Within the FE system in the UK, the role of employers in shaping policy has been prioritised by successive governments with the aim of establishing a more demand-led and employer-led system (Keep 2006). However, a number of recent studies from employer bodies have concluded that many students leave FE poorly equipped for the workplace (CBI 2012; UKCES 2014) and question whether the traditional model of VET is meeting the needs of today's students any longer. UKCES, for example, paints a picture of convergence and cross-disciplinary working becoming key features for work in the future, and explores how these skills might best be developed in the FE sector.

A consensus is emerging that the traditional industrial model of VET no longer serves the pedagogical needs of today's students (Zhao 2012) and the future of FE colleges, as they are currently constituted, is being challenged. With a world of expert opinion and comment available on the Internet at the click of a mouse, the question arises as to what knowledge and skills now actually need to be taught in formal settings. It would seem that colleges, if they are to remain relevant to students, employers and the wider economy, need to rethink the curriculum in terms of content and pedagogy so that it better reflects the demands of an increasingly digital and networked society (Castells 2009). What is needed, perhaps, is what Castells describes as ‘self-programmable’ workers, constantly retraining and relearning throughout working life, with the Internet firmly established as ‘the durable companion of adult life’ (Castells 2009, p. 91). The workplace increasingly demands inter-disciplinary and multi-disciplinary competencies and, to reflect this, recruitment is becoming geared much more to college leavers’ aptitudes for work rather than to qualifications obtained (CBI/Pearson 2014).

The pace and impact of technological change in the workplace, therefore, means that the VET curriculum is often out-dated by the time students actually start their courses and previously taught skills are rendered obsolete from the outset (Frey and Osborne 2013). So, against such a backdrop of ferocious automation and rapid obsolescence of equipment and skills in the workplace, as technology advances at an increasingly phenomenal rate, how best should colleges prepare students to meet these challenges? To answer this question, a brief overview of current thinking on VET pedagogy within a digital environment would help to set the framework for this project for which the preliminary findings obtained are the subject matter of this paper.

According to Lucas et al. (2012), there is a default pedagogy adopted by VET lecturers which lies in the spectrum of ‘learning by doing’ or ‘experiential learning’. Such student-centred and experiential models of learning can be seen as part of a pedagogic tradition dating back to the traditional apprenticeship model of VET which was largely based on learning by watching, working alongside peers and demonstrating competence through projects such as the production of an ‘apprentice piece’. It exemplifies what Lave and Wenger (1991) refer to as ‘communities of practice’ and promotes the notion of ‘situated learning’ which is context-driven and based on the experience of learning as a social phenomenon. Following from this, project-based learning came to the fore as a pedagogical methodology, providing simulated practice of workplace skills. Here it was hoped that students would become equipped with the skills needed for the informal learning increasingly practised in the workplace. Finally, self-directed or self-paced learning, which supports the learner progressing at their own pace and taking control of their learning, has emerged as a pedagogical approach in recent years, largely influenced by the availability of computer-based learning resources. A common strand underpins all these forms of learning: it is the emphasis on the individual learner managing their own learning whilst supported by the lecturer who adopts the role of facilitator (Rogers 1983).

Alongside this practice colleges are attempting to harness digital technologies to support learning (BIS 2014). The introduction of learning platforms, the adoption of social networking tools such as Facebook and Twitter and, more recently, the development of Massive Open Online Courses (MOOCs), point to the pivotal role that digital technologies are assuming in the 21st century VET paradigm. Given this background, and given the twin realities of the requirement for digitally competent workers for an increasingly digital workplace on the one hand, and the adoption of digital technologies by young adults on the other, the framework for a 21st century VET pedagogy cannot ignore the central role of digital technologies for teaching, learning and assessment. In short, if the workplace is to be increasingly defined by its digital architecture, then the period of learning which precedes employment needs to reflect this or risk being seen as anachronistic in its practice.

In the light of the above, this study set out to explore whether the method of Self-Organised Learning Environments (SOLEs), which is being adopted in school settings around the world (Davis 2013), could usefully form part of the pedagogical spectrum for the 21st century VET institution. The approach of SOLEs, as developed by Mitra (2013), rests on the principles of peer-to-peer learning in Internet-rich environments. Mitra's experiments established that deep learning can take place when learning is structured around the principles of self-organising systems, that groups of children can teach themselves, and that the absence of the lecturer can be a powerful pedagogical tool (Mitra, Dangwal, and Thadani 2008). This approach sits within a long tradition in FE Colleges of student-centred and discovery learning but redefines such approaches for the digital age with its emphasis on access to the Internet by students working in randomly chosen groups on a range of open questions. Interaction with the task is unstructured and the students are not directed towards resources by the lecturer. The aim of SOLE is to hand to students the total freedom to explore the question from whatever viewpoint occurs to them and see where such exploration takes them. Indeed, in the context of VET, the Internet could perhaps be described as the digital equivalent of the apprentice master– offering virtual and on demand access to expert demonstrations and explanations. Working together in groups, students collaborate in a process of collective learning – reflecting the process of ‘emergent learning’ which occurs in real workplace settings. Such learning is very similar to the realities of learning on the job when employees are required to troubleshoot emerging problems and acquire new skills independently or with colleagues. Given these similarities, the SOLE model could be seen as an appropriate methodology to develop the confident, proactive and self-programmable workers needed for today's global workplace.

The significance of research activity as a means to both personal and professional development is well documented (Barker 2005; BERA and RSA 2014; Vogrinc and Zuljan 2009). Its role as an important reflective activity for practitioners was stressed by Stenhouse (1975), whilst work in the late 1990s (Hargreaves 1996) emphasised the concept of lecturers as producers and users of research. Most of this research, however, is concerned with teachers in the school system. The FE system has tended to fall outside of mainstream research activity, with FE lecturers often experiencing institutional barriers in undertaking research (Scaife 2004). In addition, there is a question of confidence to be addressed in the FE sector in the conducting of research, given what Hillier and Morris (2010) refer to as

The real sense that researchers located in higher education (and government) did not recognise the value of, or indeed the existence of, research conducted by colleagues in FE’. (p. 92)

The absence of a core of FE research capacity in colleges was further compounded by the fact that there was no formal requirement for a lecturer qualification in the FE sector until 2001, when workforce reforms were set in train to require a teaching qualification for those entering the sector, and making mandatory continuing professional development (Foster 2005). The concepts of ‘licence to practice’ and the ‘reflective practitioner’ were both subsequently championed by the Institute for Learning, the professional body for lecturers working in further education. Latterly, however, this policy has become fragmented and, in 2013, the requirement for FE lecturers and trainers to gain a teaching qualification was removed. Furthermore, much of the research activity that occurs in FE colleges takes the form of colleges being invited to be mere participants, as opposed to actual partners, with activity being linked to incentives for funding to support lecturer release and engagement with research. Whilst the concept of ‘research practitioner’ is not a new phenomenon in FE (Hillier and Morris 2010), it is worth noting that, as far as this study is concerned, commitment to the research activity was secured without any external pressures or funding incentives. The lecturers involved have been fully ‘partners’ in the research, as opposed to ‘participants’, and have contributed at every stage of the process. This level of engagement does, we believe, provide a model for other colleges to consider as they seek to identify new modes of delivery as recommended in the recent FELTAG report (BIS 2014).

The methodology adopted for this study has been that of design experiments within an overarching case study framework. In practice, this has meant working in partnership with the college and developing the research design with the lecturers involved, based on a dual objective: to study learning in context; and to develop effective classroom interventions (Gorard, Roberts, and Taylor 2004). The potential limitations of case studies to produce generalised outcomes (Stark and Torrance 2005; Yin 2014) has been mitigated in this study by the adoption of multiple sources of evidence to enable a triangulation of findings. Classroom experiments were therefore complemented by other datasets – video footage, interviews and questionnaires. This approach embraces the methodological diversity and pluralism of mixed methods, as advocated by Johnson and Onwuegbuzie (2004), and is generally considered to be the most appropriate methodology for a study such as this, given its widespread adoption in the still nascent field of educational technology research (Jones and Kennedy 2012).

The research design consisted of running a series of controlled classroom experiments based on the techniques of random assignment administered using controlled groups of students who all completed standardised assessments so that a comparative analysis could be made. The steps outlined below were applied to all subject areas and taken together comprised the experiment cycle. At the end of each cycle data was collected in the form of test scores and audio and video interviews; the groups were then rotated and the cycle repeated so that each group experienced both the traditional and SOLE approach in a related topic area (Figure 1).

From the start of the process, the researcher has been integrated into the teaching team and adopted the role of ‘practitioner researcher’; this was mirrored by the lecturers stepping naturally into the role of ‘research practitioners’ (Cochran-Smith and Lytle 2009). Following a process of established protocols for ethical clearance, students met the researcher at the start of the term and were given written and verbal briefings to the study before it commenced. The process has been to integrate the intervention into the natural setting of the classroom and make the experience as least disruptive as possible for the students and lecturers. The experience of early classroom trials and discussions concerning suitable research methods ruled out the application of pre-intervention measurements as in all cases students had no prior experience of the skills being taught and, therefore, testing prior competency in the skills area was not appropriate. The most extreme examples of this occurred in a construction experiment which involved the erection of a mobile scaffold and a hairdressing experiment which involved demonstrating competencies in cutting hair. In both these cases the use of pure experimental design was not appropriate and therefore alternatives to present a triangulation of findings (combining qualitative classroom observations and video interviews with test data and questionnaires) led to the decision to introduce design-based thinking into the overarching methodology (Gorard, Roberts, and Taylor 2004; Reeves, Herrington, and Oliver 2005). The features of this study which were particularly pertinent in this decision were

the situating of the research within the daily practice of teaching and learning and with active collaboration of lecturers and students in the process,

This paper has attempted to describe the process in which new pedagogical approaches have emerged in vocational education and training to meet the changing needs of the modern workplace. It is too early in the research to present a detailed analysis of the results obtained from the experiments, but the following observations and reflections from the researcher and lecturer perspectives might prove instructive. Students are often surprised at how much they can learn from the Internet – we tend to assume that they are totally ‘net savvy’ but often their use is, in fact, limited to social media activity. Younger students select video content once they realise it is acceptable to make a noise in the classroom. Older students, on the whole, search for text-based content. The experiments have prompted discussions with students about how they prefer to learn. Some of the comments made at the end of the lessons are listed below:

I thought it was good. We learnt from each other and she (referring to other student taking part in the interview) could pick out something I'd not noticed. So, like for then, it all adds up in your head as you go along. (Business student)

In all of the assessments which have taken place to date, students who had been in the SOLE lesson mostly scored as well, and sometimes better, than students who had been taught by the lecturer in the lesson. Although still at an early stage of data analysis, this is an interesting finding, given that the students in the SOLE lesson did not have access to the lecturer's expertise to direct them and keep them on track in the lessons. Finally, lecturers have continued to experiment with the method themselves beyond the initial experiment cycle. Lessons have been videoed and lecturers are reflecting on the learning dynamics which takes place in SOLE lessons and a culture of supported experimentation is becoming established within the college.

Before being approached to be part of this study, the use of tools such as Skype, Google Hangouts, and YouTube, etc. was commonplace within curriculum design. However, its use was never formalised or studied as part of an experiment focussing on teaching and learning. In addition, project-based learning and group work are activities that students engage with on a daily basis but always with a strong steering from lecturers in terms of the requirements and desirable outcomes.

Initially, Self-Organised Learning Environments (SOLEs) seemed a somewhat confusing approach to something well established in our practice (namely, experiential learning and ‘learning by doing’) but, after further discussion and analysis, it was clear that this approach was something different; it was building on these foundations, taking learning as an emergent phenomenon, and combining theories of learning (Carl Rogers’ student centred learning, for example) with the possibilities that the Internet has created for learning.

At first, there was a degree of scepticism. This was mainly due to the pressures put upon all educators in the form of progression rates, success rates, value-added scores and UCAS points. However, the design of the study was such that there would be no detrimental effect on the students’ studies if the SOLE approach did not engage them and allow them to succeed – which was an important factor in deciding whether to take part in the study or not. Working on the research design from the early stages meant that the final design was something of a collaboration, which was particularly gratifying. It was also important to be given the opportunity to review the design as early iterations of the experiments influenced our future decisions about topic choices and timings of the experiments.

Seeing the first SOLE session, delivered to a level 3 BTEC National Diploma in Business, was a rather strange experience to begin with. Students were introduced, very succinctly, to the idea, given the question, asked to form groups of four to work around one computer, then literally ‘let loose’ to complete their research. The initial perception was that it wouldn't work, that students would struggle with all working on one computer and that one or two students would do all the work. The students were a little slow at first to understand the requirements of the task but they soon began working as a team to research, discuss and answer the question posed. At the end of the session each group then fed back the key points of the research completed. Most surprising was the fact that, after analysing the information presented by the students and comparing it with the material in the traditionally taught session, the students in the SOLE session had covered the majority of the traditional material that would have been delivered and also discovered a considerable amount of extra content that would not necessarily have been discussed as part of the traditional session. This in itself was an interesting outcome and something which we would see repeated in subsequent SOLE lessons.

Initially, it was thought that adopting the SOLE approach would mean topics and sessions would have to be constantly re-taught due to the prescriptive nature of the criteria that need to be covered to pass a BTEC National Diploma. There have been occasions where specific topics have had to be taught traditionally, after a SOLE session, and there have been times when areas have been re-visited where students did not understand the concept in the way that was laid down by the awarding body. However, in terms of creating deeper, rather than surface, learning, it can be argued that, although the specific theory required for the diploma was not addressed, understanding of the topic was in fact greater.

Delivering SOLE sessions can actually be a strange, frustrating experience. Lecturers are programmed to help students find the answer and have an objective for the specific knowledge that is necessary for the students to pass a test or write an assignment, and their goal is to facilitate the student to get to that desired end. However, in the SOLE sessions the lecturer simply poses the question and then facilitates the review stage of the session. It is extremely hard when asked questions such as, ‘What do I need to do?’ or, ‘Is this right?’ not to answer the question and/or guide the student in the correct direction. The notion of SOLE sessions being ‘self-correcting’, where groups naturally come to the right conclusion seemed, again, a little hard to conceive until examples of this happening occurred in several sessions.

After the initial few stages and experiments, SOLE sessions quickly became integrated into the Business team's schemes of work; students became more confident during the sessions and lecturers became more creative in their approaches. In addition, we noted that, contrary to original apprehensions and fears of time being lost, the method has actually resulted in time gains, providing lecturers with greater freedom to experiment and offer alternative lessons and enrichment activities where students can learn without the restriction of pre-defined assessment criteria.