SCAFFOLDING IN A HIGHER EDUCATION CONTEXT

Clare Stanier1

1School of Computing, Staffordshire University (UK)

Abstract

This paper describes the experience of using scaffolding with Computing undergraduates. Scaffolding is a well established teaching and learning approach within the constructivist framework. In a scaffolding approach, students are provided with supports, known as scaffolds, which allow learners to extend their knowledge and go beyond their existing skills and capabilities. The scaffolds are then removed, in a process known as fading, allowing students to develop as independent learners. Most of the literature on scaffolding discusses scaffolding in the context of early learners or school based instruction and scaffolding is often used in a task based context where the emphasis is on mastering specific skills rather than higher order concepts. There is comparatively little literature which discusses the use of scaffolding in Higher Education. Using scaffolding in Higher Education presents a number of issues and challenges as teaching and learning in Higher Education emphasises the higher order skills of analysis, synthesis and evaluation which are more difficult to scaffold than task based practical skills and tertiary level students have different expectations about the teaching and learning process. A scaffolding approach was used to introduce undergraduates to a complex, unstructured problem which required them to explore a range of different issues. Hard and soft scaffolds were used to support students as they moved from entry level concepts to more advanced concepts and the scaffolds were faded as students began to consider strategic approaches rather than discrete tasks. The scaffolds were easy to design and the scaffolding approach proved a good mechanism for supporting students in the initial exploration of material. However, the process of fading was much more challenging, partly because the scaffolding was linked to the assessment. Based on our experience, we propose that scaffolding approaches in Higher Education should not be linked to assessment and should be designed to support metacognitive and strategic rather than task based skills.

Keywords: Constructivist, Scaffolding, Fading, Higher Education.

1  Introduction

Some Computer Science topics lend themselves to an instructivist, step by step teaching approach [1] while other topics, by their nature, require a different pedagogy. In the second category are topics which deal with what are known as unstructured or wicked problems - problems which are characterised by incomplete, contradictory and changing requirements and/or solutions that are difficult to recognise because of complex interdependencies [2]. Teaching for a wicked problem requires an approach which supports students as they deal with multi-layered problems and incomplete or unsatisfactory solutions and which provides both structure and the space for reflection and conceptualisation. Teaching and learning do not exist in isolation but take place within a pedagogic framework. Constructivist theory is widely used in teaching to support active learning and the creation of deep knowledge, particularly in the fields of Science and Mathematics and in the different branches of Computer Science [3,4,5]. Deep knowledge is understood here as knowledge based upon an understanding of underlying concepts and models. This paper describes the experience of using a scaffolding approach to support undergraduate students working with a Computer Science wicked problem in the field of database security. An additional motivation is that it was noted during the literature review that there are differences in the way in which scaffolding is understood [6], that there is comparatively little discussion of scaffolding in a higher education (HE) environment [7] and little discussion of the student perspective on scaffolding at HE level. This is supported by a 2012 review of scaffolding in computer based learning environments which concluded that scaffolding is used with little attention to scaffold designs, or learner or task characteristics, although these variables have been shown to have a significant influence [8]. This is anomalous in a constructivist teaching approach which advocates active learning and user constructed knowledge

The remainder of this paper is structured as follows: section 2 discusses the domain background to the paper; section 3 presents an overview of scaffolding concepts and approaches; section 4 discusses the way in which scaffolding was used in this study; section 5 presents the evaluation and section 6 gives the conclusion and suggestions for future work.

2  Background

2.1  Wicked Problems in Computer Science

In the fields of Computing and Engineering, wicked problem issues tend to be those related to problem definition and management, typically linked to unstructured decision making [9]; the problems described as ‘wicked’ are problems which cannot be defined well enough to support a solution [10] or where the problem is poorly formulated, with multiple decision makers and conflicting values [11] or where the ‘wicked’ dimension derives from incomplete, contradictory and changing requirements and solutions which are difficult to recognise because of complex interdependencies [12]. In Computer Security, an additional challenge is lack of clarity as to what it means to secure the system, the lack of an immediate test to demonstrate system security and the large number of elements in potential solutions [13]. Database security is a subclass of the wider field of computer security. The ‘wicked problem’ element here derives from the complex interactions between elements of the problem, the need to work with multiple stakeholders, the fact that there is no one single solution and no clear test of success and also from the potential for harm and the implications of security failures. The requirement to protect and make available data in a database is usually well understood but the boundaries of the problem are much harder to define. Database security is an open ended problem with no single measure which can be used to determine whether a database is secure

2.2  Pedagogy

Historically the dominant teaching approach in Computer Science was instructivism [1] which supports what has been described as a teacher led, ‘drill-and-practice’ [1] pedagogy that is not suited to the teaching of wicked problems. This approach has been criticised as leading to surface learning and knowledge reproduction because learning is understood in terms of transmission of knowledge [12] rather than deep knowledge and problem solving. In the teaching of Computer Science at HE level, instructivism has largely given way to constructivist based approaches [4,5] but the nature of the Computer Science domain means that constructivist approaches also raise issues. The classical constructivist argument, developed chiefly in the context of primary and secondary education, is that a student learns from the perturbation or collision between his or her mental model and the authentic experience and that this perturbation enables students to develop their own understanding. The nature and roles of existing and alternative frameworks in developing understanding are outside the scope of this paper but it has been argued that the concept of perturbation is problematic when dealing with the advanced topics covered at HE level. Studies from other fields have claimed that free exploration of a complex environment may actually lead to poorer learning [14]. In Computer Science Education, it has been proposed that students may develop faulty mental models because they do not yet have a sufficient underpinning framework for true understanding and that in computing, faulty models ‘cause bugs’ [5]. A study on the teaching of programming at undergraduate level, identified non-viable mental models as a significant factor in poor performance [15]. One of the ways in which Computer Science HE teaching helps students to develop valid mental models is by controlling complexity using closed world and mini world approaches. A closed world approach is one in which the student does not need to take account of anything outside the immediate problem context. A mini-world is a scaled down version of a real world problem scenario tailored to meet the learning needs of students. These approaches allow students to explore concepts while limiting the complexity of the mental models required to understand the concepts. Wicked problems do not lend themselves to closed or mini-worlds since a Computer Science wicked problem, as described in 2.1 is typically open ended, complex and dynamic. For this reason we adopted a different approach, exposing students to complex concepts and using scaffolding to support students in their exploration of the issues. The aim of the scaffolds was to bridge the gap between the skills and knowledge students already possessed and the skills and knowledge needed to understand complex, interrelated problems

3  Scaffolding

Scaffolding, which has been described as building upon what students already know to arrive at something they do not know [16], is a constructivist based approach used to support learning. The original context of scaffolding was support for learning through child-adult interaction [17] and one analogy used is that of a parent helping a child to learn to walk [16]. A seminal paper on scaffolding described it as a ‘ process .. [which] consists essentially of the adult ‘controlling’ those elements of the task which are initially beyond the learner’s capacity” [18 p. 90]. Scaffolding is classically seen as having two stages which are used to support learners both in assuming control of the learning process and task completion. The first stage is the provision of scaffolds or instructional support and the second stage is the gradual fading of these supports as the learner becomes independent and no longer needs support [17]. Scaffolding is strongly associated with self regulated learning in computer supported environments [8] and the use of technology has led to different methods and strategies.

Scaffolding approaches include a wide range of tools, methods and techniques. Scaffolding based on student/tutor or group interactions may include question and procedural prompts, which at HE level may include modelling and Socratic questioning [7]. Technology based scaffolds include online scaffolds [19], digital scaffolds [20] and computer mediated scaffolding [8]. Scaffolds have been categorised in a number of ways as implicit or explicit [21,22] with explicit the most widely used [8]. Scaffolds may be hard, fixed and non negotiable, which usually implies a technology based scaffold or soft, customisable and negotiable, based on interactions with the learner, or may involve a combination of approaches [17]. Scaffolds may be described as strong, where they provide a large amount of guidance and support for the learner and weak, where the support is less. Scaffolds have also been classified by purpose as summarised in Table 1:

Table 1 Scaffolds classified by Purpose (based on [23])

Scaffolding Type / Purpose
Procedural Scaffolds / Support operational aspects
Conceptual Scaffolds / Help the student to identify knowledge gaps; geared towards understanding of problem content and gradually faded as the students develop understanding
Metacognitive scaffolds / Supports domain independent reflective and problem solving skills
Strategic scaffolds / Helps students to consider alternative views of the problem and alternative solutions

Scaffolding has been defined as the provision and withdrawal of expert support [17]. The withdrawal of support, known as the process of fading, is the key element which distinguishes scaffolding from other types of learning support [17]. A distinction is sometimes drawn between “Scaffolds-for-performance” - scaffolds which users continue to employ in order to deliver the performance required, typically associated with technology - and “Scaffolds-with-fading” which can be dismantled once the learner is capable of the required performance [6]. The term ‘scaffolding’ has been criticised as implying that when the scaffolds are removed or faded, the structure may collapse and it has been argued that a better analogy is that of an arch, supported by temporary structures until has been completed and can stand alone [24]. Scaffolding is the generally accepted term and is used here in the sense of temporary supports which can be removed. It has been argued that the fading of scaffolding must be overt to encourage the development of metacognitive skills [17]. Metacognitive skills are a required element at HE level. Fading features strongly in the theoretical literature on scaffolding but features much less so in the practical literature partly because of the difficulty of fading in a computer supported environment where students progress hierarchically from one stage to another. Scaffolds are ideally tailored to the needs of individual learners but this is challenging in a technologically supported environment [17] and also, based on the experience in this study, in a classroom environment where there are a number of students and where hard scaffolds are used. It has been argued that in practice the same level of support tends to be provided for all learners in a group, rather than being tailored to individual needs [25]. To deploy scaffolding effectively, a preliminary assessment of existing understanding is required, so that appropriate scaffolds can be developed. It has been claimed that this preliminary assessment of understanding is often missing and that a strict definition of scaffolding means that scaffolding approaches are less common than the literature suggests, with scaffolding sometimes used as a synonym for support [26].

Although there is an extensive literature about the use of scaffolding in school based contexts [18,27], there is comparatively little discussion in the literature of scaffolding in a higher education context, particularly for Computer Science. Studies include technology supported scaffolding with an undergraduate programming course [28] and the use of scaffolding with an online critical thinking course [7]. One finding from the critical thinking course was that students used the scaffolds for different purposes, and that where scaffolds were used as a template, students exhibited reflective understanding compared to students who used the scaffolds prescriptively to support the assignment. The latter group was felt to exhibit more superficial task based understanding [7,17].

4  applying scaffolding in a higher education context

4.1  Teaching and Learning Context

The Database Security module was studied as a second year option by computing undergraduate students. All students had studied the same core module in the first year of undergraduate study and this included previous experience with databases, but they had not all taken the same option modules and so had differing areas of expertise. From a teaching perspective, this was helpful as it reinforced the multi-layered nature of database security and allowed students to share different experiences. It did, however, present some issues in terms of scaffolding design, as discussed in 4.2. A feature of computer and database security problems is that the more the problem is understood, the more the complexity of the problem is appreciated. The teaching approach for the module was to introduce students to entry level database security and then build on this by moving to explore more complex and interconnected elements of security. The aim was to move from an initial task based approach to a more strategic view of the problem. Students completed a security portfolio which consisted of a series of graduated assessments and a research presentation, rather than the traditional end of semester summative assessment. Teaching was based around the portfolio, introducing students to the material covered in each task/problem and providing context and underpinning knowledge. With each successive task/problem, the complexity and level of understanding required increased as students moved from an initial, prescriptive, practical task to a final problem based element chosen by the student. After completion and submission of each task, students received individual written feedback with the opportunity to discuss the comments with the tutor. Turnaround for the feedback was within days of the assessment submission, so that lessons from one task/problem could be applied to the next. One potential issue with a portfolio based approach was that students might understand database security as a series of discrete tasks rather than as a whole and that this would limit recognition of the multi-layered, interconnected nature of database security. This was addressed by iterating and revisiting topics, discussing the strategic nature of the problem and incorporating a research element. Aspects of database security were introduced and then returned to later in the module to be discussed again in the light of the greater understanding developed as a result of exploring other elements.