IB Physics IA: 70+ Ideas, the Criteria, and How to Score 24

Most students start their IB Physics IA by asking the wrong question. Not "what's a good research question" — but "what can I actually measure well in my school lab?"

That second question is the one that wins marks. The IB doesn't reward clever ideas. It rewards clean measurements, honest uncertainties and a conclusion that follows from your own numbers.

This guide covers the whole task: how the four criteria work, how to build a research question that scores, how to handle uncertainties like a physicist, and 70+ example research questions sorted by syllabus theme.

Quick answer: The IB Physics IA (Internal Assessment) is a single scientific investigation worth 20% of your final Physics grade at both Standard Level (SL) and Higher Level (HL). You design your own research question, collect and process your own data, and write it up in a report of no more than 3,000 words. It's marked out of 24 across four equal criteria: Research design, Data analysis, Conclusion, and Evaluation. The IB allocates about 10 hours of class time to it. Your teacher marks it, then the IB moderates a sample from your school.

What is the IB Physics IA?

The IB Physics IA is a self-designed physics experiment, written up as a report, worth 20% of your grade.

IA stands for Internal Assessment. "Internal" means your own teacher marks it first, not an examiner in another country. The IB then re-marks a sample of reports from your school to check the marking is fair everywhere. That check is called moderation.

The official name for the task is the scientific investigation. You pick a physics question, plan a method, take measurements, process the data and reach a conclusion. The report you hand in is the only thing that gets marked. Nobody watches you do the experiment.

One thing many students miss: the IA is identical for SL and HL. Same four criteria, same 24 marks, same 3,000-word limit, same 10 hours. Higher Level students don't get a harder IA. They just cover more syllabus content in the exams.

The current rules come from the IB Physics guide with first assessment in May 2025. That guide replaced the old five-criterion rubric (Personal engagement, Exploration, Analysis, Evaluation, Communication) with four criteria. So any exemplar written before 2025 was marked against a different rubric.

The 2025 guide also replaced the old Group 4 Project with the collaborative sciences project, a separate 10-hour team task. That project is not your IA. Don't confuse the two.

How much is the IB Physics IA worth?

The IB Physics IA is worth 20% of your final Physics grade at both SL and HL.

The rest comes from two written exam papers. Paper 1 is multiple choice plus data-based questions. Paper 2 is short-answer and extended-response questions.

That 20% is the most controllable grade in the whole course. You choose the topic and the equipment. You control the conditions and the deadline. Nobody is timing you in an exam hall.

Physics is marked out of 7. A strong IA can be the difference between a 5 and a 6, and it's banked months before exams. A bad morning in Paper 2 can still be rescued by 21 out of 24 on the IA.

How is the IB Physics IA marked? The four criteria

The IB Physics IA is marked out of 24, split evenly across four criteria worth 6 marks each: Research design, Data analysis, Conclusion, and Evaluation.

Each criterion is worth 25% of the IA mark. There is no criterion for effort, presentation, or personal interest any more. Write the report to hit these four things and you score. Write a generic school lab report and you leave marks on the table.

Criterion A: Research design (6 marks)

Research design assesses how well you explain what you set out to measure and how you measured it.

For the top band the IB wants three things. Your research question must be "described within a specific and appropriate context". Your methodological choices must be explained, not just listed. And your method description must let another student reproduce the investigation.

The guide is specific about "context". Your research question should name the dependent and independent variables, describe the system briefly, and bring in background theory that is directly relevant. Not a page of textbook copying — just the equation and the reasoning that make your prediction sensible.

Methodological considerations you're expected to explain include:

• How you chose to measure the dependent and independent variables

• Why you picked that range, that interval, and that number of repeats

• Which variables you controlled and how you controlled them

• Any safety, ethical or environmental issues you had to handle

The word "explain" is doing the heavy lifting. "I measured five lengths" is stating. "I used lengths from 20 cm to 100 cm in 20 cm steps because below 20 cm the period drops under 0.9 s, and my timing uncertainty of ±0.2 s would then be over 20% of the reading" is explaining. That sentence alone moves you a band. The state–describe–explain ladder runs through all four criteria, so learn where each verb sits.

Criterion B: Data analysis (6 marks)

Data analysis assesses whether you recorded, processed and presented your data in a way that is clear, precise and relevant to your question.

Three things sit in the top band: communication that is both clear and precise, an appropriate consideration of uncertainties, and processing that is accurate.

"Precise" here has a specific meaning. It means following conventions correctly — units in table headers, consistent decimal places, sensible significant figures, properly labelled axes. Marks are lost here every year for reasons that have nothing to do with physics.

"Consideration of uncertainties" is where Physics IAs are won and lost. More on that below.

Criterion C: Conclusion (6 marks)

Conclusion assesses whether you actually answered your research question, using your own analysis and the accepted physics.

The top band needs a conclusion that is "justified" and "fully consistent with the analysis presented". The guide adds that being fully consistent "requires the interpretation of processed data including associated uncertainties". Translation: quote your gradient with its uncertainty, then say whether the accepted value sits inside that range.

You also need a relevant comparison to published physics — a textbook value, a constant, a paper. Cite it properly so the moderator can trace it.

A conclusion that says "the results supported my hypothesis" with no numbers sits in the bottom band. A conclusion that says "the gradient was 4.02 ± 0.14 s²m⁻¹, which gives g = 9.82 ± 0.34 ms⁻², consistent with the accepted 9.81 ms⁻² at this latitude" is top band material.

Criterion D: Evaluation (6 marks)

Evaluation assesses how well you critique your own method and suggest improvements.

The rubric distinguishes three levels sharply. Stating generic weaknesses is the bottom band. Describing specific weaknesses is the middle. Explaining the relative impact of specific weaknesses is the top.

"Relative impact" is the key phrase. You have to rank your problems. Which one damaged your result most, and by roughly how much? "Human reaction time added about ±0.2 s per timing, which is 4% of my shortest period, while the ±1 mm ruler uncertainty was under 0.5% — so timing dominated my total uncertainty" is exactly what the top band asks for.

Every improvement must then connect to a weakness you already named, and be realistic. "I would use a laser interferometer" is not realistic in a school lab. "I would use a light gate to remove reaction time" is.

Uncertainties and gradients: the physics-specific marks

Uncertainty analysis is the single biggest thing that separates a strong Physics IA from a weak one.

Biology and chemistry IAs often lean on statistical tests. Physics leans on measurement uncertainty, error propagation and graph gradients. Get this right and Criteria B, C and D all improve at once. Here's the minimum a top-band Physics IA does.

Linearisation deserves its own paragraph. If theory says T = 2π√(L/g), don't plot T against L and draw a curve. Plot T² against L. Theory then predicts a straight line of gradient 4π²/g through the origin. Your gradient hands you g directly, your max–min gradients hand you the uncertainty in g, and your intercept tells you whether you had a systematic error. That single decision is worth more marks than almost anything else you can do in a Physics IA.

What makes a good IB Physics IA research question?

A good IB Physics IA research question names one independent variable, one measurable dependent variable, a specific system, and the relationship you're testing.

The reliable template is this:

How does [independent variable, with range and unit] affect [dependent variable, with unit] of [specific system], and does the relationship match [the theory you're testing]?

That last clause is what pushes you from a fair question into a strong one. It gives you a prediction to test, a linearised graph to plot, and a published value to compare against in your conclusion.

Compare these:

Before you commit, run your idea through these six checks:

1. Can I change the independent variable in at least five steps? Fewer than five points makes a gradient meaningless.

2. Can I measure the dependent variable to better than about 5%? If not, your uncertainty swamps your trend.

3. Does theory predict a specific relationship? You need something to linearise and something to compare with.

4. Can I actually control the other variables? Room temperature, air currents, friction, and mains voltage all drift.

5. Do I have the equipment now? Not equipment you hope to borrow in three weeks.

6. Is it safe and allowed? No unshielded radiation sources, no mains electricity outside a lab supply, no lasers pointed at people.

Stuck here? Our free IB IA topic generator throws out subject-specific starting points, and the questions below are ready to adapt.

70+ IB Physics IA ideas and example research questions

The strongest IB Physics IA ideas come from the five syllabus themes, because the theory you need is already in your course. There are 72 below.

Each one below is written as a usable research question. Treat them as starting points — narrow the range, name your system, add the relationship you'll test.

One warning before you scroll. Popular topics are not banned, but they are crowded. A pendulum IA can score 24 out of 24 if your uncertainty work is excellent. It can also score 12 if you time ten swings by hand and plot a curve. The topic doesn't decide your mark. The measurement quality does.

Theme A: Space, time and motion (15 ideas)

1. How does the angle of an inclined plane affect the acceleration of a trolley, and does the gradient give g sin θ?

2. How does the mass of a trolley affect its acceleration under a constant applied force, and is a proportional to 1/m?

3. How does the surface material under a wooden block affect the coefficient of kinetic friction?

4. How does the normal force on a block affect the limiting static friction force, and is the relationship linear?

5. How does the drop height of a squash ball affect its coefficient of restitution?

6. How does the temperature of a squash ball affect its coefficient of restitution?

7. How does the launch angle of a spring-loaded projectile affect its horizontal range, and does the maximum sit at 45°?

8. How does the number of stacked paper coffee filters affect their terminal velocity?

9. How does the diameter of a steel sphere affect its terminal velocity falling through glycerol, and does the data support Stokes' law?

10. How does the mass ratio of two colliding trolleys affect the fraction of kinetic energy lost in the collision?

11. How does the spin rate of a table tennis ball affect its sideways deflection in flight (the Magnus effect)?

12. How does the water fill fraction of a sealed cylinder affect its acceleration rolling down a ramp?

13. How does the distance of two added masses from the pivot affect the angular acceleration of a rotating rod?

14. How does the radius of circular motion affect the tension in the string at a fixed angular speed, and is T proportional to r?

15. How does the length of a paper "tumbling wing" affect its steady descent speed?

Theme B: The particulate nature of matter (15 ideas)

1. How does the surface colour of a metal can affect the rate of cooling of the water inside it?

2. How does the volume of water in an open beaker affect its cooling constant under Newton's law of cooling?

3. How does the thickness of insulating wrap affect the rate of heat loss from a beaker of hot water?

4. How does the mass of ice added to warm water affect the specific latent heat of fusion calculated for water?

5. How does the salt concentration of water affect its measured specific heat capacity?

6. How does the carbon dioxide concentration in a sealed chamber affect its equilibrium temperature under a lamp?

7. How does the temperature of a fixed volume of air affect its pressure, and where does extrapolation put absolute zero?

8. How does the volume of trapped air affect its pressure at constant temperature, and is the product pV constant?

9. How does the length of a nichrome wire affect its resistance, and is R proportional to L?

10. How does the diameter of a nichrome wire affect its resistance, and is R proportional to 1/A?

11. How does the temperature of a copper coil affect its resistance, and what temperature coefficient does the gradient give?

12. How does light intensity affect the resistance of a light-dependent resistor, and does it follow a power law?

13. How does the external load resistance affect the power delivered by a AA cell, and does the peak sit at R equal to the internal resistance?

14. How does the number of identical cells in series affect the measured total internal resistance?

15. How does electrolyte concentration affect the EMF of a simple homemade cell?

Theme C: Wave behaviour (16 ideas)

1. How does the length of a simple pendulum affect its period, and does the data support T proportional to √L?

2. How does the release angle of a pendulum affect its period, and at what angle does the small-angle approximation break down?

3. How does the mass hung on a vertical spring affect the period of oscillation, and is T proportional to √m?

4. How does the number of active coils in a spring affect its spring constant?

5. How does the volume of water in a wine glass affect its resonant frequency?

6. How does the length of an air column in a tube affect the resonant frequency, and how do open and closed ends compare?

7. How does the tension in a guitar string affect its fundamental frequency, and is f proportional to √T?

8. How does the linear density of a string affect its fundamental frequency at fixed tension?

9. How does water depth in a shallow tray affect the speed of a surface wave, and does v depend on √d?

10. How does slit separation affect fringe spacing in a double-slit interference pattern?

11. How does the groove spacing of a CD or DVD used as a diffraction grating affect the measured diffraction angles?

12. How does the sugar concentration of a solution affect its refractive index?

13. How does the angle between two polarising filters affect transmitted light intensity, and does it follow cos²θ?

14. How does the viscosity of the surrounding liquid affect the damping constant of an oscillating mass?

15. How does the driving frequency of a mass-spring system affect its amplitude near resonance, and how does damping change the peak width?

16. How does the speed of a moving buzzer affect the measured Doppler frequency shift?

Theme D: Fields (12 ideas)

1. How does the timing method used (pendulum, light gate free-fall, or video analysis) affect the value and uncertainty of g?

2. How does the separation of two magnets affect the repulsive force between them, and what power law fits the data?

3. How does the current through a solenoid affect the magnetic flux density at its centre?

4. How does the number of turns per unit length of a solenoid affect the flux density at its centre?

5. How does axial distance from the end of a bar magnet affect the flux density measured with a Hall probe?

6. How does the angle between a current-carrying wire and a magnetic field affect the force on the wire?

7. How does the wall thickness of a copper tube affect the fall time of a magnet dropped through it?

8. How does the number of turns in a pickup coil affect the peak EMF induced by a falling magnet?

9. How does the rotation frequency of a coil in a magnetic field affect the peak induced EMF?

10. How does the plate separation of a homemade parallel-plate capacitor affect its capacitance?

11. How does the overlap area of two aluminium foil plates affect measured capacitance, and is C proportional to A?

12. How does the dielectric material between two plates affect the measured capacitance and the relative permittivity calculated from it?

Theme E: Nuclear and quantum physics (8 ideas)

1. How does the thickness of aluminium shielding affect the count rate from a gamma source, and what attenuation coefficient does that give?

2. How does the shielding material at equal areal density affect gamma count rate?

3. How does distance from a gamma source affect count rate, and does it follow an inverse-square law?

4. How does the colour of an LED affect its threshold voltage, and what value of Planck's constant does the gradient give?

5. How does the operating voltage of a filament lamp affect the peak wavelength of its emitted light?

6. How does light intensity affect the short-circuit current and fill factor of a small solar cell?

7. How does the angle of incidence of light affect the power output of a photovoltaic cell, and does it follow a cosine law?

8. How does the measured half-life of a protactinium generator compare with the accepted value?

Database and simulation investigations (6 ideas)

The IB explicitly allows investigations built on databases, simulations and modelling — not only hands-on lab work. Worth considering if your lab is short on equipment, or if you want an astrophysics angle.

1. How does the orbital radius of a planet relate to its orbital period, and do published solar system values support Kepler's third law?

2. How does exoplanet orbital period relate to semi-major axis in the NASA Exoplanet Archive?

3. How does a main-sequence star's mass relate to its luminosity in published catalogue data?

4. How does the recessional velocity of galaxies vary with distance, and what Hubble constant do published data give?

5. How does the tidal range at one coastal station vary with the phase of the Moon, using published tide-gauge records?

6. How does the damping coefficient in a simulated driven oscillator affect the height and width of the resonance peak?

Two cautions on this route. First, you still need a defensible uncertainty discussion — quote the uncertainties the database publishes. Second, "I downloaded a spreadsheet" is not a method; explain your selection criteria, your sample size and why that dataset answers your question.

How to structure your IB Physics IA report

There's no compulsory template, but the section order below maps cleanly onto the four criteria and is what moderators expect to see.

Two rules about the word count. The limit is 3,000 words maximum, and the IB's guidance excludes charts, diagrams, data tables, equations, calculations, citations, your bibliography and headers. So the limit applies to your prose, not your data.

The second rule: don't front-load. Students routinely spend 1,200 words on background theory and 200 on evaluation. That's backwards. Evaluation is 6 marks — the same as Research design. Give it the space it's worth.

Put your title, your IB candidate code and your final word count at the start of the report. If you worked in a group, list every member's candidate code too.

Our wider guide on how to write an IB internal assessment covers the writing process across subjects, and the sibling guide to the IB Biology IA helps if you're doing both sciences.

How to score 24 on the IB Physics IA

Full marks on the IB Physics IA come from depth on a narrow question, not breadth on a big one. Here's what the reports that hit 22 to 24 have in common.

1. One variable, five or more values, five or more repeats. Twenty-five to thirty measurements minimum. More data points beat a more exciting topic every time.

2. A range chosen for a stated reason. Not "0 to 100" but "0 to 60 °C, because above 60 °C evaporation changed the mass by more than 2%".

3. Every measurement carries an uncertainty from the start. Put the ± in your raw data table headers, not bolted on at the end.

4. A linearised graph. Plot the form that theory predicts as a straight line, with error bars and max–min gradient lines.

5. A final value with a ± and a published comparison. State whether the accepted value falls inside your uncertainty range, and say what it means if it doesn't.

6. Weaknesses ranked by size. Give the percentage each source contributed, so the examiner can see you know which one mattered.

7. Improvements that trace back to named weaknesses. One-to-one mapping. No generic wish lists.

8. A pilot run. Two hours of trial measurements will change your range, your interval and your repeat count. Every strong IA has one behind it.

None of that is about being brilliant at physics. It's about being careful, and honest about what your instruments could and couldn't do.

Common IB Physics IA mistakes to avoid

The mistakes below cost real marks every session, and every one is avoidable.

• Timing one oscillation instead of many. Time 20 swings and divide. It cuts your reaction-time uncertainty by a factor of 20.

• No error bars. A graph without error bars cannot support a top-band Data analysis mark. Neither can a gradient without an uncertainty.

• Plotting the curve instead of the linearised form. A curved graph tells the examiner very little. A straight line with a physically meaningful gradient tells them everything.

• Confusing precision with accuracy. Five readings that agree closely can still all be wrong by the same systematic amount. Say which problem you had.

• A research question with two independent variables. Change one thing at a time. Two variables makes the data uninterpretable.

• Writing generic evaluations. "Human error" and "equipment limitations" score nothing. Name the instrument, quote the size, rank the impact.

• Running out of time. Data collection always takes longer than planned. Book lab time early and expect to repeat a session.

• Inconsistent significant figures. If your ruler reads to the nearest millimetre, a length of "0.4237 m" is not defensible.

How long does the IB Physics IA take?

The IB allocates about 10 hours of class time to the scientific investigation, but most students spend 25 to 40 hours in total across a few months.

Those 10 hours are the school's teaching allocation, not a cap on your own effort. Here's a realistic timeline.

Your teacher may read and comment on one draft. That feedback is your biggest free upgrade, so hand in a complete draft — one missing its evaluation wastes the most valuable comments you'll get.

Most schools set their internal deadline in the first term of Year 2, well before the IB's submission window. Ask for that date now and work backwards.

How the Physics IA differs from Biology and Chemistry

All three sciences share the same four criteria, 24 marks and 3,000-word limit — but what earns Data analysis marks differs.

So don't reuse your biology approach in physics. A physics report that runs a t-test but never propagates an uncertainty has missed the point of the subject.

Where the IA fits in your IB Diploma

Every Diploma subject has an internally assessed component, and they tend to fall due in the same few months of Year 2. Add the Extended Essay, the TOK exhibition and essay, and CAS, and you get one very crowded term.

Two things follow. First, start your Physics IA earlier than feels necessary — the students who suffer are the ones whose data collection collides with EE deadline week. Second, the IA is the part of your Physics grade that responds best to help. It's the only assessed work where a specialist can read your draft and tell you which band you're in and why.

Educifly's IB Physics specialists work through IA drafts against the four criteria line by line, on the same syllabus your teacher is using. If your report is written but you don't know whether the uncertainty work is strong enough, that's a 45-minute conversation rather than a term of guesswork. Have a look at our online IB Physics tutoring or our dedicated IB IA help, and you can test the fit with a free trial class first.

Frequently asked questions about the IB Physics IA

What is the IB Physics IA?

The IB Physics IA is the Internal Assessment for the course: a single scientific investigation you design, run and write up yourself. It's officially called the scientific investigation. It's worth 20% of your final Physics grade at both SL and HL, it's marked out of 24 across four criteria, and the report has a maximum of 3,000 words.

How much is the IB Physics IA worth?

The IB Physics IA is worth 20% of your final Physics grade, identically at SL and HL. Paper 1 is worth 36% and Paper 2 is worth 44%. Because the IA is completed and marked before exam season, it's the most controllable 20% in the whole course.

How is the IB Physics IA marked?

It's marked out of 24 across four equal criteria: Research design (6 marks), Data analysis (6), Conclusion (6) and Evaluation (6). Each criterion is 25% of the IA mark. Your teacher marks the report first, then the IB moderates a sample of reports from your school so that marking is consistent worldwide.

How many words is the IB Physics IA?

The maximum is 3,000 words. The IB's guidance excludes charts, diagrams, data tables, equations, formulas, calculations, citations, your bibliography and headers from that count. So the 3,000 words apply to your prose. Most students find the limit generous once tables and calculations are excluded.

Is the IB Physics IA the same for SL and HL?

Yes. Same task, same four criteria, same 24 marks, same 3,000-word limit and the same 10 hours of allocated class time. Higher Level students are not given a harder investigation. The SL and HL difference sits in the syllabus content and the exam papers, not in the IA.

What is a good IB Physics IA research question?

A good one names a single independent variable with a range and unit, a measurable dependent variable with a unit, a specific system, and the relationship being tested. For example: "How does the length of a simple pendulum (20–100 cm) affect its period (s), and does the data support T proportional to the square root of L?" That structure gives you something to linearise, something to compare with, and a clear conclusion to write.

Do I have to do a hands-on experiment for the Physics IA?

No. The IB allows investigations based on hands-on laboratory work, fieldwork, spreadsheets and modelling, extraction and analysis of data from databases, and simulations. A database or simulation investigation can score full marks. You still need to justify your data selection and handle uncertainties properly — that expectation doesn't change with the method.

How important are uncertainties in the IB Physics IA?

Very. Uncertainty handling runs through three of the four criteria. Criterion B requires an appropriate consideration of uncertainties. Criterion C says a fully consistent conclusion needs the interpretation of processed data including its uncertainties. Criterion D rewards explaining the relative impact of specific weaknesses, which is impossible without percentage uncertainties. If you improve one thing in your draft, improve this.

Can I work with other students on my Physics IA?

You can collect data as a small group, but the investigation and the report must be your own. Each student in a group is expected to investigate a different independent variable, and group reports are not accepted. Your teacher needs to know the group structure, and each report must be individually written.

How many drafts can my teacher check?

Your teacher may give feedback on one draft. That feedback can be spoken or written in general terms, but your teacher cannot annotate the draft with the corrections you should make. Because you only get one round, submit a complete draft — including the conclusion and evaluation — rather than a partial one.

When should I start my IB Physics IA?

Ideally in the last term of Year 1 or the very start of Year 2. Choosing a topic and running a pilot before the summer break means your data collection lands well ahead of Extended Essay and TOK deadlines. Most schools set the internal IA deadline in the first term of Year 2, so work backwards from that date.

Can I use an old Physics IA exemplar as a model?

Read them for topic ideas, but not for structure. Any exemplar written for assessment before May 2025 was marked against the old five-criterion rubric, which included Personal engagement and Communication. Those criteria no longer exist. Use the current four criteria as your checklist instead.