Claim

The eleven items of practice linked to this claim are denoted by the suffix [Px], and the three analyses which form my thesis by the suffix [Ax].

Teaching

I have progressed from an individual enthusiastic and creative teacher of Mathematics and Computer Studies at Scott Lidgett School, taking part in developing educational software and practice as a member of the Microcomputers in Computer Education (MICE), to a researcher in the Computers in the Curriculum Project at Chelsea College London [P1] and leader of innovation in education as senior lecturer in Ultralab at Anglia Polytechnic University [P4], taking a full part in a developing research community.

My early work concentrated on leading the improvement of the design of individual pieces of software for addressing challenging learning in the school curriculum, expressed in a design and development tool the Procedure Library [P3].

Curiosity about learning

An interest in user-interface design and mental models through the London Mental Models Group [P2] provided a basis for analysing individual learning based on Norman's ideas (1983a, 1983b). My work on new multimodal learning based on multimedia technology in the Renaissance Project  created a foundation for an Expressive Constructivism model of learning [A1], and an account of how technology enhances such learning [A2], both created to inform design practice.

Commitment to Learner-Centred Education

In this period my increasing awareness of the wider context of education informed a holistic and systemic approach to large-scale action research addressing the development of new systems of learner-centred education online as a Senior Lecturer in Ultralab [P4], and particularly in the projects Ultraversity [P9] and Inter-Disciplinary Inquiry-Based Learning [P10]. Alongside this work, I focussed in on participation, creativity and reflection in learning through a sequence of projects. For participation I helped develop methodologies for self-translation of software to local language and culture [P5] and the development of a new toy in the Étui project [P6]. For creativity I investigated user-generated content for mainstream television in Input CBBC [P7] and advised the QCA on creativity with technology [P8] benefitting further the articulation of the Expressive Constructivism model of learning [A1]. Finally, the exposure to education as systems helped me formulate the third analysis of the learner at the centre of education design [A3].

Leadership and tacit knowledge

My increasingly responsible rôles as a teacher, software developer, media designer, team leader and director led to the development of a broad and deep, but tacit knowledge of the factors which lead to effective design methods and criteria for improving design quality in iterative design cycles. A substantial number of conference presentations, in-house workshops and teaching, at both practical and academic levels, forced me to explain this in greater detail and thus refined my tacit knowledge.

Theory and constructivism

Argyris and Schön (1978) introduced two ideas of theory of action: the concepts of theories-in-use and espoused-theory. Theories-in-use can be seen as driving actual behaviour and driven by tacit knowledge. Espoused-theory is that which is used to explain when challenged, but may not be actually followed in practice. The articulation of these design factors, criteria and models of learning & education in the three analyses  [A1], [A2] and [A3] is a development from theories-in-use. One can arguably improve effectiveness (Argyris and Schön 1978), by making a closer fit between theory-in-use and espoused-theory through reflection, and so an important facet of this thesis is that the analyses developed have been tested, reflected on and improved in practice to enhance their effectiveness and validity.

Furthermore, a recently created critical overview of learning theory for non-academic innovators in a report for the EU project Hotel  [P11], has helped to locate this thesis firmly as a constructivist approach.

Table 8: Selected items from the 1980s

In 1980 I sought a position as a university researcher to develop educational resources. I was appointed as the first developer for the Computers in the Curriculum Project [P1] at Chelsea College, University of London. Over the decade I became a project leader in software development, an author of design guidelines [P3] for the team and a teacher educator involved in teacher training. I was responsible for the design and development of many educational packages based on computer simulations, working with teams including practising teachers to offer advice on the pedagogical and practical design issues. In researching human computer interface issues, I was strongly influenced by Donald Norman’s models of user-centred design (1983b). These proved practical as applied theories in my everyday work and formed the basis for my first ideas for a simplified model of the learning process, later developed as Expressive Constructivism [A1]. In this decade I joined the ESRC funded London Mental Models Group [P2], led by the late Professor Joan Bliss and Professor Jon Ogborn, and planned to conduct a PhD supervised by Professor Paul Black to focus on modelling using computers. I took part as a lecturer in the development of a diploma course to retrain teachers for Computer Studies and finally as a half-time lecturer in Mathematics Education [P1]. I co-directed the Modus project to develop computer modelling software for learners to create their own simulations, resulting in the development of Expert Builder and Model Builder software. I acted as Research Fellow on interoperability in educational software for the national Microelectronics Education Support Unit, creating several reports and peer-reviewed publications and was a member of the Software Advisory Group for the BBC Domesday Project. As pointing devices, audio, picture, video and the CD-ROM, became available, I led the technical production and contributed to the educational design of interactive multimedia for higher education in mathematics, environmental science, theatre studies and business studies for higher education. In this decade I began to be invited to academic conferences as a speaker and to take part in international seminar and workshop activity as co-tutor.

Table 9: Selected items from the 1990s

In 1990 I joined Professor Stephen Heppell to form a new research centre, ultimately called Ultralab [ P4]. Over seventeen years I offered practical, analytical and evaluative guidance to this large and geographically distributed team, offering research leadership and developing collective knowledge, procedures, values and attitudes for the development of delightful learning approaches.

Early in this decade Stephen Heppell and I began collecting materials to form a National Archive of Educational Computing and we were funded by the National Council for Educational Technology to employ Greta Mladenova to organise the materials

I continued development of new interactive multimedia CD-ROM materials, taking responsibility for production of published learning resources for Teacher Education in the Apple funded Renaissance Project . The possibility to distribute globally  and the development of new materials using Apple's HyperCard led to a realisation that we could allow for adaptation to local culture and languages by the teacher and this became the subject of a published paper, Translating Software: what it means and what it costs for small cultures and large cultures [ P5 ].

The predominant research approach of Ultralab became applied and action research, creating small and large-scale actions involving education in formal and informal contexts. I helped formulate the conceptual framework, manage development and analyse findings in many projects including a longitudinal study of online community as a learning tool Learning in the New Millenium, the University for Industry pilot Online Learning Network, the teachers’ informal continuing professional development online community TeacherNet UK and the creation of a new toy for pre-school meta-level learning, Étui [ P6  ].

This decade saw a heavy load of masters teaching, conference and workshop presentation, which gave me the opportunity to refine analyses of learning and education [ A1, A2, A3] and develop them further based on the feedback from audiences and colleagues. 

Table 10: Selected items from the 2000s

The new millenium saw a change in Ultralab, and my practice, from medium to large scale, national and international projects beginning with the headteachers’ online community Talking Heads / Virtual Heads, the development of learner’s creativity through multimedia technology for Ultralab's Summer School project and the Children's BBC Input BBC pilot [P7], and many more.

In this period I was invited as consultant to many organisations, but most significantly, to the Qualification and Curriculum Authority's panel to identify the role of technology in creativity and learning [P81]. This work was founded on a more mature form of my learning model of Expressive Constructivism [A1] and helped me to clarify the role of technology in learning [A2]. I was also invited to take part in two Teachers' TV programmes the first relating to innovation in assessment and the second, School Matters - Happiest Days?, to 'delight' in learning related to well-being in school education.

In the middle of this decade I took over as head of Ultralab for two years before joining the University of Bolton in 2007 to further develop Ultralab’s ground-breaking Ultraversity project [P9] as the Inter-Disciplinary Inquiry-Based Learning project [P10].

At the same time, I founded my own consultancy company, Core Education UK, and continued to find national and international organisations, including the Qualifications and Curriculum Authority, the Improvement and Development Agency and UNESCO, willing to employ me for my analytical perspectives [A1, A2, A3], which of course helped them to be tested and refined.

This most recent period permitted substantial reflection, analysis and articulation of ideas through the European Project HoTEL (Millwood 2013a),  peer-reviewed publications (Millwood 2014) and enabled the development of this PhD by Practice.

A basis for a pragmatic learning theory

The iterative process of production of expressions and their evaluation (described in detail below), I contend, is a model of learning that helps the educational designer make effective decisions. This is achieved by examining their designs using technology to consider how they support either, or both, expression and evaluation. This way of looking at learning was first inspired by the analysis of a single interaction (Millwood and Riley 1988), after (Norman 1983b), extended by a reduction of the four stage cycle (Kolb 1984) and supported by the radical constructivist perspective (von Glaserfeld 1995).

Von Glaserfeld viewed:

Knowledge as mental representation:
1a. Knowledge is not passively received either through the senses or by way of communication;
1b. Knowledge is actively built up by the cognising subject;
2a. The function of cognition is adaptive, in the biological sense of the term, tending towards fit or viability;
2b Cognition serves the subject’s organization of the experiential world, not the discovery of an objective ontological reality.

(von Glaserfeld 1995, 51)

The expressive constructivist model holds that expressions are made continuously as an innate activity of the human condition - as if there were an internal 'fountain' of mental and physical expressiveness. Sometimes such expression is driven by internal motivation and at other times by external stimuli and contexts including social situations, and it can even be argued that it encompasses the imaginative, information-processing basis of perception, which far from being a passive act, can be seen to be a constructivist response in the mind to the raw data arriving at retina, eardrum, inner ear, skin, muscle, mouth and nose (Marr 1982, 329-332).

Similarly the evaluation question, 'is it right?', is frequently experienced (but not necessarily linguistically) as we wonder whether we have understood or articulated well and thus struggle for clearer and better expressions.

Validity and reliability for the practitioner

As a designer, this model has supported my practice by focussing my attention on how to make expression more creative and evaluation more powerful, and has been refined through regular exposure in conference settings. Laurillard's concept of "learning through production" is closely related and it is interesting that Laurillard contends "that this has not been thoroughly researched" (Laurillard 2012, 57). So the model of expressive constructivism relies on the literature cited above and my own experiences and observations in practice for its limited validity and reliability. Nevertheless I have found the analysis to work, and have observed and filmed the expression/evaluation loop in practice, as evidenced in these films transcribed below of children playing at the computer together and engaging in exploratory talk as defined by Wegerif and Mercer (1997):

Figure 6: An example of expression and evaluation in conversation

In this example (Figure 6), Sasha is expressing his thinking about the way the game works explicitly using natural language to his brothers. After several turns through the loop re-expressing on the basis of his own evaluation listening to his own words (although seeking the others' evaluation), eventually his brother evaluates his words.

Patrick - I'll die if I go down there!

Sasha - Like getting damaged. Getting all the way damaged do you mean? Getting damaged.

Sasha - Do you know when you die? You die when you get all damaged, is what it means, when it all gets red or the green turns into red.

Sasha - The red is damage and the green isn't damage. Do you understand?

Sasha - When you get all damaged then you die, is that right Patrick?

Patrick - Yes, yes that's right.

Figure 7: An example of meta-level learning in conversation

In this second example (Figure 7), Sasha explains how he came to know about the game by repeatedly playing a section. This simple meta-level learning shows that he knows something more than the game, that he is aware of his own processes of learning.

Sasha - That's right, when you get down there you can just go from there and then straight down to there without a single damage.

Sasha - And do you know how I know that?

Sasha - Because I tried it several times - that's how I know.

Patrick - That's very good, thanks Sasha!

Figure 8: An example of expression, constrained by programming, and evaluation by computer performance

In this third example (Figure 8), Sasha uses the computer (a small robot) to evaluate his expression of the algorithm for making a zig-zag path like 'steps'. He is constrained to simple statements - move forward, turn left 90˚, turn right and start. He enters the algorithm into the robot and then sits back to watch its execution. If there is a mistake in his expression, it will be indicated by unexpected behaviour by the robot, thus offering a clue to the correctness of his algorithm and mental model. The opportunity to repeat this experience, and make sense of his ideas without interpersonal judgement can provide a useful learning opportunity and also a platform for further engagement with peers or adults.

Types of Expression and Evaluation

In tables 11, figure 9 and table 12, the following types of expressions and evaluations are proposed: internal, natural and formal, each overlapping and extending the other:

Table 11: Types of Expression

Figure 9: The expression / evaluation loop

As expressions are made, they are evaluated in order to decide of they were 'right':

Table 12: Types of evaluation

Whichever kind of evaluations are employed, new expressions are generated - often with improvements in areas identified by evaluation -  and the cycle continued until satisfaction or distraction intervenes. It is my belief that this cycle leads to the improvement of mental models, whether they are of the simplest of factual relationships or the richest of human behavioural situations.

The expression / evaluation model, which I have termed Expressive Constructivism, has been the basis for decision making and a source of directions for improvement in design practice to consider how technology can enhance learning since 1986.

The expressive constructivism model of learning directs my attention towards the learner as an active, creative individual who is best served by tools for expression on the one hand and by tools for evaluation on the other.

This notion forms the basis of the following analysis in Figure 10 of key features of technology (the terms in the speech bubbles) that might enhance expression and evaluation.The text connected by dots to each bubble serves to explain the feature.

Figure 10: How can technology enhance learning? (Millwood 2012)

By asking whether a design for use in education exploits these features, we can evaluate design proposals or consider what might be missing when a technology does not seem to work well. The analysis also acts as a framework for decision making when comparing alternative technology choices for their likely effectiveness in learning situations.

(Hammond 2009, p12)

The analysis of relatively static, immutable environments (and tools) discussed in the literature on affordances does not necessarily extend to the dynamic and extensible nature of the computer, unless we only concern ourselves with the surface physical properties. The computer not only hides its functions behind visual interface hierarchies, thus making perception of affordance challenging, but is almost infinitely extensible and customisable through programming and preferences, thus chimeric. Nevertheless, the concept of affordance serve to remind us of the need to understand the complexity of mental models that learners (and teachers) must develop to be effective in learning with computers and how obscure and diverse the functions available can be.

Focussing on the learner's perspective meant imagining a complete set of essential questions that they need answers to in order to progress in education. Normally these questions are not asked, the answers are simply supplied by the educational organisation. But by taking this perspective, gaps in the designs I developed for educational organisations could be identified and closed in the development process. For example, a design may make provision for formative assessment (7 Assessment, in the analysis) but forget that it may also be important have some means to gain an award (8 Recognition) through summative assessment.

Figure 11: The Learner at the Centre

Few real learners will articulate such concerns in the form of questions, partly because we usually do not ask them to participate in such issues, but by posing these questions as ‘constants’ we may test future proposals more effectively than starting from the status quo of a current or historical solution. In this diagram, the learner is imagined to be concerned with eight questions, relating to eight areas that an innovator should consider if they wish to make an impact on learning with an invention in technology or practice.

But the issues at the bottom of the diagram – responsibilities and rights – are at the heart of the modern dynamic that education and technology present. Learners’ entitlement to access knowledge was at the heart of the development of the National Curriculum, but we must add to that the entitlement to opportunities for access, creativity and communication and the responsibilities that a free and open education brings.

(Millwood 2009b, 20)

The lists of issues related to each question will form part of a future research & development to elaborate this analysis more fully.

In qualitative research, much of the debate about validity and reliability depends on triangulation (Golafshani 2003). The analyses presented as this thesis have been exposed to considerable 'stress testing' in numerous projects, many on the large scale and the national stage with many collaborators. In addition they have been shared with colleagues, students  and professional contacts in practice, teaching and professional dialogue. Some have been peer-reviewed in the context of projects deliverables. Thus their external validity derives from this exposure and the critique that has been offered by others leading to improvements and good evidence of generalisability.

In each analysis there is more to be developed in detail and more empirical evidence to be found to make them more rigorous, but there is enough cohesion and precision to make the task of inspiring design proposals and making design decisions possible and indeed effective. In this way I would argue that the thesis has strength in internal validity.

In the sense of truthfulness regarding my contribution, validity has been demonstrated by the triangulation of colleagues agreement, despite concerns about the crudity of percentage measures.

As for reliability, there is doubt about its relevance in qualitative research that is descriptive in nature, such as this dissertation reports, but it can be  argued that reliability derives from validity (Golafshani 2003). This thesis suffers from being my point of view and analysis, and the potential is there for bias. Some of this is offset by the nature of my practice as collaboration (others have had a hand in the development of the ideas) and other aspects can be triangulated through the evidence offered for each portfolio item. Another indicator of reliability is that the thesis is couched in terms that do not depend on particular educational contexts or developments in technology and have stood the test of time in my career.

My new post at the time of writing (November 2013) is as Course Director at Trinity College Dublin for their Masters in Technology and Learning. The course provides me with the best opportunity to make use of my background and, in designing and preparing learning experiences for the students, to improve my theoretical analyses in three ways: flesh out the detail; enhance coherence; and provide interpretation with respect to design decisions that educational developers must make.

At the same time, I hope to seek PhD supervision rôles that encourage development of these analyses including correction and extension. This may include seeking empirical evidence to strengthen what I claim is a valid designer's tool-set to have greater scientific reliability.

The other major direction forward is to develop the potential of the National Archive of Educational Computing to become a trusted and rich research resource. This includes continuing to design new software, but based on some of the ideas developed and forgotten over the last four decades which are stored in the archive. The comparison of a modern design with its updated pedagogical thinking against the original material will in itself clarify trends in education, but also provide a basis for critiquing both past and present. This strategy has the added value of providing a relevance, currency and forward direction to the archive which is easy to dismiss as simply backward looking.

In order to achieve this objective, I will need to make new alliances with disciplines I have so far not touched upon and learn about historical interpretation, artefact curation & preservation, cataloguing and knowledge engineering - a challenge I relish!