A week into the IA, everything feels possible.
You sketch a bold idea in your notebook, the kind that sounds like a real research project. Then reality shows up with a lab timetable, missing equipment, and a teacher who says, gently, “That’s… ambitious.” The best IA topics don’t win because they’re flashy. They win because you can actually finish them, repeat them, analyze them, and evaluate them like a scientist.
This guide helps you choose an IA experiment that’s both manageable and original -- the sweet spot examiners reward.
The IA experiment filter (quick checklist)
Before you fall in love with an idea, run it through this IA checklist:
-
Safety and ethics: Can you justify risks and follow school rules?
-
Resources: Can you access equipment/materials more than once?
-
Time: Can you collect usable data within 2--3 weeks?
-
Trials: Can you repeat it at least 3--5 times per condition?
-
Analysis: Will your results produce patterns, uncertainty, and evaluation points?
-
Personal angle: Is there a reason this question matters to you?
If you want to calibrate what “high-scoring” looks like across subjects, browse Comparing IA Expectations Across Different IB Subjects.
Start with curiosity, then shrink it into an IA-sized question
Most students start the IA backwards: they try to invent something “impressive,” then scramble to make it doable.
A calmer method is to start with something you already notice in daily life, then compress it into one clean relationship.
A simple prompt that generates strong IA ideas
Finish this sentence:
“I keep noticing ____ and I wonder if ____ changes when I change ____.”
Examples that often lead to solid IA experiments:
-
Sports and recovery (heart rate, reaction time, fatigue indicators)
-
Food and household chemistry (emulsions, corrosion, browning, conductivity)
-
Home environment (insulation, light wavelength, sound absorption)
-
Plants and growth conditions (pH, salinity, fertilizers, LED color)
Once you have a direction, use IB Internal Assessment Guides to sanity-check what your subject’s criteria actually reward in an IA.
Manageable beats complicated (and usually scores higher)
A top IA is not a PhD thesis. It’s a controlled, defendable investigation.
The most common reason students panic mid-IA is scope creep: too many variables, too much equipment dependency, or data that takes months.
What “manageable” looks like in a real IA timeline
A manageable IA usually has:
-
One independent variable with 4--6 levels (e.g., concentration, temperature, angle)
-
One dependent variable measured consistently, with units
-
3--5+ repeats per level
-
A method you can run again if trial one goes wrong
If you’re also trying to balance exam prep, pair your IA sessions with short content reinforcement using Questionbank so the coursework doesn’t swallow your revision.
Red flags your IA experiment is too big
If you recognize these, simplify now:
-
Requires rare sensors or specialized chemicals you can’t book twice
-
Depends on living systems that vary wildly (and you can’t control conditions)
-
Needs long-term data collection (weeks of growth with no backup plan)
-
Produces data that might be mostly qualitative (hard to analyze)
A strong IA feels like: “I can run this again tomorrow if I need to.”
Original doesn’t mean new science (it means your twist)
Students hear “original” and assume the IA must be never-before-seen. That’s not the goal.
Originality in an IA usually comes from one of these:
-
A local context (your school, climate, water source, materials)
-
A specific material choice (a certain brand, soil type, insulation)
-
A measurement upgrade (better controls, clearer quantification)
-
A comparative design (two conditions that reflect a real decision)
Here’s what that looks like:
-
Instead of “How does temperature affect enzyme activity?”
- Try: “How does temperature affect catalase activity in potato extract when substrate concentration is fixed at X%?”
-
Instead of “How does light affect plants?”
- Try: “How does blue LED vs red LED affect basil growth rate over 10 days, controlling photoperiod and soil moisture?”
When you want to see how top students phrase research questions, steal structure from IB Biology IA Examples or IB ESS IA Exemplars. Even if your subject differs, the focus and specificity translate.
Variables: the backbone of a high-scoring IA
Examiners can forgive a messy result. They don’t forgive a messy design.
For your IA, write variables like you’re writing instructions for a stranger:
-
Independent variable (IV): what you change (include units and levels)
-
Dependent variable (DV): what you measure (include units)
-
Controlled variables: what must stay constant (and how you keep it constant)
Example structure:
-
IV: caffeine concentration (0.00, 0.05, 0.10, 0.15, 0.20 mol dm⁻³)
-
DV: Daphnia heart rate (beats per minute)
-
Controls: temperature, light exposure, solution volume, acclimation time
If you want a step-by-step method for design, use How to Design Effective Experiments for a Science IA alongside 10 Expert Steps to Design a Science IA Experiment That Scores a 7.
Plan evaluation before you collect data
The smartest IA students think about evaluation while the experiment is still flexible.
Ask yourself:
-
Where will uncertainty come from (instrument precision, human reaction time, environmental variation)?
-
What will you do if the trend is weak (more trials, tighter controls, alternative measurement)?
-
What limitations are unavoidable (and how will you acknowledge them honestly)?
This is where RevisionDojo becomes a practical advantage. Use AI Chat to pressure-test your method, then use Grading tools to check whether your evaluation sounds like an examiner expects. If you need models, the IB Coursework Examples: IA, EE and TOK Exemplars library shows what strong evaluation actually looks like.
Common IA experiment mistakes (and the simple fixes)
Choosing a question that’s too broad
Fix: reduce to one IV and one DV. If you have “and” in your research question twice, it’s probably too wide for an IA.
Choosing something too simple to analyze
Fix: add more levels of the IV, improve measurement precision, or introduce uncertainty calculations and statistical comparisons.
Choosing something you can’t repeat
Fix: change materials to something accessible, or redesign so each trial is quick and independent.
Forgetting exam season exists
Fix: run a “minimum viable revision routine” during your IA weeks: 20 minutes on Flashcards plus a short Questionbank set after each lab session.
Closing: choose the IA you can finish twice
A good IA experiment has a quiet confidence to it. You can run it, repeat it, explain it, and critique it without inventing excuses.
So choose the investigation that fits your life, your lab access, and your time. Make it original with a personal twist, not a fragile design. Then build the rest of your IA around clear variables and evaluation you can defend.
When you’re ready to lock in your idea, use RevisionDojo as your control center: browse the IB Coursework Examples: IA, EE and TOK Exemplars, check expectations in the IB Internal Assessment Guides, and keep exam performance steady with Flashcards and Questionbank. Your IA becomes manageable the moment your plan stops trying to impress and starts trying to work.
