Major themes that emerged from these brief vignettes are summarised in Table 3. The themes also echo some of the key ideas that emerged in earlier discussions within IFIP (See Table 1).
Table 3 Emerging Themes in relation to curriculum (in 2015)
Other themes, not included in Table 3, that emerged particularly from the panel discussion but also in some of the vignettes include: reasons and drivers for curriculum change, assessment as a constraint on curriculum development, the need for a more flexible curriculum incorporating personalisation and informal learning, how to incorporate careers advice and the importance of pedagogical considerations in determining curriculum structure and curriculum change. All the emerging themes and some of their interrelationships will be discussed in relation to key ideas from general curriculum theory, mentioned earlier, as they might relate to considerations for the design of the computing curriculum.
Emerging themes: Entitlement
According to Young (Young 2013), in order to harness the emancipatory capacities of learners, the curriculum should take them beyond their own experience. Thus the goal of the curriculum becomes to define its content in a world in which the entitlement to knowledge is the goal. In this endeavour “powerful knowledge” is key, defined as specialised discipline-based knowledge which is different from the experience-based knowledge that pupils bring to school (Young 2013). Clearly, as Young argues, this knowledge is not fixed nor is it equally easily identifiable across all subjects but in Computer Science, as in other disciplines, there are people committed to creating and evaluating the knowledge base, some of which is relatively stable and other aspects are changing quite rapidly.
Answers to the question of entitlement varied across the countries considered in the vignettes above. The Royal Society report emphasised entitlement and opportunity for all students including the “opportunity to learn concepts and principles from Computing” (page 10). Likewise Poland emphasises the importance of opportunity and the goal of motivating all students to use computational thinking and to engage in solving problems as well as to prepare students to consider computing and related fields as disciplines of their future study and professional career. Similarly, the Australian curriculum “Digital Technologies” rationale (http://www.australiancurriculum.edu.au/technologies/introduction) emphasises entitlement as well as economic needs and includes the importance of all students being able to “develop knowledge, understanding and skills to respond creatively to current and future needs”. While the UK, Poland and Australia and have taken the firm view that Computer Science within the curriculum is for all, Israel has opted for a segregated model based on students’ capabilities. It is noteworthy however that in Israel basic “computing and technology literacy” incorporating computational thinking is an entitlement for all based on the importance of developing problem-solving and analytical thinking skills. This “literacy” goes beyond the typical digital literacy that focuses only on using computers rather than problem-solving and computational thinking. New Zealand is actively considering incorporating Computer Science for all from age 5. Young’s argument regarding entitlement, outlined above, is that the curriculum question “what knowledge?” is primarily an epistemological one about what should constitute students’ entitlement to knowledge (Young 2013).
Learners’ entitlement implies entitlement for all and therefore we need to consider: do all students need to understand the powerful knowledge in Computer Science that we have begun to identify? There are three particularly compelling arguments, related to entitlement, for Computer Science in compulsory education. First if learners are never introduced to Computer Science as a disciplinary area and to the knowledgebase and approaches that computing academics and professionals use, then they are unlikely to be able to determine whether this is for them, and this is widely regarded as a key factor in the lack of gender diversity in the computing industry (Fisher and Margolis 2003). This therefore is an entitlement and equity issue. Second, as many in the profession have argued, programming is difficult and it takes many years to learn to program (see Robins et al. 2003 for a review). While programming is only one element of Computer Science, it is an essential element and it is inconceivable that an introductory course in Computer Science would not contain programming. Furthermore, while computing professionals do not necessarily do the programming themselves, they need to understand essentials of programming. The third argument is based on the ubiquitous nature of computing: since so much of life today is dependent upon computing everyone needs to develop understanding of Computer Science as well as any computing skills essential for participation in society.
Identification of the powerful knowledge that will constitute entitlement (Young 2013) is evident in the vignettes and the reports mentioned earlier in relation to the development in the UK, Europe and the USA in particular. The Royal Society’s description of the discipline of Computer Science encompasses foundational principles, widely applicable ideas and concepts as well as techniques and methods for solving problems and advancing knowledge as well as a distinct way of thinking and working (The Royal Society 2012). Using these headings, key areas of knowledge identified by recent curriculum reports are compared in Table 4. Similarities can be seen in the overview of the discipline, concepts and techniques and methods. Methods and techniques also incorporate ways of thinking and approaching problems. In addition the Joint Informatics Europe & ACM Europe Working Group include the importance of various more general intellectual practices such as tolerance for ambiguity and the ability to communicate and work with others (Joint Informatics Europe and ACM Europe Working Group on Informatics Education 2013). Communication and collaboration are also strands in the Australian and Polish curricula for Computer Science.
Table 4 Analysis of key areas of knowledge computer science in curriculum reports
Thus we are seeing consensus emerging about the key concepts and techniques of the discipline although perhaps not yet agreement about the importance of more general intellectual practices and social competences. The decision about whether to incorporate these more general skills and practices into curriculum frameworks for Computer Science is a dilemma that is associated with: 1) identifying what is unique about Computer Science and 2) how Computer Science is perceived in relation to other curriculum areas (Denning 2009; Voogt et al. 2015; Bell 2014). Computer Science as a relatively young discipline has suffered from loss of identity owing to various incomplete or inadequate perceptions of its nature (Denning 2009). For example problems have been identified where perceptions of Computer Science become limited to programming or to computational thinking (ibid.).
Emerging themes: Starting age, order and structure
Following the determination of powerful knowledge, a further step towards defining a curriculum is to determine some order and structure in the content. The age at which students should start to study in this field was an important emerging theme from the vignettes and a key question in determining order and sequence is when do the key knowledge elements need to be introduced? According to Winch (2013) curriculum design is about the management of growth of expertise within a subject. Winch argued that gaining a coherent view of this “epistemic ascent” within a subject by identifying schemata that are at least conceptually and normatively sustainable even if they are not yet empirically ratified is a key element in curriculum design. Moreover Winch argued the need to explore the constraints that the conceptual structure of the subject might impose on pedagogically and cognitively coherent schemata of epistemic ascent and then explore the implications of such constraints within conceptualisations of the subject.
The constraints identified by Winch depend on three interrelated issues. First, it is necessary early in a curriculum to introduce all three major types of knowledge: concepts, propositions and know-how. This is because knowledge of individual propositions implies some understanding of the concepts that such propositions express and this in turn implies a significant ability to understand and make inferences within the subject. This latter knowledge type is knowing how to do something. It would not be appropriate, for example, to focus only on practical tasks with the intention of developing know-how without developing understanding of concepts. As we have seen from an analysis of recent curriculum reports (see Table 4) there is general agreement about key concepts and techniques of the discipline. Furthermore it is common for topics in computing curricula to be classified into broad areas that are generally some variation of: programming/algorithms; data representation; digital infrastructure; digital applications and human factors (Duncan and Bell 2015). This reflects the nature of computing: a digital device runs a program (i.e. embodies algorithms); the memory and processor store and manipulates data; components are connected through buses and networks that place limitations on the system; people use them through software and this overall “system” has an impact on individuals and society through issues such as privacy, safety, access to information, and empowering people. For example, grappling with the broad issue of artificial intelligence potentially destroying human society (Future of Life Institute 2015) requires engaging with all of these areas: understanding what the nature of a computer program is (does it “think”?), what kind of knowledge base (data) it could collect and store, its access to our digital infrastructure, the way that humans would interact with it, and the philosophical concerns about human rights and what intelligence is. Many other topics around our digital future require this broad view: government surveillance, secure commerce, the environmental impact of technology (positive and negative), etc.
The second issue according to Winch (2013) is a need for a structured approach to progression in learning the basic facts and central concepts of the subject because knowledge is systematic in terms of 1) classification of its various conceptual elements; 2) the relationships between the elements and 3) the procedures required to gain and validate knowledge. The relationships outlined above between the elements in the computer systems and the human systems within which they operate point towards the need for introducing all of these elements and their relationships as early as possible in order that students may come to develop a broad view. Such an approach might be the spiral curriculum approach explained in the vignette from Poland which aims to develop understanding of all these areas and interrelationships but with different emphases in each stage of education. The arrangement of the specific concepts within a spiral curriculum depends on their complexity as well as the context in which they are introduced and is an area where further empirical research is needed.
Winch’s third issue is that the kind of knowledge required to expand and manage subject matter requires a profound understanding of the subject including all of the interacting knowledge types. This therefore is not accessible to school students but comes in more advanced studies beyond school. A constraint following from this third issue requires that the relationship between the ways in which pupils learn by simulating procedures for the acquisition of knowledge in their learning and the actual processes of expansion of disciplinary knowledge should be clarified. For example, project work in computing often involves the systems development life cycle. Winch argues that simulating such procedures may be pedagogically important in developing acquaintance with the knowledge set of the subject as well as building understanding of techniques used in knowledge management. However these simulations should not be seen as simplified versions of expert practice as that might propagate an illusion that high-level design and planning activities are generic and can be used free of the reality of the skills and materials that are needed to execute the plan. Instead it should be recognised that such expertise requires extensive knowledge and is therefore only possible in higher level courses that build upon previous structured development.
As mentioned earlier there is a view among Computer Science educators that learning programming is difficult but at the same time there is a view emphasised in the New Zealand experience that coming to programming late in students’ development is disadvantageous and that if they were to learn some of the techniques, approaches and thinking involved in programming at an earlier stage more students would be successful (Duncan et al. 2014). This therefore is both an entitlement issue for individuals looking to a fulfilling, creative and potentially lucrative career as well as of concern to countries in terms of their economic performance and prosperity. In New Zealand the reintroduction of Computer Science as a high school subject for seniors has highlighted problems associated with this late introduction. Decisions about the early introduction of Computer Science in the UK and Poland were partly based on a view from school teachers and computer scientists in higher education that aspects of Computer Science, especially programming, require gradual acquisition and development over many years. Furthermore there is evidence that exposing students to key computational thinking concepts before 12 years old is not only possible but important for developing their self-efficacy, lack of which can disadvantage girls particularly and thus impact gender diversity in University courses and the IT industry (Duncan et al. 2014).
The evidence that students will be able to learn programming before 12 years old includes: the greater ability to learn natural languages prior to adolescence combined with the similarities between learning natural languages and programming languages; the availability of programming environments and other software designed to support younger learners in learning programming and the experience and views of practitioners who have taught these topics in primary schools (ibid.). Primary school level also affords easier opportunities to apply computing concepts in an integrated environment, whereas subjects tend to be delineated at high school. For example, programming environments for primary school children often use turtle graphics, which also teach coordinate systems, geometry, number concepts as well as ideas around planning and movement, binary representation teaches patterns and properties of whole numbers and leads on to physiological questions such as human perception of colour and sound, and the topic of data compression investigates patterns in human languages.
In summary there is an emerging consensus in Computer Science education that starting to learn Computer Science in primary education, probably around 5 years old, and certainly well before age 12 is not only possible but beneficial for learning as well as developing self esteem and motivation. Furthermore all three types of knowledge need to be introduced early in the curriculum i.e. concepts, propositions and know-how and this can be achieved by a balance of theory and practice. Moreover, given the nature of Computer Science i.e. analysing, designing and developing systems and understanding their human contexts, a spiral curriculum addressing increasingly complex systems and/or introducing more complex concepts is indicated as a curriculum structure.
Emerging themes: Curriculum content and balance
As outlined previously, consensus has emerged about the key concepts and techniques of the discipline of Computer Science and this translates into curricula that incorporate broad areas of: programming/algorithms, data representation, digital infrastructure, digital applications, and human factors (Australian Curriculum Assessment and Reporting Authority (ACARA) 2014; Duncan and Bell 2015; Seehorn et al. 2011; The Royal Society 2012). More broadly still there is the issue of balance across Computer Science, IT, digital literacies and computational thinking. As we have seen the UK, Australia and Poland have incorporated elements of all these in their curricula for all students although the balance is only likely to be clear from more detailed analysis of curriculum documents including schemes of work and the curriculum in practice. Previous research comparing Computer Science curricula in different countries revealed the complexity and range of factors affecting the curriculum and how it is implemented and in particular the importance of contextual factors that have resulted in much diversity (Hubwieser et al. 2015).
Another issue emerging from analysis of the five vignettes and associated discussions concerned the importance of more general intellectual practices and social competences. The Polish curriculum incorporates a whole area on developing social competences including project based learning and taking various roles in group projects. The importance of cooperation, collaboration and communication were discussed during the development of the UK curriculum but were not included in the specified content owing to concerns about assessment. IFIP discussions generally agreed that, whereas these intellectual practices and social competences are very important in learning Computer Science, the challenges for their assessment can lead to their under-emphasis. This problem is not only evident in learning Computer Science but more generally in curricula (Webb and Gibson 2015).
Emerging themes: Teacher professional development
All of the vignettes emphasise the importance of appropriately trained teachers and the major challenge that this provides in countries currently engaged in curriculum change, e.g. UK, Australia, where there are not enough teachers with appropriate knowledge and expertise. The challenge for these countries is threefold. Firstly existing teachers who have taught a very different curriculum may not have sufficient technical knowledge and skills. Secondly their pedagogical content knowledge (Shulman 1987) has not developed in relation to the new curriculum content. Thirdly there are few new Computer Science graduates coming into teaching owing to the general shortage of Computer Science graduates. It is beyond the scope of this paper to examine approaches to teacher professional development in any depth but it is clear that there are resource implications and a need for further research and evaluation of professional development programmes. Early indications, from the vignettes discussed here, suggest that involving teachers in analysing new curricula, creating resources and exploring pedagogical approaches while being supported to develop their own subject and pedagogical knowledge and programming skills can be successful.
Emerging themes: Risks and drivers of curriculum change
As explained earlier various constraints have been considered in relation to decisions about curriculum change. Change with digital technologies in education is part of a wider global arena that includes bureaucratic as well as commercial and political drivers (Davis and Mackey 2015). For example, a lack of appropriately qualified and trained teachers increases the risk that the curriculum innovation will fail and therefore inhibits change. In the UK and Australia the discussions emphasised the risk of failure owing to inadequate teaching knowledge and expertise but chose to take the risk and attempt to mitigate it rather than delaying curriculum change. For example, developments in Australia stalled for some months following concerns raised by agencies including teacher unions. Thus decisions about desired curriculum content need to be accompanied by pragmatic considerations including capacity building and economic feasibility. Related to these decisions are considerations of the drivers for curriculum change. In the UK concerns over lack of appropriately trained teachers were high but strong drives from industry prompted political action to drive forward change on the basis of future economic advantages.
Assessment systems can impose serious constraints on curriculum development in the view of many who attended the panel. Assessment is often viewed as existing in a 3-way relationship with curriculum and pedagogy. Thus, where assessment systems impose strong requirements on how assessments may be performed, some curriculum elements may be constrained because they do not lend themselves to the assessment techniques allowed. In this way the creative aspects of Computer Science, for example, may become lost from a curriculum because they are not easily assessed through traditional exams and this concern was expressed by many who attended the panel. Opportunities provided by new computer-based assessments may help to solve this problem but significant research and development challenges remain (see for example Webb and Gibson 2015).
A further set of constraints, which merits research is about how the conceptual structure of the subject might impose on pedagogically and cognitively coherent schemata of epistemic ascent (Winch 2013). We have identified that in different countries there are different views about how well students might cope with concepts and approaches in computing. There is a body of research on learning aspects of computing extending back over at least 30 years but there is a need for continual updating as technological developments not only promote curriculum change but also provide new pedagogical opportunities making some concepts and skills easier to learn. For example new programming languages that simplify the syntax or use visual and/or block-based programming techniques are removing issues that were previously regarded as constraints by making programming easier to learn (Ben-Ari 2013; Mannila et al. 2006; Weintrop and Wilensky 2015). Furthermore there is evidence that students are able to progress from these simpler languages to the more sophisticated languages with more complex syntax generally used in business and industry (Mannila et al. 2006). However the need for teachers to understand not only the programming principles and concepts but also the ways in which these are supported through different languages remains (Ben-Ari 2013).
A further consideration discussed by the panel in Lithuania is how to take account in the school curriculum of increasing out of school and online opportunities for students to develop their computing expertise. Initiatives in some countries have emphasised these in order to make up for perceived deficits in the formal curriculum (Hubwieser et al. 2015). However our view of entitlement discussed above indicated that it is necessary to identify and specify the formal curriculum with an emphasis on entitlement and then extend learning opportunities through technology-supported environments including school learning platforms, open online courses, community-based maker spaces etc. These online resources can enable “any time” and “anywhere” learning and support teacher-directed activities such as homework, project work, after school clubs etc. as well as providing additional informal learning opportunities for enthusiastic students.