A group to discuss, share and develop the new Higher Computing Science course.
Lee Murray.
DCullenThe SQA have updated the course support notes and added a specimen question paper
https://www.sqa.org.uk/sqa/56924.html
“11 June 2018
We have added course support notes to the revised course specification for Higher Computing Science. The course support notes provide further advice and guidance for teachers delivering the Higher Computing Science course from session 2018-19 onwards. This includes examples of the skills, knowledge and understanding developed in the course, and links to teaching materials and resources, that we previously published in the Resources to support the Higher Computing Science course document.
In appendices 10 and 11 of the course support notes, we have provided further examples of what candidates could be assessed on for database design and implementation.
We have also published the Higher Computing Science specimen question paper, which reflects the changes to the question paper that are being introduced from session 2018-19 onwards.”
Course Support notes added to the Course Spec – https://www.sqa.org.uk/files_ccc/HigherCourseSpecCompSci.pdf
Specimen Question Paper – https://www.sqa.org.uk/files_ccc/HigherSQPComputingScience.pdf
Lee MurrayI couldn’t reply to Greg’s post for some reason, so I’ll drop my concerns here. I have more concerns and haven’t even started on the ‘Support’ notes, this is JUST the specimen question paper:
2. Still using the term ‘iterative’ as opposite of agile (or other name for ‘waterfall’). Where does this come from? I’ve never come across it and can’t find it online. Agile is iterative. In fact MORE iterative than the waterfall method.
4. I don’t like the width of the second column. It looks like it’s designed purely to trick people. If this is how we assess computing knowledge, then I’m 100% against it. Make both columns the same width.
5. Don’t see how global variables make it any harder for two programmers than 1 programmer. – provided all programmers can see all the code (but why wouldn’t they be allowed to see it?) – or is this just another way of wording the same answer (accidental use of the same variable in different modules)?
6. The complexity of this problem is WAY higher than anything I’ve seen before at Higher level. Why is a course with more content than ever before being made more complex than ever before when pupils have had less exposure to computing than ever before? (possibly only 1 year at N5 level, but then again maybe not even that much)
Not to mention that the pseudocode isn’t the SQA reference language. Has that been scrapped? Or is just a case of “it’s used when it suits, it’s not when it doesn’t”?
6 b. Now pupils need to know the names of specific exceptions. Crazy. Not to mention that the exception in the marking scheme may not be the exception that actually appears. Different programming languages may handle it differently. For example, Python enters an infinite loop (if implemented as close to the design as possible).
7. Another question which will confuse people. To PERMANENTLY display a graphic at a certain dimension when there are other functions that change the size, you would need to alter the other functions too. Turns out the answer doesn’t want this at all and it seems the code provided is just an example. Why isn’t this made clear? Terribly worded question.
8. The wording is awful. So much second-guessing is involved and it’s basically so ambiguous that nobody will be certain if they have it correct or not. It sounds like it’s implying a 1 to many relationship between president and club, but clever pupils will know that just because it’s implied, doesn’t mean it is definitely the case. The question cannot be definitively answered because we don’t know how many clubs a president can be president of. The marking instructions show 1 to 1, but where is that mentioned in the question? Further to that, the marking scheme states “showing each president to one club”, but this doesn’t consider that a club could have two presidents, so what is the answer?
9. I didn’t notice that this was in the specification until now (conversion of two’s complement numbers from binary to denary and vice versa). This is even more stuff to add to the pile.
13. Don’t see how specifically a low fidelity prototype helps find navigation problems (specifically dead/orphan links – the only answer accepted). I automatically assumed it was a usability thing, especially since it talks about ‘navigation design’. Dead links is an implementation issue rather than design.
14 a. The compound key could also be tournamentID and position. It’s what I automatically thought of and I imagine most pupils would too.
14 b. The first table shows the ‘Table(s) and query’ cell with a value of ‘Result’, but this doesn’t exist in the example.
The answer includes “position = 1”, but this is deceptive. Position is a function that finds where a string appears in another string. There are many ways to return only the first result (LIMIT, ROWNUM, TOP etc.), but none of these is in the specification. Would they be accepted in an answer?
14 c ii. Didn’t expect it to be the heading (using an alias) that was the issue. Not a major problem, but felt like a trick question as ‘Error’ implies it’s the logic or the syntax, not the appearance.
14 d. The answer references ‘Result’ table again, but it still doesn’t exist.
14 e. I dislike the ‘relationship’ mark. Stating ‘member of’ is more of an English question than a Computing one. It doesn’t add anything to the design and it doesn’t help the developer to create the database. I worked as a database developer for years and not once did anyone include the ‘relationship’ name in their designs.
15 a. The wording of this question is very poor. I had no idea what the horizontal navigation bar design would look like. I can see from the answer what it was trying to ask, but without the marking instructions it’s just a terribly worded question. I’m guessing the poor wording is an attempt to make it more challenging, but I am against the deliberate obfuscation of tasks just to increase the challenge. If a clear question cannot be created to assess a certain aspect of the course, then should it even be an assessed element?
15 b. The question references efficiency through removal of repetition, but the (only) answer introduces repetition that didn’t exist before. Repeating footer twice and section twice just so you can group two selectors that have a padding of 5px is weird. It’s much better to repeat padding:5px once than it is to repeat footer, padding and the brackets with it. Not only is it better, but it just makes sense. Nobody in their right mind would create a separate grouped rule JUST for padding. Not only does it decrease efficiency because you’re loading more bytes of data (due to having more characters), but it makes it less maintainable by having the footer and the section have two rules each. This almost exposes the writer’s lack of understanding regarding efficiency and is confusing that term with ‘What I personally like code to look like’; something I’ve vehemently complained about since the introduction of the updated N5 course (yet not one person has ever addressed my concerns about this).
15 c i. Why would restricted choice not be a correct option? Users are still restricted to the options in the drop-down, even if they are allowed to select multiple options.
16 e. Would pupils be penalised for using code or something very similar to code? I know it says ‘recognised design technique’, but the accepted pseudocode is very loosely defined and is what the subject seems to have been moving away from in recent years (with the introduction of Haggis Reference Language).
What if the candidate was to use built in tools to reverse a string. Many languages have this feature and many pupils know their chosen language’s built in tools.
Also, why does it say ‘The function could be re-written to…”? The function would be effectively scrapped in order to fulfil the new functionality. Different function name, different number of parameters, completely different algorithm…
I think the mention of ‘re-written’ will add confusion to the question (a bad idea).
17 b. This form of ER diagram looks ridiculous. I’ve never seen this ‘bubble’ technique before in my life (with good reason). Why not just have a 1×2 table for each entity with the entity name in the top cell and the fields in the bottom cell – PKs underlined/bold, FKs with an asterisk – like what happens in reality? Also, the answers seem inconsistent. Driver.numberOfWins is not required, but why is Team.championshipWins? Surely both can be calculated.
18 a i. I’m assuming it has been considered, but there are many languages that don’t support Records. Many different schools/teachers will implement their workarounds in different ways. How can teachers know which ways will be accepted and which ways won’t?
18 a ii. Some programming languages don’t have arrays, or may create them in different ways, so may not need to declare a data type or an array size. I know that python answers don’t gain full marks, despite being perfect for the job:
Participants = []
How do you expect pupils who learned python to gain all marks? Which of the many possible varieties will be accepted as full mark answers?
18 b. The additional guidance mentions ‘Answer may use 1D array as candidate may not have answered the previous questions correctly.’. The previous question ASKS for a 1D array. Does it mean a 1D array of integers instead of a 1D array of records?
18 c. I have always understood the Computer Misuse Act offenses to be on a scale, meaning you could only commit one of the offences (i.e. if you actually modify data, you have committed that offense but not the others, since in order to commit the third offense, you MUST have committed the other two, but you cannot be charged 3 times for a single crime). If that is the case, then this question doesn’t make sense. Even if that’s not the case and that someone committing offense 3 is also charged with committing offense 1 and 2 too, it’s still a weird question because of the inclusion of the word ‘different’. This implies that the offenses are almost unrelated and this will definitely confuse candidates. It confused me until I saw the marking scheme.
19 a. Marking is focused on smartphone devices but there’s no indication that this is the only design (most websites would have a desktop version and a mobile version), so you can’t say this particular design is flawed because of that. Also, the content will adapt to fit the resolution of the device, as well as the orientation of it, so stating ‘too much information’ or ‘landscape’ design doesn’t really count. The ‘detailed description’, if it really did have a lot of information, would narrow and get taller, creating a portrait page – ideal for phones. Also remember that modern phones have 1080*1920 or greater resolution which isn’t far off the width of the most popular desktop resolution (1366 * 768). Phones also have zooming functionality, so this question is really pretty poor. ‘Flaws’ and ‘ineffective’ are probably the wrong words. Maybe ‘suboptimal’ is a better term.
19 c. The first two answers are perfect but ‘Fits the width of the section’ is a strange one. I doubt anyone would think to mention that given that, by default, all text will attempt to fit the width of the element it’s in (minus the padding).
20 a. The design seems deliberately vague. ‘Find and count names’ isn’t the same as ‘Find, store and count names’. ‘threeA()’ also seems deliberately meaningless to increase the challenge of the question. I really dislike that pupils are being asked to interpret the question setter’s intentions.
20 b. Same as 16 b, are pupils penalised for writing code? Some languages have features that help with this type of problem so designing these solutions may appear as expected. Pupils may be tempted to write code or something very similar to it.
a1an.robertsonWrt Q9, it appears that the ambiguity arises from the use of the terms “Describe and exemplify” in the spec. It appears that they should be interpreted as “Describe, exemplify and convert”. It is especially confusing that directly above the floating point part of the spec it states “Conversion of two’s complement numbers from binary to denary and vice versa.”, suggesting that there is only the requirement for students to convert 2’s complement numbers but not within the context of floating point.
Is there any good reason why the spec should be written in such an opaque manner, or is the interpretation by the spec paper setter incorrect?
Lee MurrayYeah, I stated two’s complement above when it’s a floating point number. I suppose it shouldn’t be in there at all. I still haven’t had a really thorough look over it, those questions were just the ones I could find issues with at first glance.
I’ll have a look at the specification and support notes too, but I already have issues with them. I’m especially insulted by the ‘Suggested learning activities’ section after a quick skim over it. I suspect I’ll be furious by the time I’ve had a chance to look properly.
You must be logged in to reply to this topic.