fit into that layout: the introduction holds the aim, hypothesis and background physics, the methodology section holds the variables, method and risk management, and the results, discussion and conclusion sections hold the rest. Your logbook, assessed with the poster, holds your planning, risk management, raw data and preliminary analysis.

Title. Specific, descriptive.
Research question. One sentence.
Hypothesis. Predicted relationship with theoretical justification.
Variables. Independent (range, measurement method, precision), dependent (measurement method, precision), controlled (with rationale).
Methodology. Labelled diagram of setup; step-by-step procedure; justification of design choices.
Risk assessment. Identified hazards and mitigation.
Data tables. Raw and processed, with units and uncertainties.
Graphs. Linearised where applicable; uncertainty bars; best-fit line.
Analysis. Gradient, intercept, derived quantities with uncertainty.
Discussion. Uncertainty sources, limitations, suggested improvements, comparison to theory.
Conclusion. Direct answer to research question.
References.

Research question

Form: "How does [IV] affect [DV] for [system] with [controlled variables] held constant?"

Good examples:

How does the length of a simple pendulum affect its period?
How does the slit separation affect the fringe spacing in a Young's double-slit setup?
How does the inclination of an inclined plane affect the acceleration of a cart down the plane?

Each has a measurable IV, a measurable DV, and a clear system.

Hypothesis

Should be specific. "Period will increase as length increases" is weak; "Period will be proportional to $\sqrt{L}$ (so a plot of $T$ against $\sqrt{L}$ will be linear with gradient $2\pi/\sqrt{g}$ )" is strong.

Include the theoretical basis. The marker should see that the hypothesis arises from theory, not just intuition.

Variables and method

For each variable:

Range (e.g., 5 to 45 degrees in 5-degree steps).
Measurement method (e.g., protractor with 1 degree precision).
Number of trials per IV value (typically 3-5).
Why each controlled variable matters.

The methodology should include a clear labelled diagram. Step-by-step procedure must be detailed enough to replicate.

Uncertainty

Random uncertainty: From measurement-to-measurement variability. Estimate by half-range or standard deviation of repeated measurements.
Systematic uncertainty: From instrument bias (zero error, calibration). Estimate from instrument specifications.
Instrumental uncertainty: Half the instrument precision (typically).
Propagation rules

Addition/subtraction: add absolute uncertainties.
Multiplication/division: add fractional uncertainties.
Powers: multiply fractional uncertainty by the power.

Reporting. Value ± uncertainty, both to consistent decimal places.

Linearisation

Many physics relationships are non-linear. Linearise before plotting:

$y = k x^2$ : plot $y$ vs $x^2$ .
$y = k / x$ : plot $y$ vs $1/x$ .
$T = 2\pi \sqrt{L/g}$ : rearrange to $T^2 = (4\pi^2/g) L$ . Plot $T^2$ vs $L$ , gradient $4\pi^2/g$ .

Linearisation lets you fit a straight line and extract a meaningful gradient.

Figure 1. A log-log plot linearises a power law; the gradient gives the exponent and the intercept gives the prefactor k. Six computed points sit exactly on the best-fit line for y = 2.0 x^1.5.

Gradient and uncertainty

Best-fit line: standard linear regression through the data.

Gradient uncertainty: draw a max-slope line (steepest line through error bars) and a min-slope line (shallowest). The half-range of these slopes is the uncertainty in the gradient.

Figure 2. Each point carries a vertical and a horizontal uncertainty bar. The accent line is the best fit; the two dashed lines are the shallowest and steepest slopes consistent with all error bars, bracketing the gradient uncertainty.

Discussion

Strong discussion sections:

Uncertainty sources: Name specific sources tied to specific steps. Random vs systematic distinction. Major contribution vs minor.
Limitations: What the investigation cannot conclude. Are controlled variables truly constant? Is the range of the IV sufficient? Is the data sufficiently varied?
Improvements: Specific changes that would reduce uncertainty or extend the result. Use a longer pendulum; use a photogate instead of stopwatch; use more trials per IV value.
Comparison to theory: Calculate the predicted value; compare experimental result within uncertainty. State whether the result supports the hypothesis.

Figure 3. A random scatter of residuals around zero supports the linear model. A curved or wedge-shaped pattern would flag a missing physical effect or a poor fit.

Conclusion

A direct answer to the research question. Concise. Include the experimental value with uncertainty, and the comparison to the accepted value within uncertainty.

Marking criteria

VCAA's published criteria reward:

Investigation design. Quality of research question, hypothesis, methodology.
Quality of data. Range, precision, repeats.
Analysis. Linearisation, gradient extraction, uncertainty handling.
Discussion. Sophisticated treatment of uncertainty, limitations, improvements.
Scientific communication. Poster design, scientific writing.

Top band requires excellence in all five.

Check your knowledge

A focused set on the Unit 4 practical investigation in the VCAA poster-rubric style: uncertainty, log-log graph reading, projectile/circular motion data analysis. Attempt under exam conditions before checking the solutions block.

Distinguish between systematic and random uncertainty, and state how each is reduced. (3 marks)
A student measures the period of a simple pendulum five times and obtains 1.42, 1.39, 1.45, 1.41, 1.43 s. (a) Calculate the mean and the absolute uncertainty (use half the range as the simple estimate). (b) Express the result as $T \pm \Delta T$ with appropriate significant figures and units. (3 marks)
A practical investigation tests the relationship $T = k L^n$ for a pendulum's period vs length. The student plots $\log T$ versus $\log L$ and obtains a straight line of gradient $0.501 \pm 0.005$ and $y$ -intercept $0.305$ . (a) Read $n$ from the gradient and confirm consistency with theory. (b) From the intercept, calculate $k$ and compare with the theoretical value $2\pi / \sqrt{g}$ taking $g = 9.80 \ \text{m s}^{-2}$ . (c) Calculate the percent error in $k$ and identify the most likely source of any discrepancy. (6 marks)
A trolley is rolled down a ramp from rest. Photogate timings give: $d = 0.50, 1.00, 1.50, 2.00 \ \text{m}$ at $t = 0.80, 1.10, 1.32, 1.55 \ \text{s}$ . (a) Plot in your head $d$ versus $t^2$ and explain why the relationship is expected to be linear. (b) Use the first and last points to estimate the acceleration. (c) Identify two possible reasons the acceleration is less than the value predicted by $a = g \sin\theta$ for a frictionless ramp. (6 marks)
(a, 3) A Year 12 student investigates the projectile range of a foam dart fired horizontally from a launcher 1.20 m above the floor. Average range from five trials: $2.85 \pm 0.06 \ \text{m}$ . Calculate the launch speed. (b, 2) State two assumptions that must hold for the calculation to be valid. (5 marks)
(a, 3) State the difference between the accuracy and the precision of a measurement, and give an example where a result is precise but inaccurate. (b, 3) An ammeter records 2.34, 2.35, 2.33 A on three trials. The known true value is 2.40 A. Comment on the precision and the accuracy, and suggest one calibration step. (6 marks)
A student measures the spring constant of a small mass-spring system by recording the period of vertical oscillation for five different masses. Data (mass kg, period s): (0.100, 0.282), (0.200, 0.401), (0.300, 0.491), (0.400, 0.568), (0.500, 0.635). (a) State the expected linear relationship between $T^2$ and $m$ . (b) Calculate the gradient of $T^2$ vs $m$ from the smallest and largest mass, and from it determine $k$ . (c) Estimate the uncertainty in $k$ given a $\pm 0.005$ s uncertainty in each period reading. (7 marks)
(a, 2) Define proper uncertainty for a derived quantity $z = x^a y^b$ . (b, 4) A pendulum of length $L = (1.00 \pm 0.01) \ \text{m}$ has measured period $T = (2.005 \pm 0.005) \ \text{s}$ . Calculate $g$ and its absolute uncertainty using $g = 4\pi^2 L / T^2$ . (6 marks)

Solutions

Q1: Systematic uncertainty shifts every measurement in the same direction by an unknown but fixed amount (e.g., a stopwatch that runs 1 percent fast, an incorrectly zeroed metre rule). It is reduced by calibrating instruments against known standards or improving the experimental procedure. Random uncertainty produces unpredictable scatter from trial to trial (e.g., reaction time in starting a stopwatch, vibration of the apparatus). It is reduced by averaging many trials, since the random error scales as $1/\sqrt{N}$ .
Q2: (a) Mean $= (1.42 + 1.39 + 1.45 + 1.41 + 1.43)/5 = 7.10/5 = 1.420 \ \text{s}$ . Range $= 1.45 - 1.39 = 0.06 \ \text{s}$ . Half-range $= 0.03 \ \text{s}$ . (b) $T = (1.42 \pm 0.03) \ \text{s}$ to the same precision as the uncertainty.
Q3: (a) $\log T = \log k + n \log L$ . The gradient is $n = 0.501 \pm 0.005$ . The theoretical exponent for a simple pendulum is $1/2$ ( $T = 2\pi \sqrt{L/g}$ ), in excellent agreement. (b) $\log k = 0.305$ , so $k = 10^{0.305} = 2.018$ . Theoretical $k = 2\pi/\sqrt{g} = 6.283 / 3.130 = 2.007$ . (c) Percent error $= |2.018 - 2.007| / 2.007 \times 100 = 0.55$ percent, well within experimental precision. The small discrepancy is consistent with finite-amplitude correction (the small-angle approximation breaks down for large swings) and air drag.
Q4: (a) For motion from rest with constant acceleration, $d = \frac{1}{2} a t^2$ , so $d$ versus $t^2$ is linear with gradient $a/2$ and passing through the origin. (b) $a = 2(d_4 - d_1)/(t_4^2 - t_1^2) = 2(2.00 - 0.50)/((1.55)^2 - (0.80)^2) = 3.00 / (2.4025 - 0.64) = 3.00 / 1.7625 = 1.70 \ \text{m s}^{-2}$ . (c) (i) Rolling friction and axle friction at the wheels dissipate kinetic energy. (ii) Air resistance opposes motion. (iii) The trolley may not start exactly at rest; reaction-time error in the start trigger adds a small systematic offset.
Q5: (a) Time to fall: $h = \frac{1}{2} g t^2 \Rightarrow t = \sqrt{2h/g} = \sqrt{2 \times 1.20 / 9.80} = 0.495 \ \text{s}$ . Horizontal launch speed $u = R/t = 2.85 / 0.495 = 5.76 \ \text{m s}^{-1}$ . Absolute uncertainty in $u$ is dominated by the range uncertainty: $\Delta u / u = 0.06 / 2.85 = 2.1$ percent, so $u = (5.76 \pm 0.12) \ \text{m s}^{-1}$ . (b) (i) Air resistance is negligible (the foam dart is light and slow but not in vacuum). (ii) The launch is exactly horizontal and at exactly the recorded launch height; tilt or measurement error in the launch height would systematically shift the inferred speed.
Q6: (a) Precision is the spread of repeated measurements about their mean (small spread = high precision). Accuracy is closeness to the true value (small offset from true = high accuracy). A bathroom scale that reads 75.3, 75.4, 75.3 kg for a person whose true mass is 70.0 kg is precise but inaccurate. (b) Mean $= 2.34 \ \text{A}$ , spread $\pm 0.01 \ \text{A}$ . Precision is high (0.4 percent spread). Accuracy is low: the mean is 0.06 A (2.5 percent) below the true value. Suggested calibration: pass a known reference current through the meter and apply an offset, or replace with a calibrated standard.
Q7: (a) $T = 2\pi\sqrt{m/k}$ , so $T^2 = (4\pi^2 / k) \times m$ . A plot of $T^2$ versus $m$ is linear with gradient $4\pi^2 / k$ and intercept zero (for a massless spring). (b) $T^2$ at $m = 0.100$ is $0.0795 \ \text{s}^2$ ; at $m = 0.500$ is $0.403 \ \text{s}^2$ . Gradient $= (0.403 - 0.0795)/(0.500 - 0.100) = 0.324 / 0.400 = 0.809 \ \text{s}^2 \text{kg}^{-1}$ . $k = 4\pi^2 / 0.809 = 39.48 / 0.809 = 48.8 \ \text{N m}^{-1}$ . (c) The relative uncertainty in $T^2$ is $2 \Delta T / T \approx 2 \times 0.005 / 0.45 = 2.2$ percent on a typical reading. Propagated to $k$ (also 2.2 percent at first order), $\Delta k \approx 1.1 \ \text{N m}^{-1}$ , so $k = (48.8 \pm 1.1) \ \text{N m}^{-1}$ .
Q8: (a) For $z = x^a y^b$ , relative uncertainties add (in quadrature for independent random errors, or absolutely as a conservative bound): $\Delta z / z = |a| (\Delta x / x) + |b| (\Delta y / y)$ . (b) $g = 4\pi^2 L / T^2 = 4\pi^2 \times 1.00 / (2.005)^2 = 39.48 / 4.020 = 9.82 \ \text{m s}^{-2}$ . Relative uncertainty: $\Delta g / g = \Delta L / L + 2 \Delta T / T = 0.010 + 2(0.005/2.005) = 0.010 + 0.005 = 0.015$ , i.e. 1.5 percent. Absolute $\Delta g = 0.015 \times 9.82 = 0.15 \ \text{m s}^{-2}$ . So $g = (9.82 \pm 0.15) \ \text{m s}^{-2}$ , consistent with the accepted value $9.80 \ \text{m s}^{-2}$ .