Common misconceptions or difficulties students may have

Conceptual links to other areas of the specification – useful ways to approach this topic to set students up for topics later in the course.

The most important technique is the ability to break down a complex task into simple sub-tasks and write (or recycle) selfcontained code in the form of functions and procedures (and classes/objects at the higher end of ability). The code that isn’t made modular becomes quickly overwhelming and leads to endless debugging rather than program improvement, despite having an added benefit of learners becoming familiar with the debugging tools of their IDE.

It is advisable to introduce functions and procedures as early as possible, before branching and iteration. Left too late, modular programming tends to put learners under stress as it requires them to relearn the skills they have just become comfortable with. Teaching learners that we define our routines in one place and trigger them in another place in a program is great for focusing them and reduces the number of blocks they need to track at any one time. While it might be easier to introduce procedures before functions, it is important that learners get a chance to use both, so they can later use the most appropriate ones for the task at hand.

The use of functions requires more planning and confidence in coding and will not come easy to all learners. Objectoriented programming takes modularity to a new level and introduces another layer of variables – object variables. It is important that learners understand that every task can be done either through procedures with global/local variables, or functions with local variables only and parameter passing between the functions, or with objects and it is advisable to ask pupils to produce multiple versions of the same program using all three paradigms.

Sequence, iteration and branching are best introduced through the various validation routines and any mistakes, especially logical mistakes, become obvious quite soon, simplifying the debugging. Nested branching (an if statement inside another if statement) is not trivial but the process of validating input data requires it, e.g., the type check often needs to be performed before range check and needs planning and sequencing of nested branching.

Non-conditional Iteration is best illustrated through lists/ arrays, especially 2d arrays. Being able to generate a times table or read a CSV file are the rights of passage, while conditional Iteration lends itself to validation and user interface duties.

Understanding recursion requires confident knowledge of functions, parameter passing, selection and iteration. In other words, it brings together almost all of the concepts of this section. While recursion can always be replaced by simple iteration, conceptually, it requires more meticulous program planning and hence is a good discriminator in assessment and is also needed for efficient sorting and searching.