16. Craig'n'Dave 'Unscripted' - Writing code in written exams

In this video we discuss some of the difficulties that come with marking written code questions in exam papers. We provide our own thoughts on the process of when to award a mark, and while we don't always agree, we both agree in a "positive marking" approach!

02:04 Introducing the question
03:45 Craig's advice as a past examiner
07:28 Syntax errors and alternative approaches are acceptable
09:00 The OCR reference language
14:38 General comments about the algorithm produced as an answer
18:25 Initialising variables
21:00 Iterations: while or for
25:00 Comparison operators: ≤
29:26 Array out of bounds
33:45 Outputs and missing parenthesis
34:30 Input
34:45 Storing a value
36:21 Conditions
37:52 Incrementing variables: x = x + 1, x++, x+, x=+1, x+=1
41:35 Back to comparison operators: = or ==
42:42 Output
45:37 7/8 marks
48:10 Exam marking isn't as accurate as it should be

[Arganoid]

If the language is 1-indexed it's still an error, because the first iteration will try to access student zero. So according to the letter of the mark scheme it's wrong as it's iterating over 21 students, not 20. But how actual markers would mark it, I don't know

[gloshisgreen]

re: ligatures vs <= etc in code - mono-space fonts with ligatures design for coding have been around for a while now.
Fira Code, Hasklig, Jetbrains Mono, Iosevka, VIctor Mono etc... to name a few.
It's supposed to make it easier to scan through code quickly.
When I've use them in the past not all my tools supported a custom font and the cognitive overhead of switching between fonts was jarring so I stopped using them. There's wider support for them now though so I may go back and give them another go.