Engineering & Technology: Grading simulation

What you'll be able to do: Find out where the marks in an engineering answer actually are, which is rarely where students put their effort.

The mark distribution students don't believe

In most engineering questions, the arithmetic is worth surprisingly little. Marks are distributed roughly like this:

• Setting up correctly: the right free body, the right control volume, the right circuit simplification. Substantial, and it's where wrong answers originate.
• Stating assumptions: often explicitly allocated, frequently ignored.
• Method: using the right approach, visible in your working, earned even if the number is wrong.
• Execution: the algebra. Fewer marks than students expect.
• Sanity and units: carrying units, stating the answer with sensible precision, commenting on whether it's reasonable.
• Justification: why this choice, why this factor, why this standard.

Students spend their effort on execution, which is the cheapest band, and lose marks in every other category, mostly for things they knew and didn't write.

Method marks are the specific tragedy. You did the work in your head, wrote the number, and got zero for a correct approach because nobody could see it.

How to run it

Here's the mark scheme [paste] and my answer [paste]. Mark it as a strict examiner. Itemise: which marks I earned and where, which available marks I didn't earn and why. No benefit of the doubt.

Then the three engineering-specific questions:

Would I have earned the method marks? Is my approach visible in my working, or did I do steps in my head?

Where did I write something true and correct that earned nothing, and what was being asked for there that I didn't give?

Did I lose marks on units, precision, or failing to comment on whether the answer is reasonable?

That last category is nearly free marks and is forfeited constantly. An answer of 4.7 m for a beam deflection with no comment is worth less than the same number followed by "this is unreasonably large; check the units on the second moment of area", because the second demonstrates engineering judgement, which is what's being assessed.

Across the disciplines

Structural and civil. Assumption and load-case marks. Also: showing the load path (article 03) frequently earns marks students don't know exist.

Mechanical. Free-body diagram marks. The diagram is often worth more than the solution and is drawn hastily.

Electrical. Circuit simplification steps, and stating which analysis method and why.

Chemical. Control volume definition and stating the basis, per hour, per mole, per batch. Basis errors cost the whole question and stating it earns marks and prevents them.

Thermodynamics. Stating the system boundary and the process assumptions. Nearly always marked, nearly always skipped.

Fluids. Stating the flow regime and which correlation applies, with justification.

Design coursework. Justification and trade-off discussion carry the upper bands. A technically competent design with no rationale sits mid-band permanently, and students find this baffling until they see it itemised.

The marking of a design report

Worth separating, because the distribution is different again:

• Requirements and constraints identified, often heavily weighted.
• Options considered and rejected, with reasons. Students present one option, which forfeits the entire band.
• Justification of the chosen approach against the constraints.
• Analysis: usually less than students expect.
• Evaluation: limitations, what you'd do differently, what you're uncertain about.

A report that presents a single design confidently with no alternatives and no limitations will lose in two bands at once, and every year students are surprised.

Pitfalls

1. No real mark scheme. You'll get a plausible number with no basis.
2. Accepting generosity. Ask for strict; ask twice.
3. Taking the grade instead of the itemisation. The itemisation is the whole value.
4. Marking only finished work. Marking a plan is cheaper and changes more.
5. Trusting the simulated number in absolute terms. Use it for where marks go, not to predict your result.
6. The tell: your simulated marks are consistently above your real ones. Stop using the number and use only the itemisation.

Try this today

Take an engineering answer you've already had marked, and the real mark scheme. Have it marked strictly before you look at the real feedback.

Then ask only: would I have earned the method marks? For most students, on most papers, the answer is "not all of them, and I knew the method."