Your Scientific Investigation is worth 20% of your final IB Physics grade. That is not a small number. Done well, it can be the component that pulls your grade up. Done poorly, or rushed, it drags everything else down with it.
The good news is that the IA is one of the most predictable parts of the entire IB. There is a clear structure, a clear set of marking criteria, and a very specific type of research question that examiners reward. You are not trying to guess what they want. It is all here.
This guide covers everything: choosing your topic, collecting your data, writing your report, and understanding where marks are awarded and lost. Work through it from the start, or jump to the section you need.
1. Choosing Your Research Question
Does your topic have to be original?
No. This is one of the most persistent myths in IB Physics, and it costs students time they do not have.
Around 25,000 IB Physics students submit a Scientific Investigation every year. All of them are working within the same syllabus, investigating the same phenomena. It is genuinely impossible for every research question to be original, and the IB does not require it.
The IB examiners have been clear on this point in their published subject reports. Using an idea found online, in a book, or from a previous student is perfectly acceptable, provided you reference it properly, collect your own data, perform your own analysis, and write your own report.
You will not lose marks because your topic is not original.
What matters is the quality of your method, your data, and your analysis. Pick a topic quickly, pick one you can engage with, and then put your energy into doing the work well.
For a ready-made list, 100 IB Physics IA ideas covers every theme in the current IB Physics syllabus, with topics that have been tried and tested to produce clean data and high marks.
The format every top-scoring research question uses
IB examiners have been consistent in their feedback across multiple subject reports. The best investigations share one thing in common: a focused, well-defined research question with a single independent variable and a measurable dependent variable.
Investigations with multiple independent variables consistently score poorly. The examiners are explicit about this.
The format that works is straightforward:
“How does [Variable X] affect [Variable Y]?”
For example: “How does the volume of water in a wine glass affect the resonant frequency of the glass?”
This format immediately identifies your independent variable (volume of water), your dependent variable (resonant frequency), and implies a clear method and graph. It focuses your background theory, simplifies your method, and makes your conclusion easier to write.
If you are starting from scratch, structure your question in this format before you do anything else.
SL vs HL: is there a difference?
No. The marking criteria are identical for SL and HL. Examiners mark the work without knowing which level you are studying at. There is no such thing as an “HL topic” or an “HL standard” for the Scientific Investigation.
Do not waste time trying to find a topic that feels “HL enough.” Focus on finding a topic with a clean research question, accessible equipment, and measurable variables. That is all that matters.
2. Collecting Your Data
Identifying your variables
Before you run a single trial, get your variables clearly defined on paper. This is the foundation everything else is built on.
Using the research question format above makes this straightforward:
- Independent variable: the variable you deliberately change (Variable X)
- Dependent variable: the variable that responds (Variable Y)
- Control variables: everything else that must stay constant to keep the experiment fair
For the wine glass example: volume of water is the independent variable, resonant frequency is the dependent variable, and control variables include the temperature of the water, the force used to strike the glass, and the density of the glass itself.
Once your variables are defined, think carefully about how you will measure each one. Ask yourself: which apparatus gives the most precise measurement? Can human error be reduced? Can I repeat trials to reduce random error? Is the range of my independent variable large enough to show a clear trend?
These questions will directly improve your data quality and your marks in the Research Design criterion.
Collecting data at school, at home, or using a simulation
Most students collect data using school laboratory equipment, and this is usually the best option. Physical experiments lend themselves well to error analysis, which is a significant part of the marking criteria.
If you do not have access to school equipment, a top-scoring IA is still achievable at home. The key is choosing a topic that suits simple household materials. Smartphones can serve as surprisingly effective measuring instruments for timing, frequency, and light intensity experiments.
A small number of students use a simulation or database to collect data. Examiners do not penalise this, but they do expect a strong analytical approach. If you use a simulation, your analysis needs to go further: finding a physical function, deriving a numerical value, or testing a theoretical prediction against simulated data. A simulation that simply produces numbers and graphs without deeper analysis will not score well.
Group work: what is and is not allowed
Collaboration is permitted during the data collection stage, and the IB actively encourages it. However, the rules are specific.
Groups must be no larger than three students. Each student in the group must investigate a different independent variable. The final written report must be entirely individual work. A group report is not permitted under any circumstances.
So if you and a classmate are investigating the wine glass experiment together, one of you might vary the volume of water while the other varies the temperature of the water. You share the practical setup and collect data together. You then go away and write separate reports with separate research questions, separate analysis, and separate conclusions.
Your teacher must be aware of the group structure and the role each member is playing throughout the process.
3. Writing Your Report
The exact structure to use
Follow this structure precisely. Using these section headings in this order signals to the examiner that your report is organised and complete, and makes it much easier for them to award marks across all criteria.
- Report Details (title, candidate code, word count)
- Introduction
- Theory and Hypothesis
- Variables
- Diagram
- Apparatus
- Method
- Raw Data
- Processed Data
- Graph
- Analysis and Conclusion
- Evaluation
- References
For a detailed breakdown of what to include in each of these sections, the IB Physics IA structure guide covers every section with examples and common mistakes to avoid.
No title page. No contents page.
This catches people out every year.
At the top of your first page, include four things only: the title of your investigation, your IB candidate code, the candidate codes of any group members, and your word count. That is it. Four lines of plain text.
Do not create a title page. Do not create a contents section. These are not required, they add no marks, and they take up space that the examiner has to scroll through before they reach your actual work. Since IAs are marked digitally, anything that interrupts the flow of the examiner’s reading is a disadvantage to you.
The word count limit
The maximum word count for the report is 3,000 words. This is stricter than many students expect.
The following are not included in the word count: charts and diagrams, data tables, equations and calculations, citations and references, bibliography, and headers. Only your written prose counts.
For most well-structured reports, 3,000 words fits within roughly 12 pages. Reports that run to 13 or 14 pages are acceptable if the content is concise and every section earns its space. Padding, repetition, and lengthy background lectures on topics only loosely related to your research question will eat your word count without adding marks.
State the word count honestly at the top of your report. AI-assisted word count verification is now straightforward, and the IB uses it.
How to write your introduction
The introduction is one of the most misunderstood sections of the IA. Students either write three paragraphs of forced enthusiasm about their personal connection to bungee jumping, or they write a dense literature review citing 15 papers that have nothing to do with their experiment.
Neither approach works.
The examiner wants to see: what you are investigating, why it is worth investigating, and a brief indication of how you will investigate it. That is the entire purpose of the introduction.
Five sentences is enough. Around 150 words. Here is the structure:
- State your research question clearly
- Give a brief overview of the investigation
- Explain why the research question is relevant or interesting
- Show genuine personal engagement (one sentence, specific to your context)
- Keep it concise
The word “passionate” has become almost meaningless in IB introductions. Show your engagement through the specificity of your question and the quality of your method, not through adjectives.
Plagiarism
Your report will be checked.
Before submission, your teacher will review your initial proposal, your first draft, your references, and your writing style against your known standard. After that, the report goes through a web-based plagiarism detection service.
Using an idea from online is fine. Copying someone else’s analysis, method, or conclusion and presenting it as your own is not, and it is harder to conceal than students often assume. Your teacher knows how you write.
Use example reports as a reference for structure and depth. Do not use them as a template to copy from.
4. Understanding the Marking Criteria
The IB Physics Scientific Investigation is marked out of 24, across four criteria of 6 marks each. Understanding what each criterion is actually looking for is one of the most efficient things you can do to improve your mark.
Research Design (6 marks)
This assesses whether your research question is clearly defined, your methodology is logical and detailed, and your variables are properly identified and controlled. The examiner wants to see that a reader could replicate your experiment from your method alone.
Common problems: vague research questions, multiple independent variables, methods that skip steps, control variables that are listed but not explained.
Data Analysis (6 marks)
This assesses how well your data is recorded, processed, and presented. Your tables need proper headings, units, and uncertainties. Your processing (calculating means, propagating uncertainties, deriving quantities) needs to be shown clearly. Your graph needs labelled axes, a title, error bars where appropriate, and a line or curve of best fit.
Common problems: missing uncertainties, graphs without error bars, processing that is shown but not explained, data that is collected but not connected back to the research question.
Conclusion (6 marks)
This assesses whether your conclusion directly and clearly answers your research question using your data, and whether you connect your findings to the relevant scientific theory. A conclusion that only summarises your data without comparing it to theoretical predictions will not score well.
Common problems: conclusions that restate results without interpreting them, no comparison to accepted values or theoretical predictions, no discussion of whether the data supports or contradicts the hypothesis.
Evaluation (6 marks)
This assesses your ability to critically analyse your own methodology and suggest realistic, specific improvements. The examiner wants three weaknesses, with an explanation of how significant each one is and a concrete suggestion for how to address it.
Common problems: generic improvements (“use more precise equipment”), weaknesses that are listed without being connected to specific effects on the data, improvements that are impractical or vague.
How much help can you get from your teacher?
More than you might think. The IB is clear that teachers should provide guidance throughout the entire process, not just when reading the final draft.
Your teacher can and should give feedback on your topic choice, your methodology, and your first draft. They can provide that feedback verbally or in writing. What they cannot do is edit your work directly. The draft you submit to them for feedback must be followed by a final version that represents only your own revisions.
If your teacher is not proactively offering guidance, ask for it. Asking for help from your teacher is not penalised in any way. It is expected.
Where to go from here
If you are still deciding on a topic, 100 IB Physics IA ideas has a complete list organised by theme, with topics selected specifically for their ability to generate clean data.
Once you have your topic and you are ready to start writing, the exact IA structure guide takes you through all 13 sections with detail on what each one needs to contain and the most common mistakes to avoid.
The Exam Pack includes past paper questions and mark scheme practice for every topic in the syllabus, alongside revision note templates and a mock exam. It is designed to work alongside your IA, not instead of it.
Get the GradePod Exam Pack for £39 →
Written by Sally Weatherly, IB Physics teacher since 2004, Fellow of the Institute of Physics, and founder of GradePod. I help students work smarter, not harder.