How to Write an IA That Stands Out

What the IA Tests

The Internal Assessment is worth 20% of your final IB Physics grade. Unlike the exam, the IA tests your ability to design, execute, and evaluate an original scientific investigation. You have full control — which is both an opportunity and a challenge.

The IA is marked against 4 criteria, each worth a maximum of 6 marks (total: 24):

Choosing a Research Question

This is the most important decision. A strong research question:

• Is specific and measurable (not vague)
• Has a clear independent and dependent variable
• Can be investigated with school-level equipment
• Produces enough data points for meaningful analysis (minimum 5, ideally 8–10)
• Has an underlying physics theory to compare against

Strong Research Questions

• “How does the length of a pendulum affect its period of oscillation?”
• “What is the relationship between the angle of incidence and the critical angle for total internal reflection in glass?”
• “How does the mass attached to a spring affect the frequency of oscillation?”

Weak Research Questions

• ❌ “What is gravity?” (Not experimentally testable)
• ❌ “Does music affect concentration?” (Not physics)
• ❌ “How does temperature affect resistance in a superconductor?” (Requires equipment beyond school level)

Research Design (6 marks)

For Top Marks (5–6):

1. State the research question clearly in the introduction
2. Identify all variables:
   • Independent variable (what you change)
   • Dependent variable (what you measure)
   • Controlled variables (at least 3, with specific methods of control)
3. Describe the methodology in enough detail that someone could replicate your experiment
4. Justify your range and intervals — Why did you choose these values?
5. Include a labelled apparatus diagram
6. Address safety and ethical considerations

Common Pitfalls

• Methodology too vague (“I measured the length”)
• Not enough controlled variables
• No justification for the chosen range
• Missing diagram

Data Collection (6 marks)

For Top Marks (5–6):

1. Record raw data in a clear table with:
   • Column headers including quantity, unit, and uncertainty
   • At least 5 trials per data point
   • Consistent significant figures
2. State instrument uncertainties (e.g., ruler: ±0.5 mm, stopwatch: ±0.01 s)
3. Include qualitative observations where relevant
4. Show awareness of anomalies — don’t delete outliers, flag them

Example Data Table

Data Analysis (6 marks)

For Top Marks (5–6):

1. Process the data — calculate derived quantities (e.g., T²) with sample calculations shown
2. Propagate uncertainties through all calculations
3. Plot an appropriate graph — linearise if necessary
4. Draw error bars on your graph
5. Draw best-fit and worst-fit lines
6. Calculate the gradient with uncertainty
7. Derive the physical quantity from the gradient and compare with the accepted value
8. Calculate percentage discrepancy

The Graph Is Your Centrepiece

The graph is the most important element of your IA. It should:

• Fill an entire page
• Have clear, labelled axes
• Show error bars
• Include both best-fit and worst-fit lines
• Have a title

Evaluation (6 marks)

For Top Marks (5–6):

1. State your conclusion — does the data support the hypothesis?
2. Compare with theory — calculate percentage discrepancy
3. Assess the reliability of your data:
   • Were repeat measurements consistent?
   • Is the percentage uncertainty acceptable?
4. Identify specific sources of error:
   • Systematic errors (with direction of effect)
   • Random errors (with evidence from data spread)
5. Suggest specific, realistic improvements:
   • ❌ “Use better equipment” (0 marks)
   • ✅ “Use a photogate timer (±0.001 s) instead of a handheld stopwatch (±0.2 s) to reduce the percentage uncertainty in period measurement from 22% to 0.1%”
6. Suggest meaningful extensions to the investigation

Formatting and Presentation

• Word count: 6–12 pages of text (excluding data tables and graphs). There’s no strict limit, but examiners penalise unnecessary length
• Structure: Introduction → Method → Data → Analysis → Evaluation
• Graphs: Generated digitally (Excel, Desmos, Logger Pro) for clarity
• References: Cite any sources used (textbooks, online resources) in a bibliography
• Academic honesty: All work must be your own. Plagiarism detection is used

The IA That Scores 23–24/24

The highest-scoring IAs share these qualities:

1. An original, personally meaningful research question
2. Meticulous data with repeated trials and recorded uncertainties
3. A linearised graph with error bars, best-fit, and worst-fit lines
4. A quantitative evaluation comparing experimental and theoretical values
5. Specific improvements that directly address identified weaknesses
6. Clear, concise writing that lets the physics speak