Here are my comments and thoughts on Bastien's IA.

When I first read Bastien's IA report it reminded me of years ago when I first started marking EEs. There were no criteria of guidelines to follow – we just had to give it a mark of – 1, 0, +1 or +2. I received an EE from a student from the UWC in America which had involved 45 hours of real computer time on the Los Alamos computer on an oscillating reaction. It was either complete rubbish, i.e. deserved – 1 or absolutely brilliant and worth +2. I felt I couldn’t make the decision on my own so I sent it to Ron Ragsdale, a professor in Utah, who was the Chief Examiner at the time. He wrote back that he couldn’t properly understand it either and suggested that we play safe and give it +2. Some weeks later he sent me a cutting from a newspaper. The student had won an award for being the most promising young scientist in America – his comment was “phew, luckily we guessed right!”

James knows his student well so is completely happy that it is all Bastien's own work. It is highly sophisticated piece of work and is in fact a master class in how to address all the IA criteria. In some ways addressing the criteria seems the rationale for the whole IA as it ticks every single one of the boxes in terms of what must be addressed to score full marks. To obtain that standard including the fluid way in which it is set out and expressed suggests to me that it would have taken a lot longer than 10 hours work – but that is not really an issue. So applying IA criteria both James and I feel it is worth 24/24 and indeed it was awarded a Grade 7 by the IB. Note that it was submitted and assessed under the old 2014 Guide criteria but I'm certain that it would still gain a Grade 7 if it were to be assessed under the new 2023 Guide criteria.

How about the underlying chemistry? Here is where I have a slight problem. As Bastien correctly points out, simulations are inherently flawed as they rely on an algorithm which is constructed by humans. Reaction kinetics by definition depends upon experimental data and order of reaction cannot be determined any other way – all that can be arrived at is a theoretical model. Human progress is made when experimental results are not predictable. The discovery of cis-platin could never have come from a model arrived at by ChemReaX neither would ChemReaX have predicted Fleming’s attributed discovery of penicillin. In other words scientific progress actually depends upon serendipity which is an anathema to computer generated models. The IA says you must show you how you actually controlled the controlled variables. There seems to be a weakness in the IA here (on page 3). The likely impact of the controlled initial pressure (which is set at 1 bar for every simulation) does not make sense to me as it seems to be related to temperature, not pressure. The rate constant is not actually a variable.  For a simulation I can see that you can put in different values for the rate constant and see what effect it has but this is artificial. By its very definition for a chemical reaction it is a constant, not a variable, provided the temperature remains constant. Perhaps the fact that the simulation allows you to alter it is a real weakness in the whole methodology?

However this is nit-picking.. It is a very well-constructed investigation and raises the real problem as to how the IB is coping with marking simulations as opposed to hands-on chemistry. Many examiners – me included! – find it hard to reach a consensus on how to mark these simulation IAs fairly and equitably so that all students are marked to the same standard. If they just follow the ‘ticky box’ IA criteria then it is easy – but does it result in the 'correct' mark and is it chemistry?