Testing, test data & error types – GCSE Computer Science exam questions with mark scheme

1Grade 4–54 marks

(a) Explain the difference between a syntax error and a logic error. [2]

(b) For each line below, state whether it contains a syntax error or a logic error. [2]

(i) IF score > 50 THN

(ii) A program meant to find the mean of 5 numbers uses mean ← total / 4

(a) Syntax error: code breaks the rules / grammar of the programming language, so it cannot be translated / run(1)
(a) Logic error: the program runs but produces an incorrect / unexpected result(1)
(b)(i) Syntax error(1)
(b)(ii) Logic error(1)
Answer: (a) Syntax: breaks language rules, won't run; Logic: runs but gives wrong output (b)(i) syntax (ii) logic
Examiner tip: Saying a logic error 'is a mistake in the code' is too vague. Say the program still runs but gives the wrong result.

2Grade 4–54 marks

A program accepts the age of a student at a secondary school. Valid ages are 11 to 18 inclusive.

(a) Give an example of each type of test data and state the expected result. [3]

(i) normal (ii) boundary (iii) erroneous

(b) Explain why a programmer tests a program with erroneous data. [1]

(a)(i) Normal: a typical valid value, e.g. 14 (any of 12–17), with expected result accepted. Value and result both needed(1)
(a)(ii) Boundary: 11 or 18 with result accepted, or 10 or 19 with result rejected. Value and result both needed(1)
(a)(iii) Erroneous: a value that should be rejected, e.g. 25, −3, 'twelve', 14.5, with expected result rejected / error message. Value and result both needed(1)
(b) To check that the program rejects invalid data / does not crash / handles it correctly (e.g. displays an error message)(1)
Answer: (a) e.g. 14 accepted; 11 accepted; 'abc' rejected (b) to check invalid data is rejected without crashing
Examiner tip: Each piece of test data needs its expected result. A value with no outcome may not score.

3Grade 6–76 marks

A subroutine should return True if a test mark is valid. Valid marks are from 0 to 100 inclusive.

SUBROUTINE validMark(mark)
   IF mark > 0 AND mark < 100 THEN
      RETURN True
   ELSE
      RETURN False
   ENDIF
ENDSUBROUTINE

(a) The subroutine is tested with the marks 50, 0, 100 and 101. For each test, state the expected result and the actual result. [2]

(b) State the type of error in the subroutine. [1]

(d) Explain why testing with only normal data would not have found this error. [1]

(e) The program reads mark using STRING_TO_INT(USERINPUT). Explain why the program could still fail if a user types ten. [1]

(a) Expected: True, True, True, False (all four correct; allow valid/accepted for True and invalid/rejected for False)(1)
(a) Actual: True, False, False, False (all four correct; allow valid/accepted for True and invalid/rejected for False)(1)
(b) Logic error(1)
(c) mark ≥ 0 AND mark ≤ 100 (accept >= and <=; accept mark > −1 AND mark < 101; accept the conditions in either order; accept any logically equivalent condition, e.g. NOT (mark < 0 OR mark > 100))(1)
(d) Normal data (e.g. 50) gives the correct result; the error only affects the values at the edges of the range (0 and 100), so it only shows up with boundary data(1)
(e) 'ten' is not a number / not in digit form, so STRING_TO_INT cannot convert it to an integer, so the program crashes / gives a (runtime) error before the range check happens. Either the failed conversion or the crash earns the mark; stating that a type check is needed is not required(1)
Answer: (a) expected T,T,T,F; actual T,F,F,F (b) logic (c) mark ≥ 0 AND mark ≤ 100 (d) only boundary values 0 and 100 are affected (e) the conversion of 'ten' fails
Examiner tip: Boundary testing exists to find exactly this kind of off-by-one mistake with < and ≤.

Book a free introductory session and we'll find exactly which step is costing you.