Stretch tasks The four techniques

Round 1: Quiz Blitz

1. Abstraction Abstraction removes detail that does not matter for the problem. A driver needs the roads, not the paint.
2. Decomposition Breaking one huge problem into smaller ones that separate people can solve is decomposition.
3. Pattern recognition Spotting that several problems share the same structure means one method can solve them all.
4. False Abstraction leaves out detail that does not matter for this problem. The same detail may be essential for a different one.
5. Decomposition → Abstraction → Pattern recognition → Algorithmic thinking Break the problem down, strip away what does not matter, look for patterns to reuse, then write the steps.
6. Where the exits are; The corridors between rooms In a fire drill people need to know how to get out. Everything else is detail the map can leave out.
7. Travellers only need the order of stations and where lines meet Passengers change trains and count stops; they never walk the tunnels. The map keeps exactly what they need.
8. The engine model and service history of each bus Abstraction depends on the purpose. Detail that is clutter for a passenger is essential for a mechanic.
9. decomposition Decomposition turns one big problem into several smaller ones that are each easy to solve.
10. Different people can work on different parts at the same time Small, separate parts can be shared out, built in parallel and tested one at a time.
11. True Solve it once, reuse it everywhere: that is why programmers write functions.
12. Algorithmic thinking An algorithm is a set of steps in the right order. A recipe is an algorithm for food.

Round 2: Quick Fire

1. • Removing or hiding detail that is not needed
   • An example: a map keeps the roads but leaves out details such as buildings or colours
   Model answer

   Abstraction means removing detail that is not needed for the problem, so you can focus on what matters. A road map keeps the roads and junctions but leaves out details such as the colour of the buildings, the trees and the parked cars.

2. • Show the times each teacher is free
   • Let parents choose and book a time
   • Stop two parents booking the same time
   • Send a confirmation or a reminder
   • Log parents in, or identify the parent and child
   • Give each teacher a list of their appointments
   Model answer

   The system could be decomposed into: showing the times each teacher is free; letting parents choose and book a time; checking that two parents cannot book the same slot; and sending each parent a confirmation email and a reminder.

3. • It makes the problem simpler, so it is easier and quicker to solve
   Model answer

   It makes the problem simpler, so you can focus on what matters and write less code.

4. • Each one multiplies the width by the height
   • So one method can be written once and reused for all of them
   Model answer

   In every case the area is the width multiplied by the height. That means one method can be written once and reused for every rectangle, instead of solving each one separately.

Round 3: Big Think

1. • Decomposition: split the game into smaller parts
   • Gives a sensible example of a part: questions, scoring, a timer or a menu
   • Each part is easier to build and test on its own
   • Abstraction: leave out detail the quiz does not need
   • Gives an example of what to keep or leave out, such as keeping the questions and answers but not fancy graphics
   • The result is simpler and quicker to build
   Model answer

   Decomposition would let the student split the quiz into smaller parts, such as the questions, the scoring, a timer and a start menu. Each part is easier to build and test on its own, one at a time. Abstraction would help them focus on what the quiz really needs, which is each question, its possible answers and the correct answer, and leave out unnecessary detail such as fancy backdrops and animations at first. Together the two make the game simpler and quicker to build.

2. • For very small or simple programs, starting straight away can be quick
   • You can learn by experimenting and trying ideas
   • Decomposition makes a big problem manageable
   • Mistakes in the logic are cheaper to fix in a plan than in code
   • Pattern recognition lets you reuse a solution
   • Abstraction stops you wasting time on unimportant detail
   • A plan can be shared, so others can understand it and give feedback
   • A reasoned conclusion
   Model answer

   For a very small or simple program the student has a point, because writing a few lines straight away can be quick, and experimenting is a good way to learn. However, for anything bigger they are wrong. Decomposition breaks a big problem into smaller parts so it becomes manageable, and abstraction keeps you focused instead of wasting time on unimportant detail. Pattern recognition lets you spot where one solution can be reused, for example as a function. Planning also catches mistakes in the logic, which are far cheaper to fix in a plan than after hundreds of lines of code, and a plan can be shared so that other people understand it and give feedback. Overall, I disagree: a few minutes of planning usually saves much more time than it costs.

This or That

1. Decomposition Breaking one big problem into smaller ones that you can actually do is decomposition.
2. Abstraction Abstraction removes the detail that does not matter and keeps what the user needs.
3. Pattern recognition Spotting what several problems have in common lets you reuse one solution.
4. Algorithmic thinking Algorithmic thinking means working out the steps needed, in the right order.
5. Deciding what to leave out Abstraction keeps what matters for this problem and removes everything else.
6. False: another problem may need it A fire drill map does not need the plug sockets, but an electrician's plan of the same school does.
7. Splits a problem into smaller problems Decomposition breaks a problem into parts. Removing unneeded detail is abstraction.
8. Use width x height for all six All six follow the same pattern, so one method can be reused six times.

Memory Match

• A 4 Decomposition: Breaking a problem into smaller problems
• B 6 Abstraction: Removing detail that does not matter
• C 2 Pattern recognition: Spotting similarities so a solution can be reused
• D 3 Algorithmic thinking: Working out the steps, in the right order
• E 1 Algorithm: A sequence of steps that solves a problem
• F 5 Computational thinking: Solving problems in ways a computer could help with