IB Sciences: Simulation and the Scientific Investigation

Among the major science qualifications, the IB Diploma is unusually explicit about simulation. The Scientific Investigation, the single internal assessment that carries 20% of the subject grade across Biology, Chemistry and Physics, allows students to build their investigation on hands-on laboratory work, on databases, on modelling or simulation, or on a hybrid of these (IBO).

That is a materially different position from the one most UK specifications take, and it is worth understanding precisely, because it is frequently either overlooked by schools with laboratories or over-read by schools without them.

What Exactly Does the IB Permit?

The Scientific Investigation is a single individual investigation, allocated ten hours, reported in no more than 3,000 words, marked out of 24, internally assessed by the teacher and externally moderated. The methodology is the student's choice, and simulation-based and database-based investigations are assessed against the same criteria as wet-lab ones.

The important word is same. A simulation route is not an easier route, and it is not a concession for schools that lack facilities. The investigation still has to show a focused research question, a methodology the student justified, data handled properly with uncertainties addressed, and analysis that engages with the limitations of the approach taken. A student who chooses simulation inherits a specific set of limitations to discuss, and examiners can tell whether they have thought about them.

Where Does a Simulation-Based IA Genuinely Excel?

Several categories of investigation are better done in simulation than at a school bench, and not for reasons of convenience.

Investigations requiring large sample counts. A student testing how reaction rate varies across a range of concentrations, with repeats sufficient for meaningful error bars, is limited at a real bench by reagent, time and patience. Ten hours of laboratory access rarely produces a data set with enough points to support the analysis the criteria reward.

Investigations across a wide parameter range. Physics investigations in particular suffer from apparatus that works well across one decade of a variable and badly outside it. A simulation lets a student explore where a relationship holds and where it breaks down, which is a considerably more interesting investigation than confirming it in the region where it was always going to work.

Investigations that are hazardous, slow, or need controlled substances. A school cannot hand a sixteen-year-old the reagents or the timescales that some genuinely interesting questions require. Simulation removes the constraint without removing the science.

Investigations where the variable of interest is the method itself. Comparing techniques, or quantifying how a specific procedural error propagates through to a result, is difficult to do at a bench because you need to make the error deliberately and repeatably. In a physics-driven simulation that is straightforward, and it produces exactly the kind of methodological reflection the top mark bands ask for.

Where Does Simulation Still Fall Short?

Two places, and schools should be clear-eyed about both.

The first is manipulative skill. The IB sciences course as a whole expects students to develop practical technique, and the Diploma is a preparation for university science where that technique is assumed. An IA is one piece of work within a two-year course; a student can produce an excellent simulation-based investigation while still needing real bench hours across the rest of the programme. Using simulation for the IA is a legitimate methodological choice, not a reason to stop doing practical work.

The second is the quality of the simulation itself. An investigation built on a platform that plays scripted animations has a fatal problem at the analysis stage: there is nothing genuine to analyse. If the outcome was authored rather than calculated, there is no measurement uncertainty to characterise, no systematic error to identify, no noise floor, and no honest discussion of the model's limitations to be had. A student can only interrogate a simulation that is actually simulating something.

This is the practical test to apply before letting a class base IAs on any platform: does the simulation produce data with realistic variation, and can the student explain where that variation comes from? If every run returns the same clean number, it is unsuitable, whatever else it does well.

What About the Collaborative Element?

The IB sciences course includes collaborative project work across the science subjects, designed around a shared problem approached from different disciplinary angles. For schools whose students are geographically distributed, such as international schools with multiple campuses or online schools, this is logistically the hardest part of the Group 4 experience to deliver, because the whole point is students working together on something.

A shared simulated environment handles this well, since students in different countries can work on the same apparatus, on the same problem, with the same equipment available to all of them, which is something no distributed set of physical laboratories can offer.

How Should an IB School Actually Use This?

Our recommendation, for what it is worth from a vendor with an obvious interest:

• Do not present simulation to students as the easy option. It has its own demands, and students who choose it expecting less work produce weak IAs.
• Use it to widen the question space. The best argument for simulation in the IA is that it makes genuinely more interesting research questions viable, not that it saves the school money.
• Teach the limitations explicitly. A student who can articulate what their model does not capture is demonstrating exactly the understanding the criteria reward.
• Keep the bench. Where you have laboratories, use simulation to prepare students for them rather than to replace them. The flipped practical model applies as well to the IB as to anything else.

Where WhimsyLabs Fits

WhimsyLabs runs on a physics engine, which is the property that matters for the IA. Reagents carry realistic concentrations and impurities, measurements have genuine uncertainty, and outcomes are calculated rather than authored, so a student investigating a relationship gets scatter they have to account for, and a student who introduces a procedural error sees it propagate. That gives an IB student something to analyse and something honest to write in their evaluation.

It also runs in a browser on ordinary hardware, which matters for international and online IB cohorts working across time zones on their own devices. And because our assessment is based on what a student did rather than what they wrote, teachers moderating a set of investigations have a record of the actual process behind each one.

The IB has made a considered judgement that a well-constructed simulation-based investigation can demonstrate scientific thinking as well as a wet-lab one. It is worth taking them up on it, carefully, and with a platform that can bear the weight.

Related Articles

• Virtual Labs for Online Schools: Science at a Distance
• The Flipped Practical: Arriving at the Bench Already Competent
• Why Other Virtual Labs Fail: The Physics Engine Solution
• Grading the Process, Not the Answer