National 5 Computing Science course. A group to discuss, share and develop the new National 5 Computing Science course.
DCullen.
Peter ThoresenSpecimen paper for revised N5 course is now on SQA website.
http://www.sqa.org.uk/files_ccc/ComputingScienceSQPRN5.pdf
Peter ThoresenOverall I am ok with the paper, I just have to teach some things differently from before.
Some questions raise issues:
Q2 straight from Intermediate 2 Computing course
Q3 straightforward, but why are we teaching DPA instead of GDPR
Q7 still not sure exactly what “low-fidelity prototyping” is. Needs a clear definition from the SQA.
Q8b good that answers outwith course specification still get credit (xradius, yradius, rotation, layer, …)
Q14b marking scheme does not require initialisation of running total variable
Q15a marking scheme requires that wireframe blocks for media must specify file type. For images I just draw a box with a cross! We need a clear specification of what an SQA wireframe must contain.
Q17a E-R diagrams are now completely different from before (two labelled boxes with a crows-foot line).
Q17ci Use of language “evaluate”. Pupils may just write actual effect of the statement, and not compare intention.
Q17cii Why “describe” instead of writing an SQL command?
Q18 The browser image show body in a box. The code does not do this – text will stretch full width of browser window.
Q18dii Although “font-style:italic;” is easy to understand, it is not in the course specification.
Marking scheme is odd – there are 4 effects for three marks, but pupil needs both “italic” and “left” for one of them.
Q18e Technical accuracy – bit map images pixels do not store colour of each pixel. They store a number representing the colour (remember CLUTs?)
Q19c Answers are obvious when you see them, but this one confused me.
Q19d The course specification required pupils to “explain the need for interpreters and compilers to translate high-level program code to binary”. This question requires pupils to compare the two.
Also, when can we stop teaching pupils that you use an interpreter to develop a program, and then compile it. Change the question to “describe one advantage of using an interpreted language and one advantage of using a compiled language”.
Q20c Pupils should be writing SQL code. However the equijoin might be a problem in this case.
Ronnie RossThe ERD thing threw me, I taught these the way you described Peter, does this mean we need to change things, or will the other way still be ok? There are different ways to do this, and this has me slightly worried that I might be doing it the “wrong” way, I don’t want pupils to get penalised.
Generally, it doesn’t seem bad, I just feel that some further clarification on some course content would be welcome, as I think there is the potential for some slip ups on one or two areas (though mostly it looks ok).
Enrico VanniQ.10 – the answers in the marking scheme are just plain wrong. As predicted by Richard Scott in the N5 Course Spec thread the vague statement “Compare…audio standard file formats WAV and MP3 in terms of compression, quality, and file size” was a problem waiting to happen. We didn’t have to wait very long.
A high bit-rate encoded MP3 could potentially have a greater file size and will almost certainly have better sound quality than a low bitrate audio file of unspecified format – potentially also MP3 – encapsulated in a WAV wrapper file.
An indeterminate format WAV file re-encoded into MP3 format at a suitably high bit-rate will almost certainly not suffer a reduction in sound quality (the subjective arguments surrounding the sound quality implications of MP3’s lossy algorithms are for audiophiles to have – it is not the place of Computing Science courses to take sides in the debate).
The assumption being made by the question setter (!!) is of course that the WAV file contains LPCM encoded lossless audio, but even then the lack of detail on sample rate and size of the encoding of the files being compared means the answers proffered cannot be taken as absolute.
I anticipated and warned of these sorts of vagaries cropping up and made the point directly to Raymond at the CASS consultation meeting in February. Excising contextually important content (in this case the aforementioned sample rate and size) would lead to these orphaned and therefor inaccurate teaching points. Pardon my integrity, but I am entirely unhappy about teaching my students to write garbage in the exam in order to comply with an inadequate course spec.
I will of course be shunned for saying so, with the usual suspects once again saying we should applaud the effort and turn a blind eye to the lack of efficacy of the people involved in this.
#vindicated
Darren BrownIn my opinion the questions are better overall – simpler wording, clearer and better scenarios.
I still can’t get my head around 2hrs length, 110 marks and I’ve counted pupils needing knowledge of 5 different coding languages… an already difficult subject is being made even more inaccessible. Any form of prelim analysis tells me the majority of pupils struggle with coding style questions the most – they have not had enough experience and most will not do N5.
Even if pupils had done full BGE I don’t believe they could be expected to have much more than an understanding of a text SDD Language and HTML/CSS. A fair bit of this paper feels like Higher to me – I’m going to have trouble learning all this for this year and I’m worried about where Higher is going to have to go.
Obviously people feel we need to raise the bar and it will force all schools to do more Computing Science but in reality staff/pupils keep leaving the subject year on year and our subject will reduce further.
Tracy RennieI totally agree. Having taught Computing for over 25 years we have gone from a broad subject where pupils could expect to learn a lot about different aspects of computing and therefore follow several different types of careers to a highly technical subject which the majority of pupils are not interested in or will not need in future careers. I am aware that the subjects needs updating regularly and enjoy this aspect of the subject but I feel that we have shot ourselves in the foot with the direction Computing Science has taken and can only see pupil numbers decreasing.
Enrico Vanni
” I’ve counted pupils needing knowledge of 5 different coding languages… an already difficult subject is being made even more inaccessible. Any form of prelim analysis tells me the majority of pupils struggle with coding style questions the most – they have not had enough experience and most will not do N5.”
It’s the big elephant in the room that the people steering this change simply do not want to acknowledge. Learned wisdom from way back when I completed my Computing Science degree said that you should have a firm grounding in one programming language (and that means months or even years of experience) before attempting to learn another because you are in fact learning three skill sets with that first language (constructs, syntax, algorithmic) , two of which are then transferrable to other languages but must be grasped first. Nothing about that has changed in the intervening years.
The answers given in this link (https://www.reddit.com/r/learnprogramming/comments/2r7g79/is_it_a_bad_idea_to_learn_multiple_programming/) sum things up very well. The approach to learning coding being expected in this new N5 course is totally counter-intuitive and demonstrates a disjoint between the understanding of the people driving these changes and the reality of the process. Again I question their efficacy.
The response that will once again be trotted out to this question though is that the BGE/Benchmarks will prepare pupils and ground them in coding, but that is a convenient fiction as it has been pointed out numerous times already – Benchmarks are not a prescriptive mandatory list of content to be covered by every student by the end of S3, and with head-teachers having more autonomy and staffing resources being squeezed to the absolute limits their priorities will be elsewhere and Computing Science considerations, if they start to pose a challenge, will simply be sidelined.
These arguments have already been presented though, and the fingers are now securely in the ears of the people who hopefully now realise what they have presided over. I’m putting a marker down now that the N5 Course Specification will change again within the next 3-4 years, probably sooner, when the issues caused by this ill-considered rush job (the fault of John Swinney and us teachers, allegedly) come to bear.
The practical outcome of this in the interim is that we will see tri-level teaching of three disparate computing science courses in S4 – National 4, Old National 5 Units and New National 5 in order to ensure that every student achieves an outcome. If this was the real intention of the SQA (and yes I know the SQA are real people but it seems they have made a choice to become faceless and unaccountable as far as this forum goes) all along then applaud their guile, but as the old saying goes – never put down to conspiracy what was more likely caused by cock-up.
DCullenIs there still time for the SQA to clarify or omit any ambiguity in the course specification, specimen paper and specimen coursework or even fix mistakes? Probably not.
For example the wav vs mp3 debate. I like Enrico think we are doing our pupils a disservice if we just teach them that wav is better quality when most people cannot tell the difference between a wav file and a 320Kbps mp3. It is a simplification to suit exam questions, like in old Higher where the exam paper was looking for a difference between ascii and unicode and pupils where taught to write ascii is 8 bit and unicode is 16 bit, despite the fact there are different versions of unicode, including 32 bit.
Another point in the course specification:
Describe extended ASCII code (8-bit) used to represent characters.
That brought about a question in the specimen paper asking the pupil to name a control character. I have always told pupils that the first 32 characters are control characters. Why a pupil would want/need to memorise and name one specific control character is beyond me. I do not recall that ever being done before. Now I know I will have to teach my pupils to memorise some control characters as it might come up in the exam. It’s back to finding out what you have to teach the pupils from the exam answers rather than getting a clue from the course specification.
Th ERDs look quite different from what has gone before and are a bit complex going by the paper example. If pupils create ERDs in class do they have to conform to this style? I think putting the entity and attributes in the one box looks much neater but that’s just a personal preference.
You must be logged in to reply to this topic.