Producing Robust Programs - Worksheets, Questions and Revision

20 original exam-style questions - 11 pages of questions with a full mark scheme - free printable PDF.

Download PDF Jump to mark scheme (page 12)Read the revision guide

Next in this strand

Revision Library

revisionlibrary.co.uk

GCSE · Computer Science

P2.3 Producing Robust Programs

OCR J277 · Calculators not allowed · about 115 minutes

Total Marks

Name: _______________________________ Date: ____ / ____ / ______

Answer ALL questions. Show all your working.

A software company is developing a new app. The lead developer tells the team that the app must be 'robust'.

(a)State what is meant by a robust program.(1)

(b)Give two examples of situations, other than invalid data entry, that a robust program should be able to handle without crashing.(2)

(Total for Question 1 is 3 marks)

A programmer is choosing the most suitable input validation check for different fields on an online form.

(a)Which check is most suitable to make sure a UK postcode is entered in the correct pattern, e.g. SW1A 1AA?(1)

A) Presence check
B) Format check
C) Range check
D) Length check

(b)Which check ensures that a customer's date of birth field is not left empty?(1)

A) Presence check
B) Format check
C) Range check
D) Check digit

(c)A 'number of tickets' field must only accept whole numbers from 1 to 8. Which check identifies that 12 has been entered?(1)

A) Length check
B) Format check
C) Range check
D) Presence check

(Total for Question 2 is 3 marks)

Amara is writing a program in an IDE.

(a)Amara's program will not run. The IDE highlights a missing ENDIF and displays an error message before the program starts. State the type of error this is.(1)

(b)Amara fixes this, and the program now runs, but it displays the wrong total every time. State the type of error this is now most likely to be.(1)

(c)Explain why the IDE's syntax checker did not detect the error described in part (b).(2)

(Total for Question 3 is 4 marks)

A program accepts a whole number representing a student's exam mark. Valid marks are whole numbers from 0 to 100 inclusive.

(a)State the term used for realistic test data that would be expected during normal use of the program.(1)

(b)Give two boundary data values that should be used to test this input.(2)

(c)Give one example of erroneous test data for this input.(1)

(d)State why it is important to test a program using erroneous data as well as normal data.(1)

(Total for Question 4 is 5 marks)

A registration form has three input fields: Email address, Password (must be 8 to 16 characters), and National Insurance number (format: 2 letters, 6 digits, 1 letter, e.g. QQ123456C).

(a)State which validation check would be used to make sure the password is not shorter than 8 characters.(1)

(b)State which validation check would be used to make sure the National Insurance number matches the required pattern.(1)

(c)A user submits the form with the email address field left empty. State which validation check would detect this.(1)

(d)A user enters someone else's real National Insurance number, which matches the correct format exactly. Explain why the format check would not detect this as a problem.(2)

(Total for Question 5 is 5 marks)

Study the following pseudocode, which calculates the mean of five numbers stored in an array:

total <- 0
count <- 0
FOR i <- 1 TO 5
total <- total + n[i]
count <- count + 1
NEXT i
a <- total / count
OUTPUT a

(a)State two maintainability techniques, other than adding comments, that could be used to improve the readability of this code without changing what it does.(2)

(b)Rewrite the line 'a <- total / count' using a more meaningful identifier name for a.(1)

(c)The array n stores a list of numbers. Suggest one more suitable identifier name for this array.(1)

(Total for Question 6 is 4 marks)

Programmers can test their code at different stages of development.

(a)State the difference between iterative testing and final/terminal testing.(2)

(b)State one advantage of using iterative testing rather than relying only on final testing.(1)

(Total for Question 7 is 3 marks)

A bank wants to make its online banking login more secure than a password alone.

(a)Identify two authentication methods, other than a password, that the bank could add to its login process.(2)

(b)Explain how using two-factor authentication reduces the risk of a criminal accessing a customer's account, even if the criminal has stolen the customer's password.(2)

(Total for Question 8 is 4 marks)

Binary and denary conversions (no calculator).

(a)Convert the denary number 173 to an 8-bit binary number. Show your working.(2)

(b)Convert the 8-bit binary number 11010110 to denary.(2)

(Total for Question 9 is 4 marks)

Study the following algorithm, written in OCR Exam Reference Language:

total <- 0
count <- 0
numbers <- [12, 7, 25, 3, 18]
FOR i <- 0 TO 4
  IF numbers[i] > 10 THEN
    total <- total + numbers[i]
    count <- count + 1
  ENDIF
NEXT i
OUTPUT total
OUTPUT count

(a)Complete a trace table for this algorithm, showing the value of i, numbers[i], total and count for each pass of the loop.(4)

(b)State the two values output by the program.(2)

(Total for Question 10 is 6 marks)

A programmer is using an IDE to try to find the cause of a bug in their program.

(a)Identify two tools, other than a syntax checker, provided by an IDE that can help a programmer to find the cause of an error in their code.(2)

(b)Explain how using breakpoints together with single stepping can help a programmer to find the cause of a logic error.(3)

(Total for Question 11 is 5 marks)

A warehouse uses a 7-digit product code. The final digit is a check digit calculated as follows: each of the first 6 digits is multiplied by a weight (7, 6, 5, 4, 3, 2, reading left to right), the six results are added together, and the check digit is the remainder when this sum is divided by 11 (if the remainder is 10, the check digit is 0). The first 6 digits of a new product code are 3 4 5 2 1 6.

(a)Calculate the check digit for this product code. Show your working.(3)

(b)State the name of this type of validation check and give one real-world example of where a check digit is used.(2)

(Total for Question 12 is 5 marks)

Binary, hexadecimal and denary conversions (no calculator).

(a)Convert the 8-bit binary number 10111010 to hexadecimal.(2)

(b)Convert the hexadecimal number 3F to denary.(2)

(c)State one reason why programmers often use hexadecimal, rather than binary, to represent values such as colour codes or MAC addresses.(1)

(Total for Question 13 is 5 marks)

Write an algorithm, using pseudocode (OCR Exam Reference Language), that repeatedly asks the user to input a whole number from 1 to 10 inclusive. The algorithm must reject and re-ask for any number outside this range, and once a valid number has been entered, it should output that number. Any correct logic will be accepted; do not worry about minor syntax errors.

(Total for Question 14 is 5 marks)

Study the following algorithm, written in OCR Exam Reference Language:

count <- 0
REPEAT
  INPUT mark
  count <- count + 1
  IF mark < 0 OR mark > 100 THEN
    OUTPUT "Invalid, try again"
  ENDIF
UNTIL mark >= 0 AND mark <= 100
OUTPUT "Valid mark entered after ", count, " attempt(s)"

A user enters the following values, in order, when the program runs: -5, then 150, then 68.

(a)Complete a trace table to show the value of mark, count and any output produced for each iteration of the loop.(4)

(b)State the final message that the program outputs.(1)

(Total for Question 15 is 5 marks)

A programmer runs their test plan. One of the tests has failed: the expected result and the actual result produced by the program do not match. Describe the steps the programmer should take, after this failed test, to resolve the problem.

(Total for Question 16 is 3 marks)

An online shop is developing a website where customers enter their payment card details to make a purchase. Discuss the defensive design techniques the programmers should use, and explain why each technique is important for producing a robust and secure program.

(Total for Question 17 is 6 marks)

A program validates a user's age before allowing them to register for a website. A valid age must be a whole number from 16 to 65 inclusive. Complete a test plan for this validation check by giving one suitable piece of test data for each of the types below, and stating the expected outcome for each.

(a)Normal data: give a value and state the expected outcome.(2)

(b)Boundary data: give a value and state the expected outcome.(2)

(c)Erroneous data: give a value and state the expected outcome.(2)

(Total for Question 18 is 6 marks)

Write an algorithm, using pseudocode (OCR Exam Reference Language), that repeatedly inputs whole-number exam scores from the user. Only scores from 0 to 100 inclusive should be included in a running total. The user enters -1 to indicate they have finished entering scores. Once the user has finished, the algorithm should output the average of the valid scores that were entered (assume at least one valid score is entered). Any correct logic will be accepted; do not worry about minor syntax errors.

(Total for Question 19 is 6 marks)

A software company is developing a large program. Evaluate the effectiveness of using iterative testing throughout development, compared with relying only on final/terminal testing once the whole program is complete.

(Total for Question 20 is 6 marks)

Mark scheme · P2.3 Producing Robust Programs

Question 1

(a) B1 A program that continues to run / does not crash even when given unexpected or invalid input (oe)
(a) Answer: A program that keeps running correctly (does not crash) even if the user enters unexpected or invalid data.
(b) B1 First valid example, e.g. a required file is missing
(b) B1 Second valid example, e.g. a network connection is lost / division by zero / peripheral not connected
(b) Answer: Any two of: a required file is missing; a network connection is lost; a division by zero is attempted; a peripheral (e.g. a printer) is not connected.

Question 2

(a) B1 B) Format check
(a) Answer: B) Format check
(b) B1 A) Presence check
(b) Answer: A) Presence check
(c) B1 C) Range check
(c) Answer: C) Range check

Question 3

(a) B1 Syntax error
(a) Answer: Syntax error
(b) B1 Logic error
(b) Answer: Logic error
(c) M1 A syntax checker only checks that the code follows the grammar/rules of the programming language
(c) A1 It cannot check whether the logic/algorithm produces the correct or intended result
(c) Answer: A syntax checker only checks that the code follows the grammar/rules of the programming language; it cannot tell whether the logic of the algorithm actually produces the correct/intended output.

Question 4

(a) B1 Normal (test) data
(a) Answer: Normal test data
(b) B1 0
(b) B1 100
(b) Answer: 0 and 100
(c) B1 Any value outside 0-100 (e.g. 105 or -3) or of the wrong data type (e.g. 'ninety')
(c) Answer: e.g. 105 (a whole number outside the valid range) or 'ninety' (text instead of a number)
(d) B1 To check the validation routine correctly rejects/detects invalid data and the program does not crash or accept it (oe)
(d) Answer: To check that the validation routine correctly rejects invalid data and the program does not crash or accept it by mistake.

Question 5

(a) B1 Length check
(a) Answer: Length check
(b) B1 Format check
(b) Answer: Format check
(c) B1 Presence check
(c) Answer: Presence check
(d) M1 The format check only tests that the data follows the correct pattern/structure
(d) A1 It cannot check that the data is true/accurate/belongs to the correct person
(d) Answer: The format check only tests that the data matches the correct pattern; it cannot check whether the data is genuine or belongs to the correct person, so a valid-looking but stolen number would pass.

Question 6

(a) B1 Consistent/correct indentation of the FOR loop body
(a) B1 Using meaningful/descriptive identifier names instead of a, n and i
(a) Answer: Indenting the lines inside the FOR loop; using meaningful identifier names instead of a, n and i.
(b) B1 Any meaningful name replacing a, e.g. average <- total / count
(b) Answer: average <- total / count
(c) B1 Any suitable meaningful name, e.g. numbers, scores or marks
(c) Answer: e.g. numbers

Question 7

(a) B1 Iterative testing is carried out throughout development, testing each module/section as it is written
(a) B1 Final (terminal) testing is carried out once the whole program has been completed, testing it as a whole
(a) Answer: Iterative testing tests each module/section of code as it is written during development; final (terminal) testing tests the whole completed program at the end.
(b) B1 Errors are found and can be fixed earlier/closer to when they were introduced, making them easier/cheaper to locate and correct (oe)
(b) Answer: Faults are found soon after they are introduced, so there is less code to search through and they are cheaper and easier to fix.

Question 8

(a) B1 First valid method, e.g. biometrics (fingerprint/facial recognition)
(a) B1 Second valid method, e.g. a one-time PIN/code sent by text or app (two-factor authentication) / security question / physical token
(a) Answer: Any two of: biometrics (e.g. fingerprint or facial recognition); a one-time code sent by text/app (two-factor authentication); a memorable/security question; a physical security token or card reader.
(b) M1 The criminal would also need the second factor, e.g. a code sent to the customer's own phone or a biometric feature
(b) A1 The criminal is very unlikely to also have access to this second factor, so access is still denied even with the correct password
(b) Answer: Even with the correct password, the criminal would also need the second piece of evidence, such as a one-time code sent to the customer's own phone or a fingerprint, which they are very unlikely to have, so access is still blocked.

Question 9

(a) M1 Correct method shown using place values 128, 64, 32, 16, 8, 4, 2, 1
(a) A1 10101101 cao
(a) Answer: 10101101
(b) M1 Correct method shown, e.g. adding place values for each set bit
(b) A1 214 cao
(b) Answer: 214

Question 10

(a) B1 i=0: numbers[0]=12, condition true, total=12, count=1
(a) B1 i=1: numbers[1]=7, condition false, total and count unchanged (12, 1)
(a) B1 i=2 and i=3: numbers[2]=25 true (total=37, count=2); numbers[3]=3 false (unchanged)
(a) B1 i=4: numbers[4]=18, condition true, final total=55, count=3
(a) Answer: i=0: 12, true, total=12, count=1 | i=1: 7, false, total=12, count=1 | i=2: 25, true, total=37, count=2 | i=3: 3, false, total=37, count=2 | i=4: 18, true, total=55, count=3
(b) B1 total = 55 cao
(b) B1 count = 3 cao
(b) Answer: total = 55, count = 3

Question 11

(a) B1 First valid tool, e.g. breakpoints
(a) B1 Second valid tool, e.g. single stepping (step through/step into/step over) / variable (watch) window / cross-reference tool
(a) Answer: Any two of: breakpoints; single stepping (stepping through code line by line); a variable/watch window; a cross-reference tool.
(b) M1 A breakpoint pauses execution of the program at a chosen line
(b) M1 Single stepping then executes the remaining code one line/instruction at a time while the programmer watches variable values
(b) A1 This lets the programmer identify exactly which line causes a variable to take an unexpected/incorrect value, pinpointing the cause of the logic error
(b) Answer: A breakpoint pauses the program at a chosen line, then single stepping runs the remaining code one instruction at a time so the programmer can watch how each variable's value changes and identify exactly which line causes a value to become incorrect, pinpointing the logic error.

Question 12

(a) M1 Correct products shown: 3x7=21, 4x6=24, 5x5=25, 2x4=8, 1x3=3, 6x2=12
(a) M1 Correct total sum = 93
(a) A1 Check digit = 5 cao
(a) Answer: 5
(b) B1 Check digit (check-digit validation)
(b) B1 Any one valid real-world example, e.g. barcodes, ISBNs, bank account numbers
(b) Answer: Check digit; used, for example, in barcodes, ISBNs or bank account numbers.

Question 13

(a) M1 Correct method: splitting into nibbles 1011 and 1010, or converting via denary 186
(a) A1 BA cao
(a) Answer: BA
(b) M1 Correct method: 3x16 and F=15 identified
(b) A1 63 cao
(b) Answer: 63
(c) B1 Hexadecimal is a shorter/more compact way of representing the same value than binary, so it is quicker to read/write and less prone to transcription errors (oe)
(c) Answer: Hexadecimal represents the same value using far fewer digits than binary, so it is quicker and easier for humans to read, write and check.

Question 14

1 Uses a loop that repeats the input process, e.g. REPEAT...UNTIL or WHILE...ENDWHILE
1 INPUT statement used inside the loop to accept the number from the user
1 A condition correctly tests whether the number is outside the range 1 to 10, e.g. number < 1 OR number > 10
1 The loop correctly continues while/until the number is invalid, and stops once a valid number (1 to 10) is entered
1 OUTPUT statement used after the loop to display the valid number

Answer:

REPEAT
  INPUT number
  IF number < 1 OR number > 10 THEN
    OUTPUT "Invalid, please try again"
  ENDIF
UNTIL number >= 1 AND number <= 10
OUTPUT number

Question 15

(a) B1 Iteration 1: mark=-5, count=1, outputs 'Invalid, try again'
(a) B1 Iteration 2: mark=150, count=2, outputs 'Invalid, try again'
(a) B1 Iteration 3: mark=68, count=3, no 'Invalid' output produced
(a) B1 Loop correctly shown to end after iteration 3, since 0≤68≤100 is true
(a) Answer: Iteration 1: mark=-5, count=1, outputs 'Invalid, try again'. Iteration 2: mark=150, count=2, outputs 'Invalid, try again'. Iteration 3: mark=68, count=3, condition true, loop ends.
(b) B1 'Valid mark entered after 3 attempt(s)' cao
(b) Answer: Valid mark entered after 3 attempt(s)

Question 16

M1 Use IDE debugging tools (e.g. breakpoints, single stepping) or trace the code to locate the line(s) where the actual result diverges from the expected result
M1 Identify and correct the fault/logic error in the code
A1 Re-run the failed test (and ideally the whole test plan) to confirm the fix works and no new faults have been introduced
Answer: Use IDE debugging tools such as breakpoints and single stepping (or trace the code by hand) to find exactly where the program's behaviour diverges from what was expected; correct the fault in the code; then re-run the failed test, and ideally the whole test plan again, to confirm the program now produces the expected result and that no new errors have been introduced.

Question 17

Level 1 (1-2): Basic. Identifies one or two defensive design techniques with little or no explanation of their purpose. Answer lacks structure or detail.
Level 2 (3-4): Good. Identifies several relevant defensive design techniques with some explanation of why each helps produce a robust or secure program. Some structure evident.
Level 3 (5-6): Excellent. Identifies and explains a range of relevant defensive design techniques in detail, with clear links to why each reduces the risk of crashes or security breaches for this specific scenario (payment data). Answer is well structured and uses correct technical terminology throughout.
Indicative content:
- Input validation (e.g. presence, format, length, range and check digit checks) on the card number, expiry date and CVV fields, to reject data that is missing, the wrong length or the wrong pattern before it is processed.
- Input sanitisation, to remove or escape unexpected characters, helping to prevent malicious input such as SQL injection being used to attack the underlying database.
- Authentication (e.g. requiring the customer to log in and/or a one-time code for two-factor authentication), to confirm the person entering the payment details is who they claim to be.
- Maintainability techniques (e.g. clear comments, indentation, meaningful identifier names) make the code easier for other programmers to review and spot flaws in.
- Thorough testing, using normal, boundary and erroneous test data, checks that the validation and sanitisation routines work correctly before the site goes live.
- Overall justification: because card details are highly sensitive, failing to use these techniques could allow invalid data to crash the program or allow fraud/data theft, so defensive design directly reduces both technical failure and security risk.

Question 18

(a) B1 Any value from 17 to 64, e.g. 30
(a) B1 Expected outcome: accepted/valid, no error message
(a) Answer: 30 -> accepted
(b) B1 16 or 65
(b) B1 Expected outcome: accepted/valid, since the range is inclusive
(b) Answer: 16 (or 65) -> accepted
(c) B1 Any value outside 16-65 (e.g. 15 or 70) or of the wrong data type (e.g. 'twenty')
(c) B1 Expected outcome: rejected, error/re-prompt message displayed
(c) Answer: 70 -> rejected, error message shown and re-input requested

Question 19

1 Uses a loop that continues until the sentinel value -1 is entered
1 INPUT statement used inside the loop to accept each score
1 Correctly tests whether the score is within the valid range (0 to 100) before adding it to the total
1 Maintains a running total of valid scores and a count of how many valid scores have been entered
1 Invalid scores (outside 0-100 and not equal to -1) are correctly excluded from the total/count
1 After the loop ends, calculates and outputs the average (total / count) using the correct values

Answer:

total <- 0
count <- 0
INPUT score
WHILE score <> -1
  IF score >= 0 AND score <= 100 THEN
    total <- total + score
    count <- count + 1
  ENDIF
  INPUT score
ENDWHILE
average <- total / count
OUTPUT average

Question 20

Level 1 (1-2): Basic. Makes one or two isolated points about testing with little comparison or justification.
Level 2 (3-4): Good. Describes relevant features, benefits and/or drawbacks of iterative and final testing with some comparison and justification, though the argument may be one-sided or not fully developed.
Level 3 (5-6): Excellent. Gives a well-balanced, justified evaluation comparing iterative and final testing, weighing benefits (e.g. earlier/cheaper error detection, easier to isolate faults) against drawbacks (e.g. time overhead, integration errors only found later), and reaches a supported conclusion.
Indicative content:
- Iterative testing tests each module/section of code as it is written, so faults are found and can be corrected close to when they were introduced, while the relevant code is still fresh in the programmer's mind.
- This generally makes faults quicker and cheaper to isolate, because there is less new code to search through compared with testing the whole finished program at once.
- However, iterative testing takes extra time throughout development and cannot detect faults that only appear when different modules are integrated/combined.
- Final/terminal testing checks the complete, integrated program, so it can reveal problems that only occur when modules interact (e.g. data passed incorrectly between modules) that testing individual modules would miss.
- Relying only on final testing means faults could have been introduced much earlier in development, making them harder, slower and more expensive to trace back and fix, and may delay the whole project if major faults are found late.
- A balanced conclusion: most professional development uses both, iterative testing to catch faults early within modules, and final testing to confirm the complete, integrated program meets requirements and works as a whole.

Mark my answers