The infamous SCHMACKOS award for questions that are a complete dog's breakfast.

The Questions That Make You Sick As A Dog Award
Questions that are wrong in a way that does not jeopardise the answer, or are just dumb.

The Stamp of Approval for questions I like.

The exciting Illiteracy Award for crimes against the English language.

New for 2016 - the Naughty Step Award, for when the examiners breaks their own rules! and are sent to think about what they have done.

Written examination

Thursday 10 November 2016

Reading time: 11.45 pm to 12:00 noon (15 minutes)

Writing time: 12:00 noon to 2:00 pm (2 hours)

• Students are permitted to bring into the examination room: pens, pencils, highlighters, erasers, sharpeners, rulers and one scientific calculator.
• Students are NOT permitted to bring into the examination room: blank sheets of paper and/or white out liquid/tape.

Materials supplied

• Question and answer book of 21 pages
• Detachable insert containing a case study for Section C in the centrefold.
• Answer sheet for multiple-choice questions.

Instructions

• Remove the insert containing the case study during reading time.
• Write your student number in the space provided above on this page.
• Check that your name and student number as printed on your answer sheet for multiple-choice questions are correct, and sign your name in the space provided to verify this.
• All written responses must be in English.

At the end of the examination

• Place the answer sheet for multiple-choice questions inside the front cover of this book.
• You may keep the detached insert.

Students are NOT permitted to bring mobile phones and/or any other unauthorised electronic devices into the examination room.

Examiners' report general comments

SECTION A - Multiple-choice questions

Instructions for Section A

Answer all questions in pencil on the answer sheet provided for multiple-choice questions.
Choose the response that is correct or that best answers the question.
A correct answer scores 1, an incorrect answer scores 0.
Marks will not be deducted for incorrect answers.
No marks will be given if more than one answer is completed for any question

Question 1

Which specific data type would be the most appropriate to represent a street number (for example, ‘5A’)?

1. string
2. numeric
3. Boolean
4. decimal

Answer is A.

The sample data is alphanumeric - containing both alphabetical data and numbers. It can only be stored as text/string.

This bit will be filled in when the examiner's report is released - in half a year or so.

Question 2

In the process of sorting an array of eight integers using the quick sort algorithm, the first partitioning with the array appears as follows.

Which one of the following statements is correct?

1. Neither 17 nor 19 was the pivot.
2. The pivot was either 17 or 19.
3. The pivot was 17, but was not 19.
4. The pivot was not 17, but it could have been 19.

Answer is B. I'll stake 5 cents on it. I'm hoping...

For most students, I suspect this was a brain-destroying question to start off section A.

I instinctively started trying to reverse-engineer the starting state of the array, before realising that it was like trying to unscramble an egg. It took a while to go back to first principles: quicksort divides an array into 2 subarrays with a point somewhere in it (the pivot) that separates the values smaller than the pivot value from all values greater than the pivot value.

This meant that having to carry out a quick sort was not necessary. It involved more basic principles inherent to quick sort.

After a partition, there will be a value (the pivot) that will only have smaller values to its left, and only greater (or equal) values to its right.

Taking each datum in turn and applying this rule...

• It can't be 12 because 11 is to its right.
• Same for 15.
• Can't be 11 because a larger value is to its left (12 and 15).
• Could be 17, because it has only smaller values leftwards and larger values rightwards.
• Also could be 19 for the same reason.
• Can't be 22 because a smaller value (21) is on its right.
• Can't be 20 because larger values are to its left.

I'm guessing 98% of you just guessed an answer. It took me an hour (and three Youtube tutorials) to finally decide. This is a luxury you do not have in exam conditions. The lesson here (apart from pivot stuff) is that if you get a multichoice question that takes more than a couple of minutes to be sure about - SKIP IT and come back if you have time. Do not spend 15 minutes stubbornly trying to earn one single mark - especially when there's a 25% chance you'll get the mark by sheer luck anyway.

Dog almighty, this was a brutal way to start off a nice, friendly exam! It was not unfair or unclear - in fact I tip my hat to its clever questioning and focus on core theoretical principles - but by golly there will be a few sad and confused puppies in the cohort after this question.

I inaugurate a new award - the 'Scare The Bejeezus Out Of Students Award'.

Use the following information to answer Questions 3 and 4.

The following algorithm has been written.

First, let me be the first to say BAD EXAMINER and bop you on the nose with a rolled up magazine.

You've broken your own pseudocode conventions!

Assignment in SD exams is marked by the symbol, not "=" which is reserved for equality.

Then, you forgot to put the output keyword in bold with a capital 'O'. One would expect better.

Now, go and sit on the naughty step.

No. Don't be a sook. Naughty step. Now!

Question 3

The algorithm needs to be tested for logic errors.
Which of the following sets of test data would provide the best test for the algorithm?

1. 0,75,80,84,100
2. 75, 76, 84, 85
3. 70, 75, 76, 80
4. 0, 74, 84, 85

Answer is B.

Another challenging question. Consider the options. None of them leap out as immediately ridiculous, so there are no quick and easy cross-outs to be made.

We need to thoroughly check ALL possible valid data inputs. We need to focus on boundary conditions where the logic of the code may fail.

These valid inputs are not defined, but we can assume 0 to 100.

The 3 valid ranges and their boundaries are (paying close attention to the >= and > signs)

1. 0 to 75 (C grade).
2. 76 to 84 (B grade)
3. 85 and above (A grade)

• Option A tests ranges 1 (0,75), 2 (80,84), 3 (100).
• Option B tests ranges 1 (75), 2 (76,84), 3 (85)
• Option C tests ranges 1 (70,75), 2 (76,80). It does not test range 3. Cross this option out.
• Option D tests ranges 1 (0,74), 2 (84), 3 (85)

Options A, B and D all test the 3 possible ranges of valid inputs. (A) has more tests and does not focus on boundary values as well as (B). Option (D) also lacks boundary-specific tests. So of the 3 complete sets, B would be the most likely to uncover boundary condition faults in the code's logic.

Another thought-provoking question. These multichoice questions are taking some time to ponder!

Question 4

The algorithm contains

1. procedures and functions.
2. procedures and instructions.
3. control structures and functions.
4. selection statements and instructions.

Answer is D.

It does not have functions. Rule out (A) and (C).

It does not contain multiple 'procedures' (B) even thought the code itself would constitute a procedure.

That leaves (D) - but check it anyway. It has selection statements ('IF' etc to choose which code to execute) and instructions (like 'output grade')

Question 5

Which data structure would be best to store the data being collected from the form above?

1. four one-dimensional arrays
2. a single array
3. a record
4. a list

Answer is C.

4 x 1D arrays would do the job, but they would be rather more difficult to manage.

Records allow a collection of fields with different data types (e.g. text, Boolean) in a single unit.

The key word in the question is "best".

Question 6

When a student upgrades their operating system to a new version of the operating system, the updating software creates a new folder called ‘OperatingSystem.old’ and copies some of the computer’s files to that folder.
This is an example of

1. archiving the computer’s files.
2. backing up the computer’s files.
3. cleaning up the computer’s files.
4. creating extra space for the computer’s files.

According to the wording of the question, the answer would be B - but the question is badly written and common sense should lead to the answer of A.

Hmmm.

An archive moves a copy of data into offline storage. It deletes the original data.

A backup copies (and keeps) the original data. It is intended to allow recovery of lost data.

The question says the old OS files are 'copied' to the new folder. That means the original files remain in place and are not deleted. That is a backup.

The question would need to have said "moves some of the computer's files" to qualify as archiving.

The problem is, in real operating system upgrades, "old" folders contain OS data that has been copied, then deleted and replaced by newer versions - which is an archive.

The question is flawed. If the original files are only copied and the originals are left in place, it's a backup. If the original files are moved and the originals are deleted, it's an archive.

The question says that files are copied, not moved. It must be a backup - but that does not happen in The Real World™. Why copy files to a new folder when the originals are left in place? It makes no sense for an operating system upgrade to retain old, incompatible versions of files in its active system software store.

Question 7

The following array is sorted in ascending order.

Starting point: 13,6, 64, 25, 12, 22, 11, 17

After the third pass: 6, 11, 12, 25, 64, 22, 13, 17

What type of algorithm is used to sort the array?

1. quick sort
2. bubble sort
3. binary search
4. selection sort

Answer is D.

Sorting - again.

Well it obviously can't be C, which is a search, not a sort. That leaves 3 little piggies.

If it were a bubble sort (B), the first few pairs would have been swapped, leading to 6, 13, 25 - which isn't right. Rule that out too.

I'm assuming a good exam would not re-examine quick sort again so quickly, so I consider it wiser to next investigate selection sort rather than binary search as next contender...

Work it through a selection sort walkthrough...

Yep. After the third pass, it matches the question. Done and done.

See the slideshow on selection sort to see how it works.

Another challenging and time-consuming multichoice question. This exam is not easy.

Question 8

If software is purchased over the internet, copies of that software can be sold provided that

1. the software is open source.
2. the software licence allows for this action.
3. slight changes are made to the software before selling it.
4. it explicitly states that the software is copyright protected.

Answer is B.

If you thought of 'open source', remember that such a licence lets you use and (maybe) redistribute the software, but not to sell it.

Use the following information to answer Questions 9 and 10.
Consider the following piece of pseudocode.

Question 9

The purpose of the pseudocode is to

1. find the sum of all numbers in an array that are higher than 10.
2. search for numbers in an array that are equal to or higher than 10.
3. find the sum of all numbers in an array that are equal to or lower than 10.
4. find the sum of all numbers in an array that are equal to or higher than 10.

Answer is D.

Once again, SD exercises the famous ">=" gotcha. Whenever you see <= or >= your shields should be up.

Again, VCAA mucks up their pseudocode: this time with with irregular indentation in the 'Display' line.

At least they got the assignment arrows right this time.

One thing I find concerning is the use of the len( ) function. Pseudocode is supposed to be completely intuitive and readable and self-explanatory. Any invented functions should either be blatantly-obvious or explained in a comment. The len( ) function is an invention that - supposedly - returns the number of elements in the named array. So if nums[ ] had 5 elements, len(nums) would return 5. But that is not immediately and blindingly obvious to students whose programming language did not use that name for the function, or used a different name for it. From my personal experience, len( ) is a function that returns the length of a string.

Let's make a deal, VCAA - you may invent any useful function you like for exams, but they must be either crystal clear or self-explanatory in meaning (e.g. 'Display') or explained in a comment, such as...

// len( ) returns the number of elements in an array

OK?

Question 10

When converting the pseudocode into a program, it would be useful to include in the programming code the

1. design specifications.
2. constraints of the code.
3. internal documentation.
4. external documentation.

Answer is B. Probably.

There is no internal documentation in the pseudocode so it can't be (C).

There is no external documentation either, and even if there were, it wouldn't be put in the code - 'cos it's external. So it's not (D)

I can't see how the design specifications (A) would be useful in the source code.

That only leaves the constraints of the code, which may be useful to later programmers who need to be aware of the limits within which the code needs to operate. For example

// Remember that nums[ ] must support floating point values!

The answer had better not be C. Sure, pseudocode is highly recommended in source code, but if there is none given in the pseudocode, how can it be put into the source code?

Notice that the question says "it would be useful to include in the programming code the" - not just any design specs, constraints, internal or external doc but those that have already been provided in the pseudocode.

If the question turns out to mean, "When converting the pseudocode into a program, it would be useful to include in the programming code some internal documentation." I will be very, VERY annoyed - in both capitals and bold face!

Use the following information to answer Questions 11 and 12.
A photographer is viewing high-quality photographs on a computer. The photographs are stored as JPEG files on a central fileserver and have an average size of 2 MB each.

Question 11

How long will it take the photographer to download one photograph to a computer that has a wireless connection running at 10 megabits per second?

1. 0.8 s
2. 1.0 s
3. 1.2 s
4. 1.6 s

Answer is D.

Rejoice, for we have reached the single question for which you brought your calculator. Let's work it through.

2 megabytes = 2,000,000 bytes (or 2,097,152 if you're a pedantic 1,024 bytes to a kilobyte person. The answer does not depend on which one you choose.)

Multiply by 8 to get 16,000,000 bits.

Divide by 10,000,000 (10 megabits)

Answer is 1.6 seconds.

I found it odd how the examiners suddenly decided to get casual and abbreviate "seconds" to "s" in the options, without explanation and for no apparent reason. Would this be obvious to ESL students?

Question 12

The photographer is considering using a wired network.
The advantages of a wired network compared to a wireless network are that it is

1. faster and more secure.
2. faster and more portable.
3. more secure and portable.
4. easier to set up and cheaper.

Answer is A.

This is a pretty easy one. Cables are nearly always faster than wireless, and - unlike wireless - cannot accidentally carry data beyond the walls of the organisation. Cables are, however, far from portable or easy to set up.

Question 13

A standalone application is being written for handheld devices. This application will store all its data on the device and does not need an internet connection.
Which type of application architecture would be best to use?

1. mobile
2. rich client
3. thin client
4. peer-to-peer

Answer is A.

I expect there will be arguments over this one, and most kids would have been scratching their heads during the exam.

The stem says that an internet connection is not needed, but is some other communications (e.g. wifi or Bluetooth) required?

The app is said to be "standalone" which means it needs no connection to remote data or code sources. So let's assume no comms are needed at all. Think of a self-contained game like Solitaire on a phone. So, which architecture would suit?

• Rich client means that the host computer has all the processing power and ability required with no support needed from external sources, such as cloud processors. The average computer is rich client since it does all of its own processing.
• Thin client is the scheme whereby the host computer has minimal processing facilities, and relies on an external source to supply the processing grunt. The host computer acts more like a dumb screen and keyboard to enter data and view the information that has been created by another computer. Think of cheap, low-powered Chromebook computers that use cloud processing. Without the external processors, a thin client terminal is little more than a paperweight.
• Peer-to-peer means that in a network of devices, each one has equal rights to initiate transactions (e.g. sharing files locally on a small LAN, or sharing files globally via Bittorrent). I can't see how it's relevant to this situation involving a standalone app.

That leaves A, which fits because the app is meant for handheld devices.

Question 14

XML stands for extensible mark-up language.
Which one of the following is true?

1. An XML data file is designed to only display data on a web page.
2. An XML data file will work as expected if some data is modified.
3. An XML data file will never work as expected if some data is modified.
4. An XML data file is interchangeable with an HTML file because they are both mark-up languages.

Answer is B. I reckon.

XML is new theory, and many people will not be very familiar with it.

The answer will not be (A). XML is used in many places other than web pages: it is a flexible all-purpose, plain-text, human-readable way to store and communicate self-contained, structured data files.

I have done a lot a work with XML because the database I made for teachers to create semester reports had to generate XML to feed into the education department's system. It was certainly possible to modify data within an XML file, considering it is plain text. Mind you, XML will break if you muck around with the tags in the file, but not the data within the tags. So (B) is true.

And therefore (C) is not true.

It is true that XML and HTML both use markup tags to provide structure to data content, but that's where the similarity ends. Otherwise you'd have to say that worms and garden hoses are basically the same because they're both long and thin with a hole in the middle.

Question 15

Inta is writing a new application for a small bicycle company so its staff can use their mobile phones to access the stock information stored on the company’s fileserver. In the company, there is a manager and four people who sell the bicycles. They all use different brands of mobile phones.
To help her with writing the software requirements specifications (SRS), Inta should

1. interview all those who will use the new application.
2. send out a survey to the manager and one other staff member.
3. observe one of those who will use the new application at work.
4. survey some of the people who have bought bicycles from the company.

Answer is A.

Hmmm. Interesting. What key knowledge is this question fishing for? What exactly is being examined?

Is it testing understanding of who should be consulted?

Or is it testing understanding of how they should be consulted?

Or could it be testing both?

I'm thinking of SD U3O2 KK 1 in the study design - "techniques for collecting data to determine needs and requirements, including interviews, surveys and observation".

The question smells very much like a fishing expedition for knowledge of KK1 - data collection techniques.

This is another arguable one. One could say that interviewing all users (A) would be over-the-top and that just one user and the manager (B) would suffice. But there is that rather odd comment in the case study - "They all use different brands of mobile phones."

"Why is that there?" you should ask, for stuff is not put into questions as filler or fun. It must be relevant to guiding your decision.

So - how could it be relevant? What options might it be intended to rule out?

If all users have different types of phones, they would all have different experiences and needs from the app. One person, for example, might have a small screen and require the app to be designed with minimal output on each page. Another might have a phone with a stylus, which would enable hand-written input. Another might have a phone with a different constraint that would rule out an important design idea. That comment in the case study makes me lean towards option (A).

But looking at (B), one realises that the manager would be an important person to consult to determine the organisation's functional requirements for the app. But will the manager be using the app - and therefore be included in option A, and be interviewed?

Sadly the wording of the case study is ambiguous - a common fault in VCAA IT exams. We are told that "In the company, there is a manager and four people who sell the bicycles."

This could mean...

1. In the company, there is a manager and four people, who (all five of them) sell the bicycles.

or

2. In the company, there is a manager --- and four people who sell the bicycles.

Try this. What does this mean...?

A and B doing C

Does it mean...

(A and B) doing C

or

A and (B doing C)

VCAA really needs a punctuation and grammar manager.
It may alleviate the frequent ambiguities in VCAA study designs and exams.

As a former English teacher, these things leap out at me. And that's why VCAA exams often annoy me more than they irritate other people.

But, it's section A and we can't explain our reasoning or give alternative answers to cope with ambiguities in the questions. We have to decide. Here I look at option B's data collection technique - survey - and wonder. Would a survey be sufficient?
Surveys are composed of closed-answer questions with limited opportunities to provide answers. Surveys are designed for mass data collection to be undertaken after detailed initial fact-finding (using interviews and observation) has uncovered what is relevant to the topic in question. No, surveys are not suitable here.

Remember - knowledge of interviews, surveys and observation are required by SD U3O2 KK 1.

Let's down-vote (B).

Option (C) is fair enough - observing users can uncover valuable ideas that may contribute to the formulation of functional or non-function requirements (e.g. sales staff may be very busy so the app would need to work more quickly than usual.) But only observing only one user would probably not be enough.

Option (D) is feasible and might yield some insight into the requirements of the new system, but it is less valuable than getting input from the actual users of the proposed system. It's not a strong option.

So, I'd say option (A) best answers the question, even though some of the others also have some merit.

Fun fact 1: Inta is a feminine Latvian given name. It's about time female Latvian programmers got mentioned in an exam question.
Fun Fact 2: Research in The Real World shows that about 92% of software developers are male. Yet in SD exams, most programmers are female.
In this exam, we have Inta, Sue-Lee and Janette as female IT professionals, compared with 2 males, Patel and Ethan.

Question 16

Just as Toni was returning to her car in the shopping centre car park, she witnessed another car reverse into her car and drive off. She quickly took a photograph of the car with her mobile phone, capturing the registration number.
Toni contacted her insurance company to lodge a claim. She supplied the registration number, but was told she also needed the driver’s name and home address in order to make a successful claim. This information is only held by VicRoads, the Victorian government agency responsible for issuing drivers’ licences. She contacted VicRoads, but was told she could not have this information.
In accordance with which Act is the privacy of Victorian driver’s licence holders’ information protected?

1. Privacy Act 1988
2. Health Records Act 2001
3. Privacy and Data Protection Act 2014
4. Charter of Human Rights and Responsibilities Act 2006

Answer is C.

The PDPA of the current study design replaces the IPA (Information Privacy Act) of the old study design. It only affects Victorian government agencies (and private organisations working for the government.)

'Health Records Act' - hehehe. I wonder how many kids will choose that one.

Anyway - why would an insurance company put the onus on the policy holder to identify the holder of a number plate? Silly. The Real Word™ - as mentioned several times already - does not work like that.

Question 17

Which one of the following correctly identifies features of type checking?

1. checks if only letters are entered into a field and if the text is spelt correctly
2. checks if data entered already exists in the database and if the data is reasonable
3. checks if a number is entered into a numeric field and if the number is within a set range
4. checks if numbers are entered into a numeric field and if letters are entered into a text field

Answer is D.

Notice that each option has two components. Both must be true for the option to be accepted.

The first part of (A) may be true, but the spelling part is not.

The existence check rules out option (B).

The first half of (C) is OK, but the bit after the 'and' is a range check, not a type check.

Both parts of (D) are fair and reasonable.

Question 18

A data flow diagram can be used as a tool when designing software.
Data flow diagrams must always show

1. the decisions that are made when sub-modules are called.
2. the sequence of tasks involved in completing a project.
3. where data originates and where it is stored.
4. the specific data flows between processes.

Answer is D.

The key word is "always".

(A) sounds like a flow chart or decision tree, neither of which is examinable. Such decisions are certainly not included in a DFD.

(B) is project management, such as a Gantt chart.

(C) A DFD does include external entities (data origins) and data stores, but the latter don't have to always appear... but any system without data stores would be a strange and rare beast indeed...

(D) Yeah, that feels good. Right there. Data flows in a DFD are specific (each data flow arrow must be labelled), and data certainly flows between processes. Yep.

Silly, silly examiner. DFDs are not used during design.

They are used during analysis.

Back to the naughty step. Go on.

Question 19

Some of the variables in a program are to be given new names.
Which documentation must also be modified?

1. software requirements definitions
2. milestones and dependencies
3. data dictionary definitions
4. functional requirements

Answer is C.

The SRS is produced during analysis and contains an overview of the purpose and aims of the software. It does not consider how these are to be actually implemented. (B) obviously refers to project management, which also is not concerned with the details of task implementation.
A data dictionary is the place where details of programming design are documented, and need to be kept up to date.
(D) is also part of analysis, like (A).

Dang! A19 is an easy one, yet A2 dislocated several of my brain cells.

Question 20

A local shopping centre is considering providing free wi-fi for its customers.
Which one of the following would be a non-functional requirement for its new system?

1. The system must have a failure rate of less than 3%.
2. The system must be easy for customers to connect to.
3. The system must be able to cope with 500 simultaneous users.
4. The system must have a speed of at least 5 megabits per second.

Answer is B. But I don't like it much.

[This section is in development - which is why it rambles.]

A non-functional requirement (NFR) is a quality or a performance characteristic that a solution must possess. It usually describes what a user requires from the system. Being 'easy to connect to' is such a quality.
(For those of you who were guessing, it's also the only option without a specific number in it. That makes it stand out.)

A functional requirement (FR) relates to what a system must be able to do. It often describes what a system's owner/manager requires from the system.

So...

FR - the system must be able to email a customer a record of a purchase. (What the system can do)
NFR - the system must be able to send the email within 5 minutes of the transaction. (How the system should do it)

But let's quibble, shall we? Do you have time? Good. Please sit down, and let's quibble together.
No, not like that. That's not 'quibbling'. That's just awful. Stop it.

I see an NFR is a global quality of a solution - it describes what the solution should be.
An NFR is probably not achieved in a single procedure (as an FR would be) - it is often an outcome often born of many parts of a solution, e.g. 'easy to use' or 'secure' as a whole may require many pieces of programming in different procedures to achieve.
An NFR is usually related to what the solution's user wants and experiences.

A functional requirement (FR) on the other hand is an action that a solution must be able to do, particularly related to inputs, outputs and processing requirements, e.g. produce a graph of weekly sales, send an email to a customer after each purchase.
FRs require active verbs.
Such activities need to be specifically programmed, and are usually found in specific modules.
If a programmer is told that a system must have a failure rate < 3% he/she might complain, "Yeah - but where? What module/function are you talking about? Or do you mean the code should be generally reliable? How do we test this achievement?"
Yes, an FR is rather specific, but does not have to have a number attached, as A20 implies. A20 assumes that if an objective is so specific that it has a number attached, it must be an FR.

Alas, FRs and NFRs are not defined in the study design. So, when did VCAA decide to define an FR as a specific, numerically-defined objective? - like a system objective? (Also see the system objective question in C2 below)

I rue the day that the study design had to cut back on the scope of the glossary. It's questions like this that make such definitions essential.

To quote Mr Wikipedia:

"Generally, functional requirements are expressed in the form "system must do <requirement>", while non-functional requirements are "system shall be <requirement>".

Another source says: (with my inserts in bold)

Nonfunctional Requirements (NFRs, or system qualities) describe system attributes such as security, reliability [option A], maintainability,
scalability [option C], and usability [option B] (often referred to as the "ilities"). They can also be constraints or restrictions on the design of the system (in which they may be referred to as design constraints)...

The exam's interpretation of FRs and NFRs in A20 seems to be new and idiosyncratic.

End of quibble.

Yes, you can stop now.

Other references:

• http://reqtest.com/requirements-blog/functional-vs-non-functional-requirements/
• http://stackoverflow.com/questions/16475979/what-is-functional-and-non-functional-requirement

Overall for section A - Difficult! Not many quick, easy-peasy questions. Most of them took time to ponder.

There were also too many time-outs for the examiner on the naughty step.

SECTION B - short answer questions

Instructions for Section B
Answer all questions in the spaces provided.

[x lines] indicates the number of ruled lines provided on the paper for the answers.

B1

Question 1 (3 marks)

A company stores the following data in a linear file for its payroll.

Name three fields contained in this file.  

Take any 3 of these:

• Name
• Staffid
• Staffname
• Salary

What is a "linear file"? I've never heard of one before - and not even good old Mr Google knows the term.
The study design certainly never mentions them.
The sample SD exam had questions on random and sequential files, which VCAA acknowledged as inappropriate since that key knowledge is not examinable.
And now the real exam does something quite similar.

Hence the "The Questions That Make You Sick As A Dog Award".
The reference to linear files was incorrect, but did not jeopardise the answer.

B2

Question 2 (6 marks)

A government department has been issuing new employees with a sequential employee number for over 70 years. The sequential employee numbers are now at just under 100 000 and the department would like to know who has the smallest number and is still working for the department.

A file has been created of the EmployeeName and the EmployeeNumber sorted on EmployeeName. The plan is to write a simple program to search this file for the smallest number. The pseudocode has been started.

Fill in the missing lines below to find the employee with the smallest number and to print this employee’s name and number.

Begin
   LowestEmployeeName ""

   LowestEmployeeNumber 100000

   Open datafile

   while not end of file

      read tempname, tempnum from datafile

      if tempnum < LowestEmployeeNumber then

         LowestEmployeeNumber tempnum

         LowestEmployeeName tempname

      endif

   endwhile

   print "Lowest number " , LowestEmployeeNumber , " belongs to " , LowestEmployeeName

   close datafile

8 lines were provided for the answer. I think more were needed. I don't know how the code could be done properly (including explicitly closing the file, which is good practice) in only 8 lines.

B3

Question 3 (6 marks)

To find an item in an array, a binary search, in most cases, is quicker than a linear search. However, a binary search requires the array to be sorted.
a. Explain why an array must be sorted to find a record using a binary search and how the record is found. 4 marks

Binary search checks whether the sought value is larger or smaller than the middle value of the dataset. It discards the wrong half and repeats the procedure with the remaining data. By repeatedly halving the dataset, the search area rapidly decreases and the value is soon found, or known not to exist. It is only possible to eliminate half of the dataset if the data is sorted.

b. A linear search, unlike a binary search, can find records in an unsorted array. Explain why. 2 marks

It checks every value in the dataset in turn.

Does anything else need to be said?

Six lines were provided for the answer. You don't need to use them all.

B4

Question 4 (5 marks)

A hash function has been created to store a set of different integers in an array. The hash function creates a key value from the integer. The hash function takes the integer to be stored, divides the integer by 5 and uses the remainder as an index value. For example, if the first integer is 11, then 11 divided by 5 gives a remainder of 1. This becomes the hash key.
When a collision occurs because two integers hash to the same array location, the second integer is placed in the next available space under the hash key.

The numbers to be stored via the hash function are: 11, 17, 23, 8, 13. The number 11 has already been entered.

a. Complete the diagram below.

17/5 = 3 with remainder 2. So 17 hashes to 2.
23/5 = 4 with remainder 3. So 25 hashes to 3.
8/5 = 1 with remainder 3, so 8 also hashes to 3, leading to the first collision.
13/5 = 2 with remainder 3, leading to the second collision.

A pretty straightforward and fair hash question.

b. Outline how a value is retrieved using a hash key. 1 mark

The slot in the hash table that corresponds to the hash key is checked.
The values stored in that slot are compared with the value being sought until a match is found (or no values remain, meaning the value does not exist in the table).

Only 3 lines were provided for the answer. I expected more space, and more than one mark for a fairly meaty answer.

Section B overall - A fair set of questions, especially on the new theory such as XML, hash. A decent pseudocode exercise. The search questions were challenging but very reasonable. The balance between difficulty and challenge was handled well. A couple of answers needed more space provided.

I award the stamp of approval.