Algorithms, Flowcharts and Pseudocode

1 State two differences between a WHILE loop and a REPEAT UNTIL loop. 2

One mark for each of the following, to a maximum of 2.

• A WHILE loop tests its condition before the body runs and a REPEAT loop tests it afterwards
• A WHILE loop may run zero times, a REPEAT loop always runs at least once

Model answer. A WHILE loop tests its condition before the body is executed, whereas a REPEAT UNTIL loop executes the body first and tests the condition afterwards. As a result, the body of a WHILE loop may never run at all if the condition is false to begin with, while the body of a REPEAT UNTIL loop always runs at least once.

2 Describe the purpose of a trace table and explain how it helps a programmer find a logic error. 3

One mark for each of the following, to a maximum of 3.

• It records the value of each variable at each step of the algorithm
• The programmer can compare the actual values with the values expected
• The step at which a value first becomes wrong identifies where the error is

Model answer. A trace table records the value held by every variable after each step of an algorithm, together with anything the algorithm outputs. Working through the algorithm by hand and filling the table in lets the programmer compare the values the program actually produces with the values it should be producing at that point. The first row where a value differs from what was expected shows exactly which instruction is at fault, which is what makes a trace table so effective at locating logic errors, since a logic error produces a wrong answer without producing any error message at all.

3 Describe three rules that must be followed when drawing a flowchart. 3

One mark for each of the following, to a maximum of 3.

• A decision is shown as a diamond containing a condition with two labelled outputs
• Input and output use a parallelogram and a process uses a rectangle
• The flowchart begins and ends with a terminal symbol and every step is joined by arrows showing the flow

Model answer. A flowchart must begin and end with a terminal symbol, a rounded rectangle containing START or STOP, so that it is clear where the algorithm starts and finishes. Each type of step must use the correct shape: a rectangle for a process such as a calculation, a parallelogram for input or output, and a diamond for a decision. A decision must contain a condition rather than an instruction and must have exactly two outputs, each labelled Yes or No, and every symbol must be connected by arrows that show the direction the algorithm flows in, with any loop drawn as an arrow returning to a point earlier in the sequence.

4 Write an algorithm, using pseudocode or a flowchart, that asks the user for ten numbers, adds them up and outputs the total. 5

One mark for each of the following, to a maximum of 5.

• A total is initialised to zero before the loop
• A loop that repeats exactly ten times
• A number is input from the user inside the loop
• The number is added to the running total inside the loop
• The total is output after the loop has finished

Model answer. SET total TO 0, then FOR count FROM 1 TO 10 DO, and inside that loop SEND 'Enter a number' TO DISPLAY, RECEIVE number FROM (INTEGER) KEYBOARD and SET total TO total + number, then END FOR, and finally SEND total TO DISPLAY. The total must be set to zero before the loop begins so that it is not reset on each pass, the input and the addition both sit inside the loop so that they happen ten times, and the output sits after END FOR so that the total is displayed once, when all ten numbers have been added.

5 Explain why an algorithm might be written as pseudocode before it is written as a program in Python. 4

One mark for each of the following, to a maximum of 4.

• Pseudocode is not tied to any one programming language
• The programmer can concentrate on the logic without worrying about syntax
• It is easier for other people, including non programmers, to read and check
• Errors in the design can be found and corrected before time is spent coding

Model answer. Pseudocode is deliberately not tied to any particular programming language, so the programmer can set out the logic of the solution without having to be correct about the syntax of Python at the same time. That separation matters because it means the design can be judged on whether the steps are right, rather than on whether a colon or a bracket is missing. Pseudocode is also far easier for other people to read, including team members who work in a different language and clients who do not program at all, so the design can be reviewed and agreed before any code exists. Finding a flaw in the logic at that stage costs a few minutes, whereas finding the same flaw after the program has been written, tested and documented costs a great deal more.