What is computational problem solving?

‘Computational problem solving’  is the iterative process of developing  computational solutions to problems. Computational solutions are expressed as logical sequences of steps (i.e. algorithms), where each step is precisely defined so that it can be expressed in a form that can be executed by a computer. Much of the process of computational problem solving is thus oriented towards finding ways to use the power of computers to design new solutions or execute existing solutions more efficiently.

Using computation to solve problems requires the ability to think in a certain way, which is often referred to as ‘computational thinking’. The term originally referred to the capacity to formulate problems as a defined set of inputs (or rules) producing a defined set of outputs. Today, computational thinking has been expanded to include thinking with many levels of abstractions (e.g. reducing complexity by removing unnecessary information), simplifying problems by decomposing them into parts and identifying repeated patterns, and examining how well a solution scales across problems.

Why is computational problem solving important and useful?

Computers and the technologies they enable play an increasingly central role in jobs and everyday life. Being able to use computers to solve problems is thus an important competence for students to develop in order to thrive in today’s digital world. Even people who do not plan a career in computing can benefit from developing computational problem solving skills because these skills enhance how people understand and solve a wide range of problems beyond computer science.

This skillset can be connected to multiple domains of education, and particularly to subjects like science, technology, engineering or mathematics (STEM) and the social sciences. Computing has revolutionised the practices of science, and the ability to use computational tools to carry out scientific inquiry is quickly becoming a required skillset in the modern scientific landscape. As a consequence, teachers who are tasked with preparing students for careers in these fields must understand how this competence develops and can be nurtured. At school, developing computational problem solving skills should be an interdisciplinary activity that involves creating media and other digital artefacts to design, execute, and communicate solutions, as well as to learn about the social and natural world through the exploration, development and use of computational models.

Is computational problem solving the same as knowing a programming language?

A programming language is an artificial language used to write instructions (i.e. code) that can be executed by a computer. However, writing computer code requires many skills beyond knowing the syntax of a specific programming language. Effective programmers must be able to apply the general practices and concepts involved in computational thinking and problem solving. For example, programmers have to understand the problem at hand, explore how it can be simplified, and identify how it relates to other problems they have already solved. Thus, computational problem solving is a skillset that can be employed in different human endeavours, including programming. When employed in the context of programming, computational problem solving ensures that programmers can use their knowledge of a programming language to solve problems effectively and efficiently. 

Students can develop computational problem solving skills without the use of a technical programming language (e.g. JavaScript, Python). In the PILA module, the focus is not on whether students can read or use a certain programming language, but rather on how well students can use computational problem solving skills and practices to solve problems (i.e. to “think” like a computer scientist).

How is computational problem solving assessed in PILA?

Computational problem solving is assessed in PILA by asking students to work through dynamic problems in open-ended digital environments where they have to interpret, design, or debug computer programs (i.e. sequences of code in a visual format). PILA provides ‘learning assessments’, which are assessment experiences that include resources and structured support (i.e. scaffolds) for learning. During these experiences, students iteratively develop programs using various forms of support, such as tutorials, automated feedback, hints and worked examples. The assessments are cumulative, asking students to use what they practiced in earlier tasks when completing successive, more complex tasks.

To ensure that the PILA module focuses on foundational computational problem solving skills and that the material is accessible to all secondary school students no matter their knowledge of programming languages, the module includes an assessment application, ‘Karel World’, that employs an accessible block-based visual programming language. Block-based environments prevent syntax errors while still retaining the concepts and practices that are foundational to programming. These environments work well to introduce novices to programming and help develop their computational problem solving skills, and can be used to generate a wide spectrum of problems from very easy to very hard.

What is assessed in the PILA module on computational problem solving?

Computational problem solving skills

The module assesses the following set of complementary problem solving skills, which are distinct yet are often used together in order to create effective and efficient solutions to complex problems:

• Decompose problems

Decomposition is the act of breaking down a problem goal into a set of smaller, more manageable sub-goals that can be addressed individually. The sub-goals can be further broken down into more fine-grained sub-goals to reach the granularity necessary for solving the entire problem.

• Recognise and address patterns

Pattern recognition refers to the ability to identify elements that repeat within a problem and can thus be solved through the same operations. Adressing repeating patterns means instructing a computer to iterate given operations until the desired result is achieved. This requires identifying the repeating instructions and defining the conditions governing the duration of the repetition.

• Generalise solutions

Generalisation is the thinking process that results in identifying similarities or common differences across problems to define problem categories. Generalising solution results in producing programs that work across similar problems through the use of ‘abstractions’, such as blocks of organised, reusable sequence(s) of instructions.

• Systematically test and debug

Solving a complex computational problem is an adaptive process that follows iterative cycles of ideation, testing, debugging, and further development. Computational problem solving involves systematically evaluating the state of one’s own work, identifying when and how a given operation requires fixing, and implementing the needed corrections.

Programming concepts

In order to apply these skills to the programming tasks presented in the module, students have to master the below set of programming concepts. These concepts can be isolated but are more often used in concert to solve computational problems:

• Sequences

Sequences are lists of step-by-step instructions that are carried out consecutively and specify the behavior or action that should be produced. In Karel World, for example, students learn to build a sequence of block commands to instruct a turtle to move around the world, avoiding barriers (e.g. walls) and performing certain actions (e.g. pick up or place stones).

• Conditionals

Conditional statements allow a specific set of commands to be carried out only if certain criteria are met. For example, in Karel World, the turtle can be instructed to pick up stones ‘if stones are present’.

• Loops

To create more concise and efficient instructions, loops can communicate an action or set of actions that are repeated under a certain condition. The repeat command indicates that a given action (i.e. place stone) should be repeated through a real value (i.e. 9 times). A loop could also include a set of commands that repeat as long as a Boolean condition is true, such as ‘while stones are present’.

• Functions

Creating a function helps organise a program by abstracting longer, more complex pieces of code into one single step. By removing repetitive areas of code and assigning higher-level steps, functions make it easier to understand and reason about the various steps of the program, as well as facilitate its use by others. A simple example in Karel World is the function that instructs the turtle to ‘turn around’, which consists of turning left twice.

How is student performance evaluated in the PILA module?

Student performance in the module is evaluated through rubrics. The rubrics are structured in levels, that succinctly describe how students progress in their mastery of the computational problem solving skills and associated concepts. The levels in the rubric (see Table 1) are defined by the complexity of the problems that are presented to the students (simple, relatively complex or complex) and by the behaviours students are expected to exhibit while solving the problem (e.g., using functions, conducting tests). Each problem in the module is mapped to one or more skills (the rows in the rubric) and classified according to its complexity (the columns in the rubric). Solving a problem in the module and performing a set of expected programming operations thus provide evidence that supports the claims about the student presented in the rubric. The more problems at a given cell of the rubric the student solves, the more conclusive is the evidence that the student has reached the level corresponding to that cell. 

Please note: the rubric is updated as feedback is received from teachers on the clarity and usefulness of the descriptions.

Learning management skills

The performance of students on the PILA module depends not just on their mastery of computational problem solving skills and concepts, but also on their capacity to effectively manage their work in the digital learning environment. The complex tasks included in the module invite students to monitor, adapt and reflect on their understanding and progress. The assessment will capture data on students’ ability to regulate these aspects of their own work and will communicate to teachers the extent to which their students can:

• Use resources

PILA tasks provide resources such as worked examples that students can refer to as they build their own solution. Students use resources effectively when they recognise that they have a knowledge gap or need help after repeated failures and proceed to accessing a learning resource.

• Adapt to feedback

As students work through a PILA assessment, they receive different types of automated feedback (e.g.: ‘not there yet’, ‘error: front is blocked’, ‘try using fewer blocks’). Students who can successfully adapt are able to perform actions that are consistent with the feedback, for example inserting a repetition block in their program after the feedback ‘try using fewer blocks’.

• Evaluate own performance

In the assessment experiences designed by experts in PILA, the final task is a complex, open challenge. Upon completion of this task, students are asked to evaluate their own performance and this self-assessment is compared with their actual performance on the task.

• Stay engaged

The assessment will also collect information on the extent to which students are engaged throughout the assessment experience. Evidence on engagement is collected through questions that are included in a survey at the end of the assessment, and through information on students’ use of time and number of attempts.  

Resources

Learn about computational problem solving-related learning trajectories:

• Rich, K. M., Strickland, C., Binkowski, T. A., Moran, C., & Franklin, D. (2017). K-8 Learning Trajectories Derived from Research Literature: Sequence, Repetition, Conditionals. Proceedings of the 2017 ACM Conference on International Computing Education Research, 182–190.
• Rich, K. M., Strickland, C., Binkowski, T. A., & Franklin, D. (2019). A K-8 Debugging Learning Trajectory Derived from Research Literature. Proceedings of the 50th ACM Technical Symposium on Computer Science Education, 745–751. https://doi.org/10.1145/3287324.3287396
• Rich, K. M., Binkowski, T. A., Strickland, C., & Franklin, D. (2018). Decomposition: A K-8 Computational Thinking Learning Trajectory. Proceedings of the 2018 ACM Conference on International Computing Education Research  - ICER ’18, 124–132. https://doi.org/10.1145/3230977.3230979

Learn about the connection between computational thinking and STEM education:

• Weintrop, D., Beheshti, E., Horn, M., Orton, K., Jona, K., Trouille, L., & Wilensky, U. (2015). Defining Computational Thinking for Mathematics and Science Classrooms. Journal of Science Education and Technology, 25(1), 127–147. doi:10.1007/s10956-015-9581-5

Learn how students apply computational problem solving to Scratch:

• Brennan, K., & Resnick, M. (2012). Using artifact-based interviews to study the development of computational thinking in interactive media design. Paper presented at annual American Educational Research Association meeting, Vancouver, BC, Canada.