Massive Open Online Courses (MOOCs) as a Disruptive Innovation in Higher Education? Name: Dr Jürgen Rudolph Student ID: 1616033 Programme: Master of Education Supervisor: Dr Francisco Ben Date: 23 May, 2014 Education Research Project (nine units), submitted in partial fulfillment of the requirements of the Master in Education at the School of Education, University of Adelaide
211
Embed
Massive Open Online Courses (MOOCs) as a Disruptive Innovation in Higher Education?
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Massive Open Online Courses (MOOCs)
as a Disruptive Innovation
in Higher Education? Name: Dr Jürgen Rudolph Student ID: 1616033 Programme: Master of Education Supervisor: Dr Francisco Ben Date: 23 May, 2014
Education Research Project (nine units),
submitted in partial fulfillment of the
requirements of the Master in Education at the
School of Education, University of Adelaide
2
Table of contents List of figures and tables 4 Preface and Acknowledgements 6 1. Introduction 9 1.1 Background to the research 9 1.2 Aim of the research, research problem and research questions 10 1.3 Significance and justification of the research 13 1.4 Methodology 15 1.5 Chapter Overview 16 1.6 Definitions of key terms 17 1.7 Delimitations and key assumptions 19 1.8 Conclusion 20 2. Literature Review 21 2.1 Introduction 21 2.2 A brief history of MOOCs 22 2.2.1 cMOOCs 22 2.2.2 The Stanford higher education experiment 25 2.2.3 Stanford University experiment spin-off #1: Coursera 26 2.2.4 Stanford University experiment spin-off #2: Udacity 27 2.2.5 edX 28 2.2.6 Comparison of the three dominant xMOOCs 28 2.3 Do MOOCs offera high quality educational experience? 31 2.3.1 Who are the participants, what is their motivation and what do the 31
completion rates signify? 2.3.2 Required MOOC literacies 34 2.3.3 What courses or content may best be taught via MOOCs, if any? 34 2.3.4 Assessment 35 2.3.5 Accreditation, certification, credit transfer and badges 38 2.4 Are MOOCs sustainable? 40 2.4.1 How much do MOOCs really cost? 41 2.4.2 Revenue models 42 2.4.3 MOOCs as marketing devices 45 2.4.4 Are MOOCs sustainable? 46 2.5 Do MOOCs offer a viable alternative to traditional education? 47 2.6 MOOCs as disruptive innovation and providing access to the 48
bottom of the pyramid? 2.6.1 Are MOOCs a revolution? 48 2.6.2 MOOCs as disruptive innovation? 49 2.6.3 Do MOOCs provide access for the bottom of the pyramid? 52 2.6.4 Unbundling and rebundling 55 2.7 Conclusions and research model 56 3. Methodology and Methods 60 3.1 Introduction and research design 60 3.2 Interpretivist research paradigm and symbolic interactionism 61 3.3 Qualitative research methodology and the circular model 64
3
3.3.1 Qualitative research methodology 64 3.3.2 Qualitative versus quantitative research 64 3.3.3 A circular model of the qualitative research process and theory 66 3.4 The six steps of the interview process 67 3.4.1 Links between research issue and interview questions 67 3.4.2 Choice of interview method: convergent expert interview 71 3.4.3 Sample Frame – discussion of selection of experts 74 3.4.4 Introduction of the eight expert interviewees 77 3.4.5 Conducting interviews 82 3.4.6 Transcribing interviews 83 3.4.7 Coding and interpreting interviews 84 3.5 Participant observation in cyberspace as a secondary method 87 3.6 Assessment of the quality of qualitative research 88 3.7 Ethical considerations 91 3.8 Conclusions 92 4. Empirical analysis 93 4.1 Introduction 93 4.2 MOOCs, cMOOCs, xMOOCs and their predecessors 93 4.2.1 The MOOC components 93 4.2.2The first MOOC 97 4.3 The quality of the educational experience in MOOCs 99 4.3.1 Who are the participants, what is their motivation and what 102 do the completion rates signify 4.3.2 Required MOOC literacies 105 4.3.3 Expertise versus wisdom of the crowds 106 4.3.4 What courses or content may be best taught via MOOCs, if any 108 4.3.5 Assessment 109 4.3.6 Accreditation, credit transfer and badges 118 4.4 Are MOOCs sustainable? 120 4.4.1 The cost of MOOCs 120 4.4.2 MOOCs as marketing devices 122 4.5 MOOCs as a viable alternative to higher education? 123 4.6 MOOCs as disruptive innovation and providing access to the 129
bottom of the pyramid? 4.6.1 MOOCs are not much of a revolution 129 4.6.2 MOOCs are not a disruptive innovation 130 4.6.3 Do MOOCs provide access to the Bottom of the pyramid? 133 4.6.4 Unbundling and rebundling 134 4.7 Conclusions 135 5. Synthesis, recommendations and conclusions 136 5.1 Synthesis of literature review and research findings 136 5.1.1 Research question 1: the quality of the educational experience in MOOCs 136 5.1.2 Research question 2: the sustainability of MOOCs 139 5.1.3 Research question 3: MOOCs are not a viable alternative to 141
traditional higher education 5.1.4 Research question 4: MOOCs are not a disruptive innovation and 142
do not provide access to the bottom of the pyramid 5.2 Benefits and recommendations for implementation 145
4
5.2.1 Challenges and benefits 145 5.2.2 Six recommendations 147 5.3 The future of MOOCs and opportunities for further research 150 5.3.1 Opportunities for further research 150 5.3.2 The future of MOOCs 151 5.4 Conclusions 151 Bibliographical references 153 Appendices 211 Appendix 1: Comparison of various qualitative methods 211 Appendix 2: The history of distance learning and education and 212
technology in the perspectives of experts Appendix 3: My own experience with MOOC assessment 216 Appendix 4: Author’s biodata 219
List of figures and tables Figure 1.1 The indescribable, indestructible, unstoppable MOOCs 9 Figure 1.2 Research questions in the qualitative research process 12 Figure 1.3 The world population and income pyramid 14 Figure 1.4 The path of disruptive innovation 19 Figure 2.1 MOOC timeline 22 Figure 2.2 Professors Thrun and Norvig 26 Figure 2.3 Thrun demonstrates Google’s self-driving cars 26 Figure 2.4 Georgia Tech’s MSc in Computer Science MOOC 28 Table 2.1 Comparison of MOOCs 30 Figure 2.5 World map of enrolment for all HarvardX offerings as of 10 32
February, 2014 Figure 2.6 Udacity’s free versus subscription model 44 Figure 2.7 College: a packaged bundle 56 Figure 2.8 Research model 59 Figure 3.1 Components of qualitative research design 61 Table 3.1 Research perspectives in qualitative research 63 Figure 3.2 Linear versus circular models of process and theory 66 Table 3.2 Constructive alignment of research question 1 and interview questions 68 Table 3.3 Constructive alignment of research question 2 and interview questions 69 Table 3.4 Constructive alignment of research question 3 and interview questions 70 Table 3.5 Constructive alignment of research question 4 and interview questions 70 Table 3.6 Constructive alignment of sections in chapters 2 and 4 with 71
research questions and interview questions Table 3.7 Convergence in convergent expert interviews 74 Table 3.8 Sampling decisions in the research process 74 Table 3.9 Research sample 76 Figure 3.3 Mayring’s process analysis of general content analysis 86 Figure 4.1 What MOOC means 94 Figure 4.2 The meaning of OER 95 Figure 4.3 Screenshot of Keth Devlin’s “Introduction to Mathematical 102
5
Thinking” Coursera MOOC Figure 4.4 Dino 101: Dinosaur Paleobiology 104 Figure 4.5 Dave Cormier’s MOOC #Rhizo 14 116 Figure 4.6 Screenshot of MOOC on “Beauty, Form and Function: An 123
Exploration of Symmetry” Figure 4.7 A future with only 10 universities 125 Figure 4.8 Screenshot from PowerPoint presentation: “Higher Education 129
Doesn’t Do Revolutions” Figure 4.9 Three core elements of higher education 131 Figure 4.10 MOOC on Mobiles for Development by IIT Kanpur & 135
Commonwealth of Learning Figure 5.1 Theses 1-4 on the perception of experts on MOOCs (Research 140
Question 1) Figure 5.2 Theses 5-7 on the perception of experts on MOOCs (Research 142
Question 2) Figure 5.3 Theses 8-9 on the perspectives of experts on MOOCs (Research 143
Question 3) Figure 5.4 Theses 10-13 on the perspectives of experts on MOOCs (Research 146
Question 4) Figure 5.5 Learn with your peers 151
6
Preface and Acknowledgements
I started the M.Ed. programme of the University of Adelaide almost three years ago.
My motivation at that time was lifelong learning and specifically, learning more about
(higher) education. I have been in education most of my life, largely on the receiving
end (having completed three postgraduate qualifications thus far), but also as an
Academic Director for two Private Education Institutions (PEIs) in Singapore and as a
lecturer, trainer and facilitator in a wide variety of subjects. Consequently, it appeared
to make sense to pursue an education-specific qualification.
When I started the M.Ed. programme, I was in no hurry, as my focus was on learning,
not the credential. Then, in 2012, I was persuaded to take up Republic Polytechnic’s
Specialist Diploma in Applied Learning and Teaching (SDALT) that I completed in
2013 – this was a practical course, specifically designed for PEI faculty like myself,
and I found it quite meaningful. On the flipside, it derailed me for nine months from
my M.Ed. studies.
My current role as Senior Lecturer for Kaplan Higher Education Singapore also kept
me rather busy, which occasionally made me question the meaningfulness of the
whole undertaking. Did I really need another postgraduate qualification, now that I
had an educational qualification (which was one of my original motivations)? I
occasionally had to remind myself that quitting was not part of my vocabulary.
Writing the thesis has been very rewarding and rather painful at times. The reason
why it has been great fun and intellectually highly stimulating is due to the support of
quite a few wonderful people who deserve to be acknowledged.
My thesis supervisor Dr Francisco Ben has been wonderfully supportive, being
infectiously enthusiastic about the topic and getting me a couple of much-needed
extensions from the university. Rhys Johnson, my awesome boss at Kaplan, created a
highly conducive work environment for my humble research – without his support,
this thesis would have never been completed. Lydia Tan also deserves special
mention for her inestimable assistance with literature search, the compilation of
7
bibliographical references, transcription of interviews, drawing of figures and other
technical assistance. Jervin Lim also thankfully did some of the transcribing.
Professor Chad Perry was very kind in sharing some of his most recent and not yet
published writings about convergent expert interviews and thesis writing with me.
Professor Jan Herrington from the Department of Education at Murdoch University
was extremely helpful in sharing with me her incredible knowledge about higher
education and MOOCS. Critically, she suggested to me many MOOC experts as
potential interviewees, and allowed me to drop her name, when contacting them,
which I am certain made my success rate much higher than it would otherwise have
been. Stephen Li, Kaplan Singapore’s Senior Director for University Relations,
provided me with some great references on MOOCs that enriched my review of the
literature.
If this project is worth reading, much of the credit is due to the expert interviewees
who so generously shared with me their time. Dr Bror Saxberg, Chief Learning
Officer of Kaplan Inc., made himself available during a visit to Singapore for a face-
to-face interview, and also previously had been most kind to comment on some of my
articles on MOOCs. Professor Jim Jackson, a former supervisor of mine, entertained
me for 90 minutes via Skype during his expert interview, and also commented on
some of my previous work. His sharp legal mind led me in some less-explored
directions. Dr Samson Tan, a Singaporean learning technologies expert and lecturer
of mine in the SDALT programme, was also very kind in letting me interview him.
I feel very honoured that all three key people who started the whole MOOC
phenomenon granted me interviews. Professor George Siemens, Stephen Downes
and Dave Cormier were all most kind in sharing their precious time for Skype
interviews, very much in the spirit of connectivism. Great thanks are also due to Sir
John Daniel, a famous thought leader on education, who also granted me an
interview.
Professor Tania Aspland taught me two modules during the M.Ed. programme and
was a greater influence on me than she would know, recommending two wonderful
books (by Flick and Biggs & Tang), and being one of my great role models as a
8
university teacher. Dr I Gusti Darmawan almost made me consider quantitative
research, which is a very big compliment indeed, and deserves much praise as a
teacher-mentor.
Two years ago, Dr Scott Gardner opened up a whole new world of knowledge
management to me, and it was he, together with Dr Amy Huang and Alfred Tseng
(all from Murdoch University), who were kind enough to permit MOOCs as a
possible additional essay topic for my favourite module as a teacher, BUS378
Knowledge and Organisational Learning. This led to intriguing discussions with some
of my students.
Lastly, and most importantly, huge thanks are due to my wonderful and long-
suffering wife Clara, who has been very understanding during the countless burnt
evenings and weekends that I spent on this project.
Any shortcomings of this work, needless to say, are entirely mine.
9
1. Introduction 1.1 Background to the research
The topic of disruptive innovation in higher education is highly current and important.
Disruptive innovation may threaten the traditional approach to higher education (of
studying on physical university campuses) and, simultaneously, it may open up
hitherto largely unavailable opportunities to people at the base of the income pyramid.
In this research project, it will be evaluated whether Massive Open Online Courses
(MOOCs) are a potential disruptive innovation in higher education.
MOOCs have garnered much interest for a variety of reasons. One is the massiveness
(which is already part of the concept) of their unprecedented scale, with some of the
courses having more than 100,000 participants. Another reason is the global demand
for knowledge and information that is manifested by highly diverse learners of all
ages, backgrounds and levels of expertise from all over the world. MOOCs also
“challenge current credit transfer and accreditation systems and practices” (Australian
Trade Commission, 2012, p. 12).
Figure 1.1 The indescribable, indestructible, unstoppable MOOCs
Source: Rudolph, 2013.
10
1.2 Aim of the research, research problem and research Questions
The objective of the research project is to shed some light on MOOCs, and to use
Geertz’s (1973) famous term, provide a thick description of the phenomenon. The
research problem is to improve our understanding of MOOCs through the
perspectives of experts. In other words: What are the perspectives of experts on
MOOCs in terms of their potential as a disruptive innovation in higher
education?
With research questions being amongst the most important parts of a research study
plan, Kervin, Vialle, Herrington and Okely (2006) have come up with some rules for
research questions, including:
“Try not to use a research question that can be answered with the words Yes or
No… Starting these questions with the words ‘To what extent…’ (or similar
exploratory words) makes them… much more researchable questions… Be
wary of research questions starting with the word Can or Should…” (Kervin et
al., 2006, p. 51).
Flick (2009, p. 98) emphasises that research questions should be formulated “in
concrete terms with the aim of clarifying what the field contacts are supposed to
reveal”. As a consequence of the above-mentioned considerations and also of my
qualitative methodological approach (see chapter 3), my research questions – or
guiding questions (O’Donoghue, 2007) – are as follows:
1. What, according to experts, constitutes a high quality educational experience
in the context of MOOCs?
2. To what extent are MOOCs sustainable (in the perspectives of experts)?
3. To what extent do MOOCs offer a viable alternative to traditional higher
education, in the perspectives of experts?
4. In the perspectives of experts, to what extent, if at all, are MOOCs
revolutionising higher education, a disruptive innovation and/or are opening
up opportunities for the base of the income pyramid?
11
O’Donoghue (2007) cautions that “there is no point in designing a study about
participants’ perspectives on something unless we are convinced before we
commence the study that it is something about which they have fairly well formed
views and that they feel sufficiently free to discuss these views” (pp. 36-37). Through
my earlier conversations with Dr Bror Saxberg, Chief Learning Officer of Kaplan,
USA, who easily qualifies as an expert on disruptive innovation in higher education, I
had one first indication that the above research questions were appropriate.
In qualitative research, the initial understanding of the topic should be regarded as
preliminary and thus be constantly reevaluated (Kleining, 1982). Consequently,
research questions are to be reviewed at various stages in the qualitative research
process (see Figure 1.2). The principle of openness questions the a priori formulation
of hypotheses (Hoffmann-Riehm, 1980). In the context of this project, the theoretical
structuring of the issue was postponed till the structuring by the interviewees had
emerged.
However, the principle of openness, by no stretch of the imagination, means that
attempts to formulate research questions should be abandoned, as Flick (2009, p. 98)
advises: “The less clearly you formulate your research question, the greater is the
danger that you will find yourself in the end confronted with mountains of data
helplessly trying to analyse them”. Research questions should be formulated clearly
and unambiguously “as early as possible in the life of the project” (Flick, 2009, p.
129). It is then that the principle of openness applies, as we should remain open to
surprises, and we can revise our research questions to become more concrete and
more focused (Flick, 2009).
12
Figure 1.2 Research questions in the qualitative research process
Source: Flick, 2009, p. 99.
What Flick (2009) and Glaser & Strauss (1967) call sensitising concepts is similar to
the key concepts that are defined and discussed in section 1.6. Other methodological
issues are discussed in Chapter 3.
13
1.3 Significance and justification of the research
For higher education to be of a good quality, a deep approach is necessary. Such a
deep approach goes beyond mere memorising and note-taking, students are to be able
to describe, explain, relate, apply and theorise in high-level engagement (Biggs and
Tang, 2011). In Outcomes-based teaching and learning (OBTL), teachers should
address six goal clusters: (1) higher order thinking skills; (2) basic academic success
skills; (3) discipline-specific knowledge and skills; (4) liberal arts and academic
values; (5) work and career development; and (6) personal development (Angelo and
Cross, 1993).
Over the past century, tertiary education has become increasingly massified and it has
become critically important for employability in the global knowledge economy. On
an individual level, earning power appears directly related to the level of education,
and research has shown that tertiary education historically forms a U-curve in terms
of return on investment (ROI). This means that at present, there may be a particularly
high ROI for tertiary education. On a national level, due to globalisation and
increased international competition, the quality and quantity of outcomes of a
country’s education system appear related to a nation’s future trajectory and
international economic position (Goldin and Katz, 2008).
In developed nations, almost everybody is a consumer of education. In the US, for
instance, it was observed as early as 1970 that higher education has been transformed
from a privilege into an assumed right, and for a growing number of adults, into an
expected obligation (Trow, 1970). However, there are widening gaps and increased
inequities in parts of the developed world, not to speak of emerging economies. The
typical disruptive innovation strategy to start with areas of non-consumption may
also be applicable to education, including higher education (Christensen, Horn &
Johnson, 2008; Richardson, 2010).
Prahalad’s base of the pyramid (BoP) theory is perhaps the easiest way to illustrate
that out of a global population of more than seven billion people, the majority
continue to be non-consumers of higher education. Figure 1.3 shows an updated
version of Prahalad and Hart’s original illustration from 2002, as the global
14
population has since dramatically grown. Prahalad and Hart’s (2002) estimation of
four to five billion people being non-consumers remains intact.
Figure 1.3 The world population and income pyramid
Source: Arnold & Valentin, 2013, p. 1906. MEP refers to moderate and extreme
poverty, BOP to base of the pyramid.
This project is not the first work to consider the concept of disruptive innovation
within the context of higher education. The concept of disruptive innovation was
originally applied to manufacturing, but later also to K12 (kindergarten to age 18)
education and to higher education (Christensen et al., 2008; Christensen & Eyring,
2011). Key results of disruptive innovations are products or services that exhibit
affordability, accessibility, convenience and simplicity (Christensen et al., 2008;
Richardson, 2010). The innovative products are good-enough, and become accessible
by the masses in the bottom of the pyramid (Prahalad, 2004).
While the concept of disruptive innovation has been applied to tertiary education
before, MOOCs as a potential major disruptive innovation in higher education have
not been systematically researched as yet. This is not surprising, as MOOCs are really
a phenomenon that has only caught global attention in 2012 (despite the fact that the
first MOOC dates back to 2008). Due to the relative novelty of the phenomenon, this
research is bound to cover new ground.
15
How exactly may higher education be disrupted? In the traditional university, tertiary
services are bundled together and may appear inseparable. However, there are
currently tendencies of them being unbundled, with the most compelling cases
perhaps being curricula, other educational content and instructional delivery, as they,
in computer-speak, consist of ones and zeros. Other educational services – such as
socialisation, skills, assessment and accreditation – may pose more complex, but
perhaps not insurmountable challenges.
MOOCs are relatively new, hence previously under-researched. In the past two
years, there has been a veritable avalanche of writings, though, due to the usual time
lag of academic, peer-reviewed publications, not yet so much academic literature. It is
not sufficiently explored whether free higher education, offered by some of the
world’s most famous Ivy League universities (Harvard, MIT, Stanford etc.) and new
online portals, will indeed disrupt higher-education-as-we-know-it.
Convergent expert interviews are suited to understand this important phenomenon
better. I am not aware that convergent expert interviews have been used in
educational research, so this may be an innovation in my research.
To sum up, (1) the research topic is novel; (2) the theoretical concept of disruptive
innovation, though not new to higher education, probably only comes to full bloom
when applied to MOOCs; (3) and finally, the method of the convergent expert
interview appears to be novel when applied to the topic.
1.4 Methodology
In chapter 3, my methodological position is detailed. To clearly anticipate it, my
research paradigm is interpretivism, my specific theoretical position is closest to
symbolic interactionism (although I would not totally disavow hermeneutics and
phenomenology, being quite eclectic and normally preferring a Foucaultian toolbox
approach), the specific methodology is qualitative, and the research method is
interviews, more specifically convergent expert interviews.
O’Donoghue (2007, p. 32) makes a very useful distinction between two basic types of
symbolic interactionist studies: on one hand, there is a type of studies that are “aimed
16
at generating theory on the perspectives which participants hold with regard to
something”; on the other hand, there are studies “aimed at generating theory on how
participants ‘manage’, ‘deal with’, or ‘cope with’ a phenomenon”.
For the purpose of my research, I will use the first, ‘frozen in time’ approach, as for
my present purpose, the second, more longitudinal approach is not feasible. The term
“perspective” is important to understand, it refers to the conceptual frameworks
through which people make sense of the world (Woods, 1983, Charon, 2001,
O’Donoghue, 2007). Interrelated parts of the framework that make up a perspective
“consist of the participants’ aims or intentions, their strategies, what they see as being
significant for them, the reasons they give for their activity, and what they see as the
expected outcomes of their activity” (O’Donoghue, 2007, p. 39).
1.5 Chapter overview
The structure employed in this research is the usual five-chapter structure. The
Introduction provides an overview of the topic. Initially, the Literature Review was
meant to be based on non-peer-reviewed journal articles by the author (Rudolph,
2012a, b). However, much additional literature was eventually evaluated, given the
exploding academic (and journalistic) interest in the topic. Hence, this section was
written completely anew.
In the third chapter on Research Methodology, I put the method of convergent expert
interviews into a paradigmatic and methodological perspective. The interview process
is discussed and justified. The Empirical Analysis in chapter 4 is predominantly
based on analysing expert interviews. To a very minor extent, some of my personal
experiences as a MOOC student were incorporated. The Conclusions in the final
chapter synthesises the literature review and my research in relation to the research
questions and provides recommendations and an outlook.
17
1.6 Definitions of key terms
In this section, the two key terms of this project are discussed: MOOCs (including
related terms such as cMOOCs, xMOOCs, connectivism and the flipped classroom)
and disruptive (versus sustaining) innovation. The term MOOC was created by Dave
Cormier in 2008 (Cormier, 2014b). All four aspects of MOOCs as Massive Open
Online Courses are worthy of careful consideration.
Massive refers to the number of participants, some MOOCs have attracted more than
100,000 participants, and Coursera reached one million participants from all over the
world within five months of its inauguration (Lawton, Ahmed, Angulo, Axel-Berg, &
Burrows, 2013).
Open. To Cormier (2014b), openness is the most important part of the MOOC
acronym. Openness is a “political act” and about “’demolishing established barriers’
of all sorts” (Cormier, 2014b). There is a big question mark whether MOOCs are
open or not, and Siemens (2013) has gone as far to state that MOOCs are
“destroying” open education (see chapter 4). The ‘open’ aspect in MOOC could often
be considered a misnomer, as ownership of course content and platform design is
protected so that the intellectual property can be monetised in some manner (Sandeen,
2013; Siemens, 2013). FutureLearn, for instance, has the following rather non-open
copyright notice posted on its website: “The Online Content and Courses IPR is
protected to the fullest extent possible by copyright laws. All such rights are reserved”
(cited in Siemens, 2013). Different from other xMOOCs, edX provides open
educational and platform resources in the normal sense of ‘open’: not only is
admission to the courses free, the material is available free-of-charge and available for
use and adaptation (Sandeen, 2013).
Online is perhaps the least controversial aspect of the acronym. However, MOOCs
can be combined with offline education, for instance in blended learning or in the
practice (that originated in Harvard in the 1990s) of the flipped classroom where
education technology is used to impart core learning and the classroom is “the locus
for coaching, mentoring and peer interaction” (The Maturing of the MOOC, 2013, p.
13).
18
Course. MOOCs are mostly individual courses, and many of them appear to be less
rigorous than for-credit courses at universities. However, Georgia Tech’s Masters
Degree in Computer Science is a course (in the broader sense of the word) that is
available via Udacity.
cMOOCs is the term that is used for the original MOOCs that started in 2008. C
stands for ‘connectivist’, as connectivism is the educational theory by Siemens and
Downes that inspired the cMOOCs. As Cormier, Downes and Siemens are all
Canadians, this is another meaning of the c (The Maturing of the MOOC, 2013).
Connectivism is Siemens’s “theory of learning emphasizing how knowledge and
skills emerge from making connections between different domains of activity such as
experience, learning and knowledge, as well as between individuals in a social
network” (The Maturing of the MOOC, 2013, p. 13).
xMOOCs are “online versions of traditional learning formats (lecture, instruction,
discussion etc.) on proprietary specialist software platforms owned by private
enterprises” that “feature contractual and commercial relationships between
Universities… and technology providers” (The Maturing of the MOOC, 2013, p. 11).
The three US xMOOCs are Coursera, edX and Udacity, with the UK’s FutureLearn
joining this group. xMOOCs are oftentimes just known as MOOCs due to their
dominance in public perception (The Maturing of the MOOC, 2013).
With the MOOC concept clarified, our attention needs to be turned to the other key
term, disruptive innovation. Christensen distinguishes two main types of innovation,
sustaining and disruptive innovation. Sustaining innovation “makes something
bigger or better”, for instance, “universities with more college majors and better
activity centers” (Christensen & Eyring, 2011, p. xxiv). In contrast, disruptive
innovation “disrupts the bigger-and-better cycle by bringing to market a product or
service that is not as good as the best traditional offerings but is more affordable and
easier to use”, for instance, online learning (Christensen & Eyring, 2011, p. xxiv).
Disruptive innovation is initially a boon to non-consumers, and via its own sustaining
innovations, eventually becomes a threat to traditional providers and may drive them
19
to their demise (Christensen & Eyring, 2011). The two types of innovation are
illustrated in Figure 1.4.
Figure 1.4 The path of disruptive innovation
Source: adapted from Christensen & Eyring, 2011, p. 16.
1.7 Delimitations and key assumptions
Literature that was published after September 2013, was not systematically included
in chapter 2’s literature review. Hence, only selected literature that was published
after September 2013, is included in it. Some key developments that occurred after
September 2013, are discussed in the concluding chapter. The topic was previously
under-researched, but in and after 2012, the literature has exploded. Thus, many
interesting phenomena related to MOOCs had to be excluded; for instance, the long
tradition of online education and Open Educational Resources (OER); related
phenomena like Khan Academy, Peer-to-Peer University, University of the People,
Western Governors University, and some of the most recent MOOC providers such as
FutureLearn could only be touched upon in passing.
20
1.8 Conclusion
In this chapter, the research topic of critically evaluating MOOCs as a potential
disruptive innovation in higher education was introduced. The problem statement
focuses on the perspectives of experts on the research topic, that is further
differentiated into an evaluation of the MOOCs’ quality of the educational experience
and their sustainability, and to what extent MOOCs are a viable alternative to
traditional higher education and disruptively innovate higher education by opening up
opportunities for the base of the income pyramid. The research is significant and has
some innovative elements, due to the novelty of systematically evaluating MOOCs as
a disruptive innovation in the context of convergent expert interviews.
21
2. Literature Review 2.1 Introduction
The New York Times coined 2012 the “Year of the MOOC” (Pappano, 2012) and
while there was relatively little research conducted on MOOCs prior to 2012, the
topic literally exploded in the past two years. When I first started researching the
topic, there seemed to be relatively little to write about, whereas now it is all too easy
to drown in a flood of information. Hence, as explained in the introductory chapter’s
delimitations, a focused approach was used that systematically addresses the state of
the literature (especially till September 2013) with regard to my research questions.
Due to the extremely fast developments of the topic, much additional literature till
April 2014 was eventually still included, albeit in a less systematic way.
A literature review is not a compilation of summaries, but rather a critical analysis of
the available literature. Purposes of literature reviews include: identifying of gaps in
the literature, avoiding reinventing the wheel, piggybacking on other researchers,
identifying other researchers in the field for networking purposes (in addition, this is
useful for identifying potential experts for the convergent expert interviews – see
chapter 3), identifying seminal works as well as opposing views, and putting our work
into perspective (Kervin et al., 2006).
Section 2.2 provides a brief history of MOOCs, before sections 2.3 to 2.6 address the
four research questions from the perspectives of the reviewed literature. After the
quality (section 2.3) and the sustainability (section 2.4) of MOOCs has been
evaluated, they are discussed as a potential alternative to higher education (section
2.5) as well as a potential revolution and disruptive innovation, offering access to the
bottom of the income pyramid via unbundling and rebundling of educational services
(section 2.6). Section 2.7 provides conclusions and discusses the research model.
22
2.2 A brief history of MOOCs
Figure 2.1 MOOC timeline
Source: The Maturing of the MOOC, 2013, p. 12.
The cMOOCs precede the xMOOCs. After giving due credit to the cMOOCs, I
continue with the ‘Stanford education experiment’ (not to be mixed up with
Zimbardo’s Stanford prison experiment from 1971 – section 2.2.2). This notable
experiment led to two spin-offs, Coursera (section 2.2.3) and Udacity (section 2.2.4).
After introducing edX (section 2.2.5), similarities and differences between these
xMOOCs will be analysed (section 2.2.6).
2.2.1 cMOOCs
cMOOCs are based on Siemens’s philosophical concept and learning theory of
connectivism. In connectivist theory, knowledge is not confined to tacit knowledge
(know-how embodied in people) or explicit knowledge (know-that embedded in text),
it is also distributed throughout networks. Connections between participants and
learning resources are fostered and supported (Downes, 2012). “The principle of
The Maturing of the MOOC
Figure 1: MOOC Timeline
(with acknowledgement to UniversitiesUK)
The consensus points on MOOCs are their importance, popularity and expansion. There is consensus on the reasons that Universities, and learners, have for engaging in MOOCs. These are: brand extension, recruitment, educational innovation and revenue (or cost reduction) opportunity. Learners mostly report satisfaction from studying in MOOCs, and curiosity about the experience. MOOC learners are not (currently) looking for awards.
That the impacts of MOOCs on HE will be profound and enduring is another consensus point. There is disagreement over whether a MOOC-induced transformation of the education landscape would be destructive or creative, and who might be the winners and losers. The possibility exists (for a minority of authors) that a Gartner Hype Cycle3 may be at work in the MOOC phenomenon – in which case the potential impacts may be overplayed.
Controversial points about MOOCs are just as important as the consensus. The field provokes some vocal and emotive polemic. This can directly influence its trajectory, with some Universities changing tack abruptly in the wake of strong opinion. The xMOOC format, which is where recent massive growth has taken place, is the subject of intense comment and speculation in the Academy and Press. Strong commitments from top university brands, stoked by large venture capital investments, have cooked up a powerful
12
͵ The Gartner Hype Cycle is a proprietary methodology from Gartner Research, widely adopted elsewhere, to account for large fluctuations in optimism and consolidation in technology-based industries. A technology trigger leads to inflated expectations, which are then followed by heavy consolidation before the sector reaches a stable plateau. See http://www.gartner.com/technology/research/methodologies/hype-cycles.jsp �
23
connectivist learning is that the learning takes place not as a result of absorbing the
course content, but rather in using course content as the basis for conversation and the
creation of additional materials” (Downes 2012, p. 536).
Learning is about belonging to a community, “the network becomes the learning”
(Siemens, quoted in Kolowich, 2013i). Education is "a connection-forming process,"
in which "we augment our capacity to know more" by adding nodes to our personal
networks and learning how to use them properly (Siemens, quoted in Kolowich,
2013i). There appears to be a strong metacognitive component – an important aspect
of learning in a MOOC is “learning how to select content” (Downes, 2012, p. 504).
In addition to Siemens’s connectivist pedagogy, other major influences on cMOOCs
were Couros’s online graduate course (Couros, 2009) and Wiley’s wiki-based course
(Wiley & Hilton, 2009) – which, importantly, opened up university courses to
outsiders and thus transcended institutional boundaries – and the emergence of
massive open online conferences (Downes, 2012). The success of Siemens’s
conference on Connectivism made Siemens and Downes consider MOOCs as “a
longer and more involved enterprise” (Downes 2012, p. 503).
Contrary to the xMOOCs that are “arguably dominated by the ‘drill and grill’
instructional methods with video presentations, short quizzes and testing”, “cMOOCs
emphasise connected, collaborative learning and the courses are built around a group
of like-minded individuals’ platform to explore new pedagogies beyond traditional
classroom settings” (Yuan & Powell, 2013, p. 7).
In 2008, Siemens ran the first MOOC with Stephen Downes, soon to be joined by
Dave Cormier, called “Connectivism and Connective Knowledge” (CCK08). The
course eventually attracted 2,300 non-paying, non-credit participants in addition to
the 25 students who took it for credit for the University of Manitoba’s Certificate in
Adult Education (Downes, 2012; Kolowich, 2013i). The three Canadian academics
(all of whom I interviewed – see chapter 4) designed the course along explicitly
connectivist lines, thus facilitating the course by providing the students with a basic
framework rather than instructing them in a more traditional way (Downes, 2012).
Different from xMOOCs, participants were not confined to a prescribed online
24
learning platform, but are “encouraged to use their own platform (blog, photo
account, social network site) to create and/or share resources” (Downes, 2012, p.
536). As Downes recalls about CCK08:
“170 of them created their own blogs, the feeds of which were aggregated
through a tool I created, called gRSShopper…,and the contents delivered by
email to a total of 1870 subscribers... Students also participated in a Moodle
discussion forum, in a Google Groups forum, in three separate Second Life
communities, and in other ways we didn’t know about” (Downes, 2012, p.
505).
Target participants appear to be rather advanced. “What we are trying to do with a
MOOC is to create an environment where people who are more advanced reasoners,
thinkers, motivators, arguers, and educators can practice their skills in a public way
by interacting with each other” (Downes, 2012, p. 509). “The idea behind the first MOOC… was to make online instruction dovetail with the
way people actually learn and solve problems in the modern world” (Kolowich,
2013i). Siemens and his collaborators wanted "to give learners the competence to
interact with messy, ambiguous contexts… and to collaboratively make sense of that
space" (Siemens, quoted in Kolowich, 2013i). CCK08 was the first cMOOC, and
there have been other courses in Canada and in the US, which follow that tradition
(Sandeen, 2013).
25
2.2.2 The Stanford higher education experiment
The vision is: change the world by bringing education to places that can’t be reached today. (Thrun, quoted in Markoff, 2011).
In 2011, two Stanford University professors, Thrun and Norvig, both world-leading
artificial intelligence (AI) experts, conducted an experiment in distributed education.
Their course “Introduction to Artificial Intelligence” drew from Stanford’s
homonymous introductory course, offered similar materials, assignments, and
examinations, and was offered free and online to students worldwide in the fourth
quarter of 2011 (AI Class, 2011).
One difference between the Stanford University experiment and earlier MOOCs was
its redefinition of being massive, its enormous outreach. While the AI course was just
one of three experimental online courses offered by the university’s computer science
department, the AI class alone attracted more than 160,000 students around the
globe, with an age range from high school to retirees from more than 190 countries
(Markoff, 2011; Udacity Press Kit 2013). Students themselves took the initiative to
translate the class into 44 languages, and over 400 students outperformed the top
Stanford student (Bennett, 2012).
In terms of credentialing, online students did not get the coveted Stanford credits, but
were ranked in comparison to the work of other online students and received a
statement of accomplishment, indicating their completion and level of achievement
(Markoff, 2011). For grading the 160,000 students, a system running on the Amazon
cloud was used, and in place of office hours, the Google moderator service that
enabled students to vote on the best questions for the professors to respond to
(Markoff, 2011).
While Thrun was amazed by the dizzying popularity of the online course, he was even
more surprised by the students’ attendance of the actual Stanford lecture dwindling
from 200 in-person attendees to just 20 or 30 students, as these students also preferred
to watch the lectures as online videos on their own time.
26
“These are students who pay $30,000 a year to Stanford to see the best and
brightest of our professors, and they prefer to see us on video? This was a big
shock to us” (Thrun, quoted in Hsu, 2012).
Figures 2.2 and 2.3: left: Professors Thrun and Norvig; right: Professor Thrun
demonstrates Google’s self-driving cars.
Sources: Markoff, 2011; Hsu, 2012.
2.2.3 Stanford University experiment spin-off #1: Coursera
Coursera was founded by Stanford computer science professors Koller and Ng, and
they have received US$85 million in venture capital (Kolowich, 2013j). From 2013
onwards, Coursera started to offer more than three dozen college courses on a wide
variety of subjects (ranging from Greek mythology to calculus). Professors from top
US universities such as Stanford and Princeton started to teach the courses under their
university’s name and adapted their most popular courses, embedding assignments
and exams into video lectures (Free Online Courses, 2012).
In terms of assessment, multiple-choice tests are computer-scored and peer grading
is used to assess more complex work. Similar to the Stanford University experiment,
students are ineligible for college credits, but can obtain certificates of completion for
a nominal fee (Free Online Courses, 2012).
27
2.2.4 Stanford University experiment spin-off #2: Udacity
Like Coursera, Udacity is a spin-off of the Stanford University experiment. It is a
private institution of higher education founded by Thrun and Evans with the original
goal of making free online classes available to everyone. Similar to Coursera, Udacity
has received venture capital, and initially hoped to provide free education, but
generate revenues by placing their top students in top companies. As of November
2013, Udacity offered 30 courses across so-called STEM (science, technology,
engineering and mathematics) subjects (Udacity Press Kit, 2013).
Different from other MOOC providers, “Udacity selects, trains and films the
professors who teach its courses” (Ripley, 2012, p. 40). Co-founder Thrun has gone
on record that he was “personally troubled by the 90 percent dropout rate”,
emphasizing that “MOOCs will only succeed if they make normally motivated
students successful” (quoted in Lewin and Markoff, 2013, p. 3). In later sections of
this thesis, Georgia Institute of Technology’s collaboration with Udacity to offer a
Master’s degree in computer science for US$7,000 via Udacity’s MOOC platform
will be further discussed, with a similar accredited degree pursued on-campus costing
approximately six times as much (Reynolds, 2014; Thrun 2014a).
Figure 2.4 Georgia Tech’s MSc in Computer Science MOOC
Source: Thrun, 2014a.
28
2.2.5 edX
edX was founded in May 2012 by Harvard University and the Massachusetts Institute
of Technology (MIT). It followed the success of MITx, an open online learning
project, whose first course had 120,000 students enrolled in March, 2012. Thus,
everybody with the capability to do academically well enough to successfully
complete an edX course, can now get a Harvard and MIT certificate, which, as usual,
carries no credit (Harvard, MIT to offer free Web courses, 2012).
edX is different from Coursera and Udacity, as it is nonprofit and open source. MIT is
famous amongst open education supporters for having started the OpenCourseWare
initiative. edX has set itself three goals, which are (1) “to open up access to quality
education globally”, (2) improve on-campus education and (3) “conduct research to
understand more precisely how students learn” (edX Media Kit, 2013, p. 5). edX has
an explicit social agenda in which MOOCs are viewed “as the great democratizer”
and in which it is believed “that in the future, economics, social status, gender or
geography won’t be a determinate of a student’s access to education or opportunity
for success” (edX Media Kit, 2013, p. 6).
2.2.6 Comparison of the three dominant xMOOCs
The three thus-far dominant MOOCs (Coursera, Udacity and EdX) have generated
astonishing student numbers in record time. As a result, they could be compared with
other big data companies such as Facebook or Twitter, being a platform for hosting
(learning content) and having proprietary algorithms (cf. Davis, 2014). All three
major MOOC platforms have distinct access missions.
One of the most distinctive differences is the way the three providers co-operate with
university partners. At one extreme of the spectrum is Coursera, which uses a
decentralised model, where the universities are mainly responsible for faculty and
content, while “Coursera provides the platform and various instructional and
29
assessment tools, format guidelines, course development support, marketing,
enrollment, and customer and technical support” (Sandeen, 2013, p. 6).
On the other extreme of the spectrum of university co-operation is Udacity, which has
the largest degree of vertical integration. Due to their highly detailed production
methods, Udacity has completed fewer courses than the other major MOOC
providers. The third MOOC platform, edX, is somewhere in between on the
continuum. While edX “also directly contributes to course and assessment design”, it
is “perhaps to a lesser degree than Udacity” (Sandeen, 2013, p. 6).
FutureLearn, the British answer to the US xMOOCs, was founded in late 2012, and
only started offering courses in September 2013, and was thus too new to include in
this research project. Table 2.1compares the three dominant xMOOCs according to
some key criteria. The MOOCs are still a very-fast developing phenomenon, and the
below table is largely based on data available on 18 January, 2014.
30
Table 2.1 Comparison of MOOCs
MOOC Udacity Coursera edX Launch January, 2012 April, 2012 May, 2012
Number of
enrolments
1.6 million
(January, 2014)
6.1 million
(January, 2014)
2.4 million
(October, 2013)
Fee (if any) Free for basics,
$150 p.m. for
value-added
services
Free for basics,
fees for value-
added services
Free for basics,
fees for value-
added services
Number of courses 30 (January, 2014) 639 (April, 2014) › 80 (October,
2013)
Number of
institutional
partners
Collaboration is
predominantly with
professors
108 (April, 2014) 29 (January, 2014)
Course design
model
vertical integration Decentralization in-between
decentralisation
and vertical
integration
Funding Venture Capital Venture Capital MIT & Harvard
Kirshstein, 2012). If students were able to transfer some credit earned by completing
free or low-priced MOOCs, this would decrease their total cost to obtain a credential
(Sandeen, 2013).
Normally, faculty appears to be a cost item to the university. For instance, Davidson
(2014) was paid US$10,000 for developing and offering her MOOC, which she
generously decided to spend on teaching assistants and other miscellaneous expenses.
It has been estimated “that professors typically spend 100 hours to develop a MOOC,
and then eight to 10 hours each week while the courses were in session” (The
Maturing of the MOOC, 2013, p. 71). However, the amount of time spent appears to
vary significantly, as, for instance, Belanger & Thornton (2013) report on Duke
University’s first MOOC that more than “600 hours of effort were required to build
and deliver the course, including more than 420 hours of effort by the instructor”. In
the Georgia Tech online Master’s co-operation with Udacity, the cost of faculty is a
42
consideration, and Udacity was reported to hire course assistants to help Georgia Tech
instructors with both academic and non-academic tasks (Kolowich, 2013m).
To the MOOC providers and the universities, substantial cost is involved. A cost
item is often the fee that the university pays the MOOC provider for hosting the
course. For instance, in a contracting model, edX is paid US$250,000 for producing a
MOOC for a participating University, and if a course is rerun, an additional fee of
US$50,000 per every new edition of the course (Kolowich, 2013e; The Maturing of
the MOOC, 2013). For Coursera, a university estimated its cost to be $50,000 per
MOOC, not including faculty time (Popp, 2013), with MOOCs that incorporate a high
degree of design, assessment, and analytics costing “much more” (Sandeen, 2013, p.
8).
The claim has often been made that due to MOOCs’ ‘one to many’ potential, a
MOOC can reach tens of thousands of students for the cost of reaching ten students at
zero marginal cost per student after the initial investment (Rollins, 2014). However,
Rollins (2014) points to research that shows that online courses can be more costly
than on-campus courses and concludes that by “ignoring history, MOOCs inherit a
tough problem in economics of higher education”.
To complete the overview of the different costs to various stakeholders, universities
and venture capital enabled the birth of xMOOCs. For edX, MIT and Harvard
contributed $60 million in startup capital, and for Coursera and Udacity, venture
capital was used.
2.4.2 Revenue models
With much of the cost being shouldered by universities, there does not appear to be a
very clear revenue model for universities and hence no easily definable return on
investment (ROI). Rollins (2014) has expressed his lack of understanding as to “what
incentive a university has to adopt an expensive course of uncertain effectiveness that
leads to faculty revolt”.
43
However, many possible sources of income have been identified for MOOC
providers (e.g. Voss, 2013b, The Maturing of the MOOC, 2013), and some of them
have been put into practice. The following 12 suggestions for revenue generations
have been made, and as the first four are related, their application will be discussed
together.
(1) Certification and credentialling – participants pay for some kind of evidence that
skills and knowledge have been acquired. This could be a badge or certificate, or it
could also be for credit transfer for advanced standing or in the extreme case, a whole
degree.
(2) Participants may pay for human tutoring or assignment marking.
(3) Participants pay tuition fees.
(4) Secure assessments via invigilated examinations are chargeable.
In 2013, for a period totaling less than one year, Coursera announced a revenue of $1
million from ‘verified’ course certificates from its Signature Track, thus far its only
significant revenue source (Kolowich, 2014c; 2013n). If students want to pursue the
Signature Track and obtain a verified certificate, they have to decide early in the
course and pay upfront a fee ranging from $30 to $100 (Young, 2013i). edX’s
XSeries and Coursera’s “Specializations” programmes are expected to create much
additional revenue for the MOOC platforms and their university partners. The MIT
XSeries certificate in computer science is priced at about $425, and Coursera’s
“Specializations” courses will be priced between $200 to $500 (Kolowich, 2014g).
Coursera continues to offer free unverified certificates to participants who pass its
courses. Financial aid may be offered to “students who demonstrated that they could
not afford the fees but could benefit from the verified certificates” (Young, 2013i). In
a dramatic development, Udacity recently started to charge for all its certificates,
discontinuing its previous practice of offering free, ‘non-identity-verified’ certificates
(Kolowich, 2014a). Udacity’s co-founder Thrun (2014b) related fee payment to
credential acceptance:
“We know that many of our hardworking students can’t afford to pay for
classes. At the same time, we cannot hope that our certificates will ever carry
great value if we don’t make this change.”
44
Udacity’s new ‘full’ courses cost $150 per month and, in addition to the courseware
(that will continue to be available for free), include contact with coaches, project-
based assignments and job-placement services (Kolowich, 2014a; see Figure 2.6).
Figure 2.6 Udacity’s free versus subscription model
Source: Udacity, 2014.
Georgia Tech’s co-operation with Udacity in the Master of Computer Science
programme was discussed earlier. The low tuition of $7,000 puts“a top-ranked
computer-science program at a price point more comparable to a typical community
college” (Young, 2013d). Course revenue is split 60-40 between Georgia Tech and
Udacity, and the partners hope to bring the economies of scale that MOOCs offer to a
postgraduate degree by aiming to enrol 10,000 students over the next three years
(Young, 2013d; Kolowich, 2013m).
However, attempts to charge for standalone MOOCs have occasionally been
unsuccessful. In 2013, the University of Texas at Austin (UT) conducted a
psychology MOOC, planning for up to 10,000 students and charging $550 per non-
UT enrollee. 125 people were engaged to produce and administer the course, but
fewer than 50 non-UT students enrolled (Rollins, 2014).
(5) Data mining, employee recruitment and applicant screening – companies pay for
access to student performance records.
45
(6) Selling the MOOC platform to enterprises to use in their own training courses.
(7) Third-party sponsors pay for selected courses. AT&T has set a precedent by
supporting the Georgia Tech Master’s programme with $2 million, and it aims to
create a pipeline of qualified applicants (Young, 2013d).
(8) Advertising could serve in similar ways like elsewhere on the Internet.
(9) Cross-selling or up-selling more advanced or certified course offerings. The
concept of freemium, where something is given away for free, but premium services
command a premium, which is used in many industries, is also being considered for
application to the marketing and recruitment of students (Australian Trade
Commission, 2012).
(10) Donations from online students and alumni (Free Online Courses, 2012).
(11) Charging universities. There are several models in place. One of edX’s models is
that edX receives US$50,000 of revenue from any course, and a 50% split thereafter
(Kolowich, 2013e).
(12) Content licensing and providing more conventional online courses (Kolowich,
2014c; 2013m). Coursera has begun to work with public universities in the US to use
its platform for “guided” or “adopted” non-MOOC online courses, charging $3,000
for development plus an additional per-student fee (Kolowich, 2013m). These courses
carry credit and students are enrolled with the universities and pay tuition. When
universities license course content from one another, Coursera is also getting a
percentage of the license fee (Kolowich, 2013m).
The provision of a platform for non-MOOC online courses sees MOOCs like
Coursera in potential competition with LMS providers(like Blackboard,
Desire2Learn and Instructure) and perhaps even with textbook publishers (like
Pearson and McGraw-Hill) (Kolowich, 2013m). “In the long term, the fate of
Coursera and Udacity's ambitions may depend on how well their platforms and
content work in a non-MOOC context” (Kolowich, 2013m).
46
2.4.3 MOOCs as marketing devices
The idea of MOOCs as a marketing device is linked to our discussion in the previous
section on cross- and up-selling as well as the freemium concept. MOOCs are part of
marketing strategies of universities, signaling, in conjunction with social media
strategies, “futurism and reinvention” (Bogost, 2012).“In cases like Caltech and
University of Pennsylvania, who have together invested some $3.7 million in
Coursera, they are buying a more explicit and long-term version of that advertising.
Likewise, that's how MIT and Harvard see edX” (Bogost, 2012).
2.4.4 Are MOOCs sustainable?
It is difficult to predict whether the current dominant players will be sustainable. edX,
because of the institutional support of Harvard and MIT and these universities’
substantial financial and human resources, may appear most sustainable. Coursera has
the most decentralised model (see section 2.2.6) and thus appears to be most
dependent on the support of its university partners and their professors, as the returns
for them have been “largely intangible” (Kolowich, 2014c). However, as a big data
company and because of economies of scale, it may be sustainable in any event.
Udacity’s sustainability may depend on the success of co-operations with industry and
university partners and its certification programmes.
However, it would be erroneous to equate MOOCs with the three dominant US
xMOOCs. There have been MOOCs before them, and there is absolutely no reason
for them to disappear entirely, even if the three existing major players turned out not
to be sustainable. The sustainability discussion is closely linked with the next research
question, which is whether MOOCs offer a viable alternative to traditional higher
education.
47
2.5 Do MOOCs offer a viable alternative to traditional higher education?
“College Is Dead. Long Live College!”
(Time's “Reinventing College” issue)
Udacity’s co-founder Stavens thought “the top 50 schools are probably safe" (quoted
in Carlson & Blumenstyk, 2012). During the dizzying heights of MOOC hype in
2012, Stavens’s colleague Thrun predicted that fifty years into the future, only ten
universities would remain, one of which would be Udacity (Leckart, 2012). Thrun
was much criticized for this and similar statements, but he was certainly not alone,
with the president of Northeastern University, Aoun, saying that with the advent of
MOOCs, “we're witnessing the end of higher education as we know it” (quoted in
Carlson & Blumenstyk, 2012).
However, there is much consensus that “MOOCs do not immediately threaten the
continuation of traditional Higher Education in campus universities offering face-to-
face teaching” (The Maturing of the MOOC, 2013, p. 18). Generally xMOOC
providers are aware of their dependency on traditional education and make
conciliatory statements such as: “MOOCs won’t replace universities, but rather
enhance the quality of education by incorporating blended learning” (edX Media Kit,
2013, p. 6). MOOCs are originally isolated courses that do not amount to a degree.
There have been some potential breakthrough developments like the Georgia Tech
Master’s programme as well as Coursera’s “Specializations” programme and MIT’s
XSeries, amongst others, that obviously aim at addressing this shortcoming.
At present, MOOCs do not seem to provide a viable alternative to traditional higher
education, as universities retain their monopoly on credit-granting privileges
(Kolowich, 2013m). Rather, MOOC providers complement and supplement
universities. The discussion about whether MOOCs are a viable alternative to
traditional higher education is inextricably intertwined with the final research
question, which is also at the heart of the problem statement, which will be discussed
in the following section.
48
2.6 MOOCs as disruptive innovation and providing access to the bottom of the
pyramid?
This section discusses whether MOOCs are revolutionary (section 2.6.1), an instance
of disruptive innovation (section 2.6.2), provide access to the base of the income
pyramid (section 2.6.3) and to what extent, unbundling and rebundling of higher
education is occurring.
2.6.1 Are MOOCs a revolution?
I can't teach at Stanford again. You can take the blue pill and go back to your classroom and lecture to your 20 students, but I've taken the red pill and I've seen Wonderland. (Thrun, quoted in Hsu, 2012)
Thrun’s humourous quotation refers to the 1999 movie Matrix, where Neo meets
Morpheus and agrees to follow him by swallowing a red pill (instead of the blue one).
Subsequently, Neo realizes that they live in the year 2199, in which humans are
fighting intelligent machines that control the Earth's surface and harvest human
bioelectrical energy. Humans are kept docile within the Matrix, a simulated reality of
the world as it was in 1999.
It is quite common to ponder whether MOOCs are revolutionising higher education
and whether MOOCs are a “high-tech engine of a transformative revolution that will
remake education as a highly engaging, open and low-cost activity” (The Maturing of
the MOOC, 2013, p. 6). Coursera’s co-founder Koller co-authored a presentation
called “MOOCs: The Coming Revolution” (Voss, 2013a), and Voss (2013b, p. 1) has
proclaimed: “This revolution is not about IT, it is about teaching and learning”. The
revolutionary concept is often linked up with the popular concept of “disruption” that
will be further explored in the next section in the context of disruptive innovation.
There is much skepticism in the literature regarding the revolutionary character of
MOOCs, with Rollins (2014) denying the existence of a “Magister ex machina
49
miracle”. Even Coursera’s Koller conceded that MOOCs cannot “really move the
needle on fundamental educational problems” (quoted in Kolowich, 2013o).
Similarly, Udacity’s Thrun self-critically remarked that a “medium where only self-
motivated, Web-savvy people sign up, and the success rate is 10 percent, doesn't
strike me quite yet as a solution to the problems of higher education” (quoted in
Kolowich, 2013m). A disenchanting collaboration between San Jose State University
and Udacity on a statistics MOOC resulted in only half of the students earning a
passing grade (which was a lower passing rate when compared to the face-to-face
version), and other courses also produced underwhelming results (Kolowich, 2013b).
Rollins (2014) critiques MOOCs from a historical perspective and asks rhetorically
why MOOCs should “fare any better than their extinct predecessors?” During the dot-
com era, universities and venture capitalists created online learning ventures that
largely went bust, and going further back to 1885, it was predicted “that mail-
correspondence students would soon outnumber students on campuses”, while later,
“radio lectures would move higher education forward by 25 years” and lavish hopes
were bestowed on “educational television” (Rollins, 2014). Kolowich (2013m) also
concludes that MOOCs may offer universities technology tools and services “that are
more helpful than revolutionary”.
2.6.2 MOOCs as disruptive innovation?
In their book The Innovative University, Christensen & Eyring (2011) view online
learning as a disruptive innovation in higher education (see section 1.6). As the focus
is shifting from professors’ credentials and institutional prestige to what students
actually learn, universities “must… master the disruptive technology of online
learning and make other innovations” (Christensen & Eyring, 2011, p. xxvi). For
universities, the “challenge is to change in ways that decrease its price premium and
increase its contributions to students and society” (Christensen & Eyring, 2011, p.
396).
In 2011, MOOCs had not been highly on the global agenda. However, in 2013,
Christensen has specifically commented on MOOCs as a disruptive innovation.
50
Horn and Christensen (2013) argue that the “question is not just whether MOOCs are
going to disrupt traditional education, but how”. They contend that MOOCs “bear the
early hallmarks of a disruptive innovation” as they serve non-consumers, march
upmarket (through improved quality) and redefine quality – in “the current university
system…, most faculty are rewarded for the quality of their research – not for the
quality of their teaching…; in the future, courses might be offered based on employer
demand, not faculty research interests” (Horn & Christensen, 2013).
Horn and Christensen (2013) characterise it as unusual – by disruptive innovation
standards – that it is some of the world’s most famous universities that lead the
MOOC movement. They point to Harvard and MIT being behind edX; Stanford’s
groundbreaking Artificial Intelligence course having morphed into Udacity; and some
Stanford professors having founded Coursera (Horn & Christensen, 2013).
In perhaps self-congratulatory fashion, it is argued that universities invest in MOOCs
“because disruption theory is finally widely enough understood that astute leaders
know how to identify and chase opportunities early” (Horn & Christensen, 2013). The
authors laud the involved universities for setting up “an autonomous business model
with different resources, processes, and priorities” and point to the useful leveraging
on the brands, as being associated with the likes of Harvard, MIT, and Stanford
signals quality (Horn & Christensen, 2013).
Horn and Christensen (2013) see MOOCs evolving into a “scale business” and
“facilitated network model” that optimises and individualises learning opportunities
for millions of students. They appear to seriously consider the possibility of the
traditional degree becoming “irrelevant”. When “learning becomes a continuous, on-
the-job learning process”, then “the need for customization will drive us toward just-
in-time mini-courses” (Horn & Christensen, 2013). This kind of transformation would
be akin to General Motors’ bringing forth choice and variety in replacing Ford’s
Model T (Horn & Christensen, 2013).
When describing the current situation of higher education, Christensen uses the
metaphor of the crevasse, a deep crack in a glacier.
51
“I think higher education is just on the edge of the crevasse. Generally,
universities are doing very well financially, so they don’t feel from the data
that their world is going to collapse. But I think even five years from now
these enterprises are going to be in real trouble” (Christensen, quoted in
Howe, 2013).
Christensen anticipates that "[s]ome” universities will “survive”, with most evolving
into hybrid models, “in which universities license some courses from an online
provider like Coursera but then provide more specialized courses in person"
(Christensen, quoted in Nisen, 2013). Even Harvard Business School does not teach
accounting anymore, as they use Brigham Young University’s (BYU)
“extraordinary” online accounting course (Christensen, quoted in Howe, 2013).
Friedman (2013b) comments on the latter aspect: “When outstanding becomes so
easily available, average is over” (Friedman, 2013b).
Christensen’s perspective that MOOCs promote an imminent process of disruptive
innovation and will cause pain to existing Higher Education players is shared by a
significant minority of academics (The Maturing of the MOOC, 2013). The disruptive
innovation has also been viewed in a different way than Christensen. Lawton and
Katsomitros (2012) have regarded it, amongst other things, as the shifting of costs
from students to institutions and future employers. Yet another aspect of MOOCs may
be an innovation in the process of knowledge creation, and them providing a “testing
ground for knowledge growth in a distributed, global, digital world” (The Maturing of
the MOOC, 2013, p. 47).
.
“They are a departure from the traditional brick and mortar Universities and
from the ‘walled gardens’ of conventional learning management systems,
insofar as they share the processes of knowledge work, not just the products.
Facilitators share their sense making habits and their thinking processes with
participants” (The Maturing of the MOOC, 2013, p. 47).
Disruptive innovation by MOOCs has also been described as a potential catalyst for
moving us “from an instruction paradigm to a learning paradigm”:
52
“By leveraging the vast resources available via the Internet and by using the
technology available today through the use of multimedia, instructional
design, automated assessment and web-based faculty-student interactive
strategies, the classroom experience is being re-created and high-quality
learning is now available to those individuals who might not otherwise have
access or the financial wherewithal – here and around the world” (DiSalvio,
2012).
The majority of academic observers does, however, appear to not regard MOOCs as a
disruptive innovation. Levin, Coursera’s new chief executive, reiterated that his
company’s MOOCs should be thought of as “additive to what universities are doing,
not disruptive” (quoted in Kolowich, 2014b). MOOCs may be disruptive only in
appearance, but they could end up not disrupting anything and rather being a teaching
aid in traditional, credit-bearing courses (Kolowich, 2013o). “As you go in the belly
of the beast, you will run into this brick wall every single time”, says Horn, a co-
founder of the Clayton Christensen Institute for Disruptive Innovation (quoted in
Kolowich, 2013m). Horn also commented on the MOOC providers’ resemblance with
technology vendors like Blackboard and Pearson that “are not disruptive at all”
(quoted in Kolowich, 2013m).
2.6.3 Do MOOCs provide access for the bottom of the pyramid?
Perhaps the most prominent motivation among professors for teaching MOOCs, is
“altruism – a desire to increase access to higher education worldwide” (Kolowich,
2013h). This motivation can also be found amongst the innovators of the first MOOC
like Downes:
“I want and visualize and aspire toward a system of society and learning
where each person is able to rise to his or her fullest potential without social or
financial encumbrance, where they may express themselves fully and without
reservation through art, writing, athletics, invention, or even through their
avocations or lifestyle… This to me is a society where knowledge and
learning are public goods, freely created and shared, not hoarded or withheld
53
in order to extract wealth or influence. This is what I aspire toward, this is
what I work toward” (Downes 2011, p. 3).
In 2014, edX announced a programme called SocialEDU in co-operation with
Facebook, Nokia (as the device manufacturer), service provider Airtel and the
Rwandan government to provide free, localized education to students in Rwanda on
affordable smart phones (Biemiller, 2014a). The mobile teaching app was to be
optimized for a low-bandwidth environment. edX’s president Agarwal explained that
“[i]mproving global access to high-quality education has been a key edX goal from
Day 1” and SocialEDU was part of the learning process how to achieve this goal
(quoted in Biemiller, 2014a).
edX’s good intentions to provide global access to higher education are shared by
Coursera in its mission statement: “We envision a future where everyone has access
to a world-class education that has so far been available to a select few” (quoted in
Kolowich, 2013l).Siemens commented on edX: “Projects like this can impact lives
around the world, for the next billion students from China and India” (quoted in
Harvard, MIT to offer free Web courses, 2012).
However, serious doubt has been cast whether the bottom of the pyramid can be
reached. Our review of MOOC participants has shown that many of them were highly
educated and that participants without access to higher education in developing
countries were highly underrepresented (section 2.3.1; G. Christensen et al., 2013).
This may be partially due to a relative lack of publicity as well as spotty Internet
connections in developing countries (Kolowich, 2013l).
More important for the difficulties of MOOCs in reaching out to the BoP may be the
link to literacies (see section 2.3.2). G. Sharma (2013) calls “the intellectual barrier”
the “elephant in the room [that] is still invisible”. He also warns against “a one-way
transfer of educational materials from the rich north to the poor south” and the
resultant “intellectual neo-colonialism” and asks rhetorically: “What magical powers
do the MOOC platforms… provide… that make all the challenges of facing students
across the world simply disappear?” (G. Sharma, 2013).
54
Altbach (2013) critically questions the MOOC phenomenon in a similar vein. There
are cultural biases in academic traditions, methodological orientations and teaching
philosophies of particular academic systems. At present, MOOCs are largely an
American-led effort, with most courses coming from Western countries, thus
potentially bolstering Western academe’s “higher-education hegemony”, and “making
it more difficult for alternative voices to be heard” (Altbach, 2013).
There are serious challenges for MOOCs to be successful even in the US. Udacity’s
co-founder Thrun commented on the failed experiment with San Jose State
University: “These were students from difficult neighborhoods, without good access
to computers, and with all kinds of challenges in their lives. … It's a group for which
this medium is not a good fit” (quoted in Schumann, 2013).
Schuman (2013), in a polemical critique, noted that MOOCs were supposed to bring
higher education to the masses, but they were not ready for the “commodified,
impersonal higher education that MOOCs offer”. She contented that not the students
were to blame for the failed San Jose experiment, but the medium.
“Successful education needs personal interaction and accountability… This
is… the same reason students feel annoyed, alienated, and anonymous in large
lecture halls and thus justified in sexting and playing World of Warcraft
during class – and why the answer is not the MOOC, but the tiny, for-credit,
in-person seminar that has neither a sexy acronym nor a potential for huge
corporate partnerships” (Schuman, 2013).
Schuman (2013) did not doubt that MOOCs were appropriate for highly educated
participants, but concluded that Thrun has involuntarily shown that higher education
cannot be automated. Carlson and Blumenstyk (2012) have predicted the perpetuation
of a two-tier system in which only the less wealthy, “less-prepared students who are
increasingly cut off from the dream of a traditional college education” will study low-
and no-cost MOOC's. The irony is that students at the base of the pyramid need face-
to-face instruction more than other students. To a university president, the “real
disruption” was “the changing demographics” in the US, with the real problem being
55
that “higher education has to repeat a whole lot of lower education” (McGuire, quoted
in Carlson & Blumenstyk, 2012).
Similarly, Bates (2013) argues that xMOOCs are likely to “increase inequality, by
undermining publicly funded education, leaving an elite of campus-based universities
for the very rich, resulting in high paid knowledge-worker employment for them, and
massive information transmission delivered to the rest, who will be confined to low-
wage service jobs because of their lack of high-level critical thinking skills”.
2.6.4 Unbundling and rebundling
With curricula increasingly becoming a commodity, MOOCs – and Open Education
Resources (OER) before them – “have opened up access to tried and tested curricula”
especially for more basic undergraduate courses (The Maturing of the MOOC, 2013,
p. 50). In the Australian Trade Commission’s (2012, p. 7) view, not only the creation
and dissemination of courses gets disrupted and unbundled, but MOOCs “may also
disrupt teaching and learning as new pedagogy is created around delivering education
at massive scale” and “how assessment is conducted by incorporating machine
learning and peer-to-peer assessment models”.
As a consequence of this unbundling, the role of the faculty may become
disaggregated. While for centuries, “postsecondary teaching has been vertically
integrated: identifying a subject area, designing a course, sourcing content, organizing
content, determining learning outcomes, designing exercises and assessments,
teaching the course, scoring assessments, and assigning final grades”, the role of
faculty may “become one of curator of information and mentor to students” if other
aspects are unbundled and outsourced to experts (Sandeen, 2013, p. 8). Figure 2.7
shows that university education is a rather complex ‘bundle’.
56
Figure 2.7 College: A Packaged Bundle
Source: Staton, 2012, p. 5.
A report by Pearson regarded unbundling as a threat and rebundling as an opportunity
as “those who rebundle well will find they have reinvented higher education for the
21st century” (quoted in The Maturing of the MOOC, 2013, p. 50). The rebundling
may lead to “new network lock-in models” (Siemens, 2013).
2.7 Conclusions and Research Model
In conclusion, MOOCs may not be a revolution or a disruptive innovation, as they are
but a chapter in the history of online learning. Christensen may have been on the right
Draft: Please do not cite without permission from the author.
succeeding within, and adding value to our society. The bundle is outlined here in the below
diagram.
Figure 1: College: A Packaged Bundle
5
57
track in his earlier publication (Christensen & Eyring, 2011) when he characterised
online learning (rather than MOOCs) as the disruptive innovation.
In section 2.6.1, Christensen’s characterisation of disruptive innovation as serving
non-consumers, reaching upmarket and redefining quality was referred to. However,
as already shown in section 2.3.1, non-consumers are not normally reached by
MOOCs – at least not yet. It is also suspected that rather than “reaching” upmarket,
the MOOCs’ early adapters were university instructors and other highly-educated and
tech-savvy participants and they may move now ‘downmarket’, which in any event, is
the stated intention in all xMOOCs’ access missions (see section 2.2.6). Christensen’s
third characteristic is the redefinition of quality. As we saw in section 2.3, the quality
of MOOCs, when compared to traditional higher education courses, is unfortunately
questionable, especially in terms of assessment and the value of their credentials.
Christensen’s own definition of disruptive innovation could thus be used against his
characterization of MOOCs as such. xMOOC providers cannot be faulted for their
publicly stated good intentions of providing access to the bottom of the pyramid, and
there are some laudable initiatives. However, they appear to be in the process of
further diluting the openness of their business and educational model. The content
(curricula, video lectures etc.) of the for-profit providers (Coursera and Udacity) was
never open. While admission was originally open as in free, there has been a tendency
of reducing the freely available offerings further.
The MOOC phenomenon is still extremely new and undergoing rapid changes. As a
consequence, the judgment in the literature is still highly ambiguous. The core of the
research model consists of the MOOC concept, with all of its components (Massive,
Open, Online, Courses) being important aspects, and the concept as a whole related to
the four research questions.
As the bidirectional arrows show, there are inter-relations between the MOOC
concept and all four key concepts that are expressed in the research questions: (1)
quality, (2) sustainability, (3) viable alternative to traditional higher education, and (4)
disruptive innovation. All these concepts come with a question mark, as they are all
contested. It was tempting to also attach question marks to all four MOOC
58
components, but this was already implicit in the initial definition and discussion of
MOOCs in section 1.6 and thus did not need to be re-emphasised.
There are also bidirectional relationships between all research questions. The question
of high quality is related to all other concepts, as can be briefly described in the
following three considerations. (1) If the MOOC concept does not deliver a quality
learning experience, it may not be sustainable. (2) If it does not deliver a high quality,
it would disqualify itself as a viable alternative to traditional higher education. (3)
There is also a relationship between quality and disruptive innovation, though
disruptive innovation is normally defined as something that just needs to be 'good-
enough'.
Similarly, the relationships between the other concepts could be elaborated on. For
instance, MOOCs cannot be a disruptive innovation if they do not constitute a viable
alternative to traditional higher education. And if MOOCs are not sustainable, then
they would not be much of a disruptive innovation. So, in the final analysis, all five
concepts are interrelated with each other.
The core of the research model (Figure 2.8) is surrounded by the three sources of data,
(1) literature review, (2) expert interviews and (3) participant observation. As is
explained in chapter 3, participant observation refers to my sampling of some
MOOCs and is not a major method in this research.
59
Figure 2.8 Research model
Source: self-developed
It will become clearer in chapter 3, that in this qualitative research, a circular model of
process and theory is applied (see Figure 3.1). This circularity further extends to the
structure of the literature review and the analysis of the empirical data. They
continued to influence each other in a circular, rather than in a linear, fashion during
the process of presentation.
60
3. Methodology and methods 3.1 Introduction and research design
This chapter designs appropriate methodology and methods to explore the research
questions. My research paradigm is interpretivism, my specific theoretical position is
closest to symbolic interactionism (although I would not totally disavow
hermeneutics and phenomenology, being quite eclectic and normally preferring a
Foucaultian toolbox approach), the specific methodology is qualitative, and the
research method is interviews, more specifically convergent expert interviews.
When constructing a research design, the eight components in Figure 3.1 should be
considered. Briefly, the goal of the study is description (as opposed to testing of
hypotheses or theory development). There is not much of a MOOC-specific
theoretical framework, as MOOCs are a rather new phenomenon and the approach
adopted here would tend to discover a grounded theory, which, however, goes beyond
the scope of this study. In terms of a general educational theoretical framework, I am
strongly influenced by Biggs and Tang’s (2011) constructive alignment approach in
implementing outcomes-based education. The relationship between the research
questions (section 1.2) and the interview guide will be discussed in section 3.4.1.
Empirical material (especially the transcribed expert interviews and my own
observations based on MOOC sampling) was collected voraciously, and not
everything could be used within the humble confines of this project. Methodological
procedures are described in section 3.4. Their degree of standardization and
control is best described as loose, which is appropriate for this study, as MOOCs are
a relatively new field and theoretical analyses of them are relatively underdeveloped
(Miles & Huberman, 1994; Flick, 2009). No goals of generalisation are pursued due
to the exploratory nature of the study.
I discuss the interpretivist research paradigm and symbolic interactionism in section
3.2 and justify the appropriateness of qualitative research methodology within a
circular model in section 3.3. In section 3.4, the six steps of the interview process are
discussed and the eight expert interviewees are introduced. In section 3.5, participant
observation in cyberspace as a secondary method is reflected upon, before quality
61
(section 3.6) and ethical considerations (section 3.7) of this qualitative research study
are discussed. The conclusion provides a brief summary of the methodological
approach and links it to the empirical analysis in chapter 4.
Figure 3.1 Components of qualitative research design
Source: adapted from Flick, 2009, p. 133.
3.2 Interpretivist research paradigm and symbolic interactionism
My research is paradigm-guided. In the context of education research, a paradigm is
a very broad term that encompasses assumptions about the social world, elements of
ontology and epistemology, theory and philosophy, methodology as well as methods,
and techniques and topics (Punch, 1998; Denzin & Lincoln, 2011). The research
paradigm provides guidelines that connect theoretical paradigms to strategies of
inquiry and also to methods for collecting empirical materials (Denzin & Lincoln,
2011).
62
In the interpretivist paradigm, individual and society are inseparable. All human
action is meaningful and has to be interpreted within the context of social practices.
There are four major assumptions of interpretivism: (1) everyday activity is the
building block of society; (2) there is always some autonomy in everyday activity; (3)
everyday activity nearly always involves social interaction; and (4) it involves
negotiation of meaning, leading to a modification of our views (Blackledge and Hunt,
1985, O’Donoghue, 2007).
Symbolic interactionism, in the words of Blumer (1969, p. 2), has three premises:
“The first premise is that human beings act toward things on the basis of
meanings that the things have for them… The second premise is that the
meaning of such things is derived from, or arises out of, the social interaction
that one has with one’s fellows. The third premise is that these meanings are
handled in, and modified through, an interpretative process used by the person
in dealing with the things he encounters”.
Another central assumption is the Thomas theorem, according to which a situation is
real in its consequences, as long as a person defines a situation as real (Flick, 2009).
There is a relationship between theoretical positions, methods of data collection and
interpretation, and field of application. In Table 3.1, relevant aspects to this research
project are highlighted in bold. This will be elaborated in the following sections.
63
Table 3.1 Research perspectives in qualitative research
Approaches to
subjective
viewpoints
Description of the
making of social
situations
Hermeneutic
analysis of
underlying
structures
Theoretical
positions
Symbolic
interactionism
Phenomenology
Ethnomethodology
Constructivism
Psychoanalysis
Genetic
structuralism
Methods of data
collection
Semi-structured
interviews
Narrative
interviews
Focus groups
Ethnography
Participant
observation
Recording
interactions
Collecting
documents
Recording
interactions
Photography
Film
Methods of
interpretation
Grounded theory
coding
Content analysis
Narrative analysis
Hermeneutic
methods
Conversation
analysis
Discourse analysis
Genre analysis
Analysis of
documents
Objective
hermeneutics
Deep hermeneutics
Fields of
application
Biographical
research
Analysis of
everyday
knowledge
Analysis of life
worlds and
organizations
Evaluation
Cultural studies
Family research
Biographical
research
Generation
research
Gender research
Source: adapted from Flick, 2009, p. 457.
64
3.3 Qualitative research methodology and the circular model
3.3.1 Qualitative research methodology
In the 1960s, social scientists like Glaser and Strauss (who coined the concept of
grounded theory) concluded that grand theories like Parsons’s system theory, “which
were originally meant to explain more or less everything”, “ended up explaining
almost nothing on the level of everyday phenomena” (Flick, 2009, p. 49). The
grounded theory approach implies that theories “should not be applied to the subject
being studied but are ‘discovered’ and formulated in working with the field and the
empirical data to be found in it” (Flick, 2009, p. 91).
The model of the process in grounded theory includes theoretical sampling,
grounded theory coding, and writing the theory (Flick, 2009). In later sections in
this chapter, it will become clear that I am not adopting the grounded theory process
in any of its more purist forms. The reflexivity of the researcher and the research is a
hallmark of qualitative research.
“Unlike quantitative research, qualitative methods take the researcher’s
communication with the field and its members as an explicit part of
knowledge instead of deeming it an intervening variable. The subjectivity of
the researcher and of those being studied becomes part of the research process.
Researchers’ reflections on their actions and observations in the field, their
impressions, irritations, feelings, and so on become data in their own right,
forming part of the interpretation” (Flick, 2009, p. 16).
3.3.2 Qualitative versus quantitative research
In this research, I am not adopting a rigid approach. Of course, some academics have
occasionally argued that qualitative research is superior to quantitative research
Below, my reasons for choosing qualitative over quantitative research are outlined,
but perceived superiority of qualitative research is not one of them.
65
(1) Originally, the focus was on decisions whether to use qualitative or quantitative
research was on philosophical and epistemological standpoints. Positivism as the
underlying epistemology of quantitative research was deemed as incompatible with
interpretivism, which is the foundation of qualitative research.
(2) However, I choose qualitative research largely for practical reasons. All three
major MOOC platforms (Coursera, edX and Udacity) collect a wealth of data on
students’ interactions and outcomes (Sandeen, 2013), and notably, researchers from
Harvard and MIT linked to edX have started to publish their analyses of the data (Ho
et al., 2014). Consequently, the main MOOC providers can be expected to continue
publishing solid quantitative research. Moreover, it may have been challenging to
convince MOOC providers to provide access to their data, as it would be
understandable that their own research departments regard them as sensitive,
proprietary, and the main source for their own publications.
(3) Also, a so-called mixed method approach (referring to a combination of
qualitative and quantitative methods) that has become quite popular in the past
decades, could have been considered. The two different sets of methods “should be
viewed as complementary rather than as rival camps” (Jick, 1983, p. 135). However,
Flick (2009) observes that developing “really integrated qualitative/quantitative
methods of data collection or data analysis remains an unsolved problem” (p. 30). In
any event, for a humble research project like the present one, it did not seem to be
advisable to mix qualitative and quantitative methods, as it would have increased the
complexity to an unmanageable extent.
(4) Flick (2009, pp. 32-33) suggests that a “decision for or against qualitative or for or
against quantitative methods… should be determined by the appropriateness of the
method for the issue under study and the research question”. The problem statement
and the research questions of this research project are about the perspectives of
academic experts with varying exposures to MOOCs. Consequently, qualitative
research is more appropriate than quantitative research for this kind of study.
66
3.3.3 A circular model of the qualitative research process and theory
A qualitative model of research process and theory must not be mixed up with the
linear model of quantitative research (theory, hypotheses, operationalisation,
sampling, collecting data, interpreting data, validation). The circular interlinking of
empirical steps (see Figure 3.2) is more appropriate for the discovery process and is
sensitive to the epistemological principle of verstehen (Flick, 2009). The circularity of
the qualitative research process forces the researcher to constantly reflect on the
whole research process. The research process continuously constructs reality. “The
central part reserved for the interpretation of data (compared with their collection or
the a priori construction of elaborated designs) takes into account the fact that text is
the actual empirical material and the ultimate basis for developing the theory” (Flick,
2009, p. 94).
Figure 3.2 Linear versus circular models of process and theory
Source: adapted from Flick, 2009, p. 95.
67
3.4 The six steps of the interview process
There are six steps with regard to interviews (Gaskell, 2000) that will be
reconstructed in this section. After (1) linking the research issue and interview
questions, (2) the choice of the convergent expert interview as a method is justified.
(3) The sample frame of the eight expert interviewees is discussed, before (4)
conducting, (5) transcribing, (6) coding and interpreting interviews are explored in
theory and application.
3.4.1 Links between research issue and interview questions
The first step of the preparation of a topic guide, or as briefly mentioned above, a
set of data collection questions (that are closely related to the guiding questions and
the general research question that were spelt out in section 1.2). In the previous
section (3.3), I highlighted the circular approach to the qualitative research process
that also applies to my interview questions. I emailed all questions to my
interviewees, but explained to them that we would not necessarily have to go through
all the questions and could also discuss other things that were of interest to them and
they thought to be of importance.
In Tables 3.2 to 3.5, I disclose the complete list of initial interview questions. In the
rightmost column in each table, the numbers refer to the initial interview questions
and indicate the eventual focus which was determined via the perspectives of the
expert interviewees. Bracketed numbers indicate issues that were not deemed as
important as issues with numbers without brackets, and omitted numbers showed a
relative lack of interest by the experts. The perspectives of the experts in turn had a
great influence on the final structure of the literature review in chapter 2 that is largely
aligned with the structure of the empirical analysis in chapter 4.
68
Table 3.2 Constructive alignment of research question 1 and interview questions
Research Question Initial Interview Questions Final
focus
(1) What, according to experts, constitutes a high quality of educational experience in the context of the MOOCs?
(1) Who are MOOC participants / students?
(2) Why do these participants / students enroll in these courses?
(3) What are MOOC participants’ persistence rates (drop out / completion rates)
(4) What skills and knowledge are needed to be successful in a MOOC?
(5) Who are the facilitators?
(6) What kind of training do these facilitators or tutors receive?
(7) What are the characteristics of effective facilitators or tutors?
(8) How do MOOC students perform compared to traditional students enrolled in the same course?
(9) Are MOOCs effective for all types of learners?
(10) How do instructors ensure quality learning experiences in MOOCs with multiple facilitators and thousands of students?
(11) What type of content is best taught in a MOOC, if any?
(12) Is the assessment valid and reliable (issue of cheating)?
(13) Accreditation and Employer Recognition
1-3, 4, (8), (9), (10), 11, 12, 13
Source: self-developed.
69
Table 3.3 Constructive alignment of research question 2 and interview questions
Research Questions Initial Interview Questions Final focus
(2) To what extent are MOOCs sustainable (in the perspectives of experts)?
(14) Who is paying for MOOCs (their instructors, platform, facilitators, and tutors)?
(15) How much do MOOCs really cost?
(16) What is the ROI for institutions providing MOOCs?
(17) For MOOCs offering certificates, how will they be received by employers?
(18) What are the major differences and similarities between MOOCs and traditional, credit-bearing online coursers offered in degree and certificate programs?
(19) When is a MOOC too big?
(20) What incentives are there for instructors to teach MOOCs?
(21) Do MOOCs take more time / effort for instructors to teach and students to learn?
14-15, (16), (17), (18), (20), (21)
Source: self-developed.
I considered Research Question 2 right from the start as the least important question,
as this thesis is an Education Research Project rather than a Business School project.
However, as higher education is a big business, I still decided to retain this question.
Interestingly, during the interviewees, the additional theme of MOOCs as a marketing
device emerged and was hence included.
70
Table 3.4 Constructive alignment of research question 3 and interview questions
Research Questions Initial Interview Questions Final focus
(3) To what extent do the MOOCs offer a viable alternative to traditional higher education, in the perspectives of experts?
(22) Or are the benefits of the MOOCs illusory, as they in reality harbour undesirable and inappropriate behaviours?
(23) How about the issue of access?
(22), 23
Source: self-developed.
The focus of the experts was on the research question, and the issue of access was
explored together with the issue of the bottom of the pyramid (see Table 3.5).
Table 3.5 Constructive alignment of research question 4 and interview questions
Research Questions Initial Interview Questions Final Focus
(4) In the perspectives of experts, to what extent, if at all, are MOOCs revolutionising higher education, a disruptive innovation and/or are opening up opportunities for the base of the income pyramid?
(24) Are MOOCs revolutionising higher ed? Are they a tipping point for higher ed?
(25) Will MOOCs lead to unbundling and rebundling?
(26) Or will MOOCs force many Higher Education players to radically transform themselves, or die, and a chaotic rout of the sector is in prospect (The Maturing of the MOOC, 2013)?
24, 25, 26
Source: self-developed.
Table 3.6 shows how chapters 2 and 4 are related to the research questions and the
above-mentioned 26 initial interview questions.
71
Table 3.6 Constructive alignment of sections in chapters 2 and 4 with research
two Australian Professors, three Canadian academics who were all involved in the first MOOC ever; one highly distinguished Canadian open learning expert
Failed to return emails
2 US See above discussion
Expressed regret at being too busy
5 US See above discussion
Source: self-developed.
Generally, when interviewing experts, it needs to be considered that they are very
busy people. As a result, I asked for an hour of their time and kept to the time frame,
with the exception of three cases – the interview with Dr Saxberg was cut short to
approximately 25 minutes because of his demanding meeting schedule, and the
interviews with Professor Jackson and Dr Tan lasted approximately 90 minutes, as
they were in no hurry to end the interview.
77
3.4.4 Introduction of the eight expert interviewees
This section introduces the eight experts in chronological order of interviews
conducted.
(1) Dr Bror Saxberg
Source: Saxberg, 2014.
Dr Bror Saxberg is the Chief Learning Officer of Kaplan Inc., USA, where he is
responsible for the research and development of innovative learning strategies,
technologies and products across Kaplan's range of educational offerings as well as
the overall academic standards (Kaplan Leadership, 2014). Dr Saxberg had previously
served in senior management positions for K12, Inc., Knowledge Universe, and DK
Multimedia. His educational qualifications include a B.A. in Mathematics and a B.S.
in Electrical Engineering from the University of Washington; an M.A. in Mathematics
from Oxford University; a Ph.D. in Electrical Engineering and Computer Science
from M.I.T.; and an M.D. from Harvard Medical School (Kaplan Leadership, 2014).
78
(2) Professor James Guy Jackson
Source: Jackson, 2014.
Professor James Guy Jackson is Emeritus Professor of Southern Cross University
(SCU). He was the foundation dean of the Law Faculty at SCU, Chair of the
Academic Board and Professor of Law. His immediate past responsibilities were as
VP Academic, Kaplan Asia Pacific. He has also held positions at Darling Downs
Institute of Advanced Education (DDIAE), and at Wollongong and Bond Universities
(Jackson, 2014). Jim Jackson is a trained AUQA/TEQSA auditor and was on the first
TEQSA panel to review an Australian university, and he has written widely on higher
education law and other fields of law. His academic qualifications include BCom
(UNSW); LLB (UNSW); LLM (Hons) (Sydney); Grad Dip in Education (Tertiary);
and PhD (Sydney) (Jackson, 2014).
(3) “Professor A”
Professor A is a Professor of Education at an Australian University and an expert on
educational technologies and authentic learning. Due to the request for anonymity, I
do not disclose more information on Professor A.
79
(4) Dr Samson Tan
Source: Republic Polytechnic, 2014.
Dr Samson Tan is an Assistant Director (Academic) at Republic Polytechnic,
Singapore. With more than 16 years of experience in teaching and training both
youths and adults, Dr Tan is presently involved in e-learning and game-based learning
projects and engages in research on education reform and teaching with technology
(Republic Polytechnic, 2014). He holds a Bachelor of Science in Physics from
National University of Singapore; a Postgraduate Diploma in Education from
National Technological University, Singapore; a Master of Business Administration
from the University of Hull, UK; and an Ed.D. from the University of Western
Australia in 2007 (Tan, 2014).
(5) Mr Stephen Downes
Source: Downes, 2014.
Stephen Downes is Program Manager, Learning and Performance Support Systems,
for the National Research Council Canada. In 2008, George Siemens and he designed
80
and taught the first MOOC. Stephen Downes was the winner of the Edublog Award
for Best Individual Blog in 2005 for his blog OLDaily and is Editor at Large of the
International Journal of Instructional Technology and Distance Learning. Associated
with many key developments in learning technology, Mr Downes, in the early 1990s,
developed Canada's first academic MUD (Multi-User Dimension); appears to have
been the world’s first 'edublogger' and is a central figure in the development of
connectivism, a network-based learning theory (Downes, 2014). He is also the author
of 'gRSShopper', an open source application designed to replace Learning
Management Systems. Stephen Downes holds BA (Honours, First Class) and MA
degrees from the University of Calgary (Downes, 2014).
(6) Sir John Daniel
Source: Daniel, 2014.
Sir John Daniel is a well-known thought leader on education. Amongst many other
leadership appointments, he was Vice-Chancellor of The Open University (1990-
2001); Assistant Director-General for Education of UNESCO (2001-04); and
President of the Commonwealth of Learning (2004-12). “His non-executive
appointments have included the presidencies of the International Council for Open
and Distance Education, the Canadian Association for Distance Education and the
Canadian Society for the Study of Higher Education” (Daniel, 2014). He has
published an incredible 350 publications and has been awarded 32 honorary degrees
from universities in 17 countries. “The three countries where he has lived and worked
81
have each recognised his contributions with national honours: France – Ordre des
Palmes Académiques (Chevalier – 1986; Officier–1991); United Kingdom – Knight
Bachelor (1994); Canada – Order of Canada (Officer – 2013)” (Daniel, 2014). He
pursued his university studies in Metallurgy at the universities of Oxford and Paris, as
well as a Master’s degree in Educational Technology at Concordia University
(Daniel, 2014).
(7) Mr Dave Cormier
Source: Cormier, 2014a.
Dave Cormier coined the term “MOOC” and was also was involved in conducting the
first MOOC, CCK08. He is a web projects lead at the University of Prince Edward
Island, Canada, co-founder and president of Edtechtalk, and president of Edactive
Technologies, a social software consulting firm (Cormier & Siemens, 2010). He holds
a Masters degree in Educational Literacies from Mount Saint Vincent University,
Nova Scotia, Canada (Cormier, 2014a).
82
(8) Professor George Siemens
Source: Penn State, 2014.
Professor George Siemens is the executive director of the Learning Innovation and
Networked Knowledge Research Lab (LINK Lab) at the University of Texas at
Arlington. Amongst other academic positions, he was professor at the Center for
Distance Education. In 2008, Siemens and Downes designed and taught the first
MOOC, and Siemens has since offered additional MOOCs that have gained popular
worldwide attention. Professor Siemens is the originator of Connectivism theory. He
holds a PhD from the University of Aberdeen, UK.
3.4.5 Conducting interviews
Six of the eight interviews were conducted via Skype (one Skype-to-phone and five
pure Skype conversations), and I recorded them with my iPhone 4 which produced a
decent sound quality. The two face-to-face interviews – which were recorded in
Kaplan Singapore’s conference room (in the case of Dr Saxberg’s interview) and in a
meeting room at Republic Polytechnic (in the case of Dr Tan’s interview) – were
recorded on the same device. Using such a machine for recording renders the
documentation independent of both the researcher’s as well as the interviewees’
perspectives and achieves a naturalistic recording (Flick, 2009).
83
None of my interviewees was averse to the interview being recorded, and in my
experience, interviewees quickly forgot about the recording, and the conversation
took place naturally. Visual recording was not deemed necessary, as the content of the
interview was the focus, and video recording can also be considered obtrusive,
reminding the interviewee constantly that s/he is being recorded.
3.4.6 Transcribing interviews
After data have been recorded on a technical device, their transcription is a fifth step
on the way to their interpretation. Some transcriptions focus so much on natural
science ideals of precision in measurement that Flick (2009, p. 299) has observed
them entering interpretive social science through the backdoor, and “the formulation
of rules for transcription may tempt one into some kind of fetishism that no longer
bears any reasonable relation to the question and the products of the research”. Overly
precise transcriptions of data absorb much time and energy that could be invested
more meaningfully in their interpretation. For the presentation of the transcripts, I
used Bruce’s (1992) criteria of manageability (for the transcriber), readability and
interpretability (for the analyst). While initially, paralinguistic utterances were
faithfully transcribed, I deleted most of them during coding because they did not add
value and also in the interest of readability of chapter 4. My omissions are
characteristed by “(…)”, emphases by interviewees are in bold, and paralinguistic
aspects such as laughter are put into rectangular brackets.
As stated in the Acknowledgements, my former degree student and her friend
thankfully undertook the arduous task of verbatim transcribing approximately eight
hours of interviews. As is a central feature of the procedure of transcription (Flick,
2009), I did a second check by going through all transcripts carefully, while listening
to the interviews.
84
3.4.7 Coding and interpreting interviews
The interpretation of data is at the core of qualitative research and also of grounded
theory research and is the final step of the interviewing process. However, in
grounded theory, both data collection and coding are considered as inextricably
intertwined with the interpretation of data (Flick, 2009). There are different
approaches to coding which I will briefly discuss before I justify the chosen process.
Coding is “the operation by which data are broken down, conceptualized, and put
back together in new ways” (Flick, 2009, p. 307).
Grounded theory coding is the central process by which grounded theories are built
from data through a process of abstraction (Flick, 2009). In Strauss & Corbin’s (2008)
approach, the first step is open coding during which concepts are identified and
developed, resulting in a list of codes and categories that are attached to the interview
transcript. Categories are more abstract than codes, but linked to them. The categories
that were identified in open coding are then further refined and differentiated in axial
codingwhich is the second step. Axial coding, very much like open coding, is
accomplished by making comparisons and asking questions, but axial coding links
subcategories to categories in a complex process that involves both inductive and
deductive thinking (Strauss & Corbin, 2008). The third step, selective coding, leads
to higher abstraction and focuses on potential core concepts. There are various
approaches to grounded theory coding (e.g. Strauss & Corbin, 2008; Glaser, 1992;
Charmaz, 2006), but they have in common that open coding is an important step; that
they base coding and analysis on a constant comparison of the empirical materials;
and that they view theoretical saturation as the end point of coding (Flick, 2009).
Qualitative content analysis is an alternative to grounded theory coding. In content
analysis, “categories are brought to the empirical material and not necessarily
developed from it” (Flick, 2009, 323). However, the categories are repeatedly
assessed against the interviews and modified where necessary. In contrast to grounded
theory coding or other approaches such as thematic coding, the goal of content
analysis is to reduce the material (Flick, 2009). Mayring (1997) proposes a detailed
procedure of qualitative content analysis (the co-called “General Content Analytic
Process Model”), beginning with the following four steps: (1) selection of material
85
from the interviews that is relevant for answering the research questions; (2) analysis
of context of data generation; (3) formal characterisation of the material; and (4)
definition of direction of analysis.
The fifth step is the theoretical differentiation of the research questions. The “research
questions of the analysis must be clearly defined in advance, must be linked
theoretically to earlier research on the issue and generally has to be differentiated in
sub-questions” (Mayring, 1997, p. 47). A sixth step, the definition of the analytical
units, is followed by a seventh that involves three types of content analysis: summary,
explication, and structuration
In summarising content analysis, the material is paraphrased. Thus, less relevant
passages as well as paraphrases with the same meanings are omitted (first reduction)
and similar paraphrases are bundled and summarized (second reduction) (Flick,
2009). In addition to summarising content analysis, there is also explicative content
analysis – which “clarifies diffuse, ambiguous, or contradictory passages by
involving context material in the analysis” (Flick, 2009, p. 327) – and structuring
content analysis, which searches for formal structures in the material (Flick, 2009).
The three concluding steps of Mayring’s (1997) approach are (8) the reassessment of
the category system against theory and material; (9) the interpretation of the results
according to the main research questions; and finally, the application of quality
criteria such as validity (see Figure 3.3).
86
Figure 3.3 Mayring’s process analysis of general content analysis
Source: adopted from Flick, 2009, and Mayring, 1983.
There are other approaches like Legewie’s (1994) global analysis. This one is less
complex than some of the others, starting with (1) a clarification of the researcher’s
background knowledge and of the research question, (2) coding the text via keywords,
refining the structure by marking central concepts or statements, (3) producing a table
of contents and alphabetically-ordered themes, (4) summarizing the text from the
viewpoints of the participants and (5) considering keywords for the entire text – all
this prior to the actual interpretation of the text.
My own approach is very much in the methodological trend of hybridisation. It is a
mix of some aspects of grounded theory coding, qualitative content analysis and
global analysis. For expert interviews, it is important that the content of the expert
knowledge is analysed and compared, and I found coding from the literature (see
chapter 2) as well as from the expert interviews useful for that purpose.
87
3.5 Participant observation in cyberspace as a secondary method
In the history of qualitative research, observation has been a popular and much-
discussed method (Flick, 2009). Denzin defines participant observation as a “field
strategy that simultaneously combines document analysis, interviewing of
respondents and informants, direct participation and observation, and introspection”
(1989, pp. 157-158). I use the term in a narrower sense, as I kept interviewing and
observation apart. While interviews and narratives make the accounts about practices
accessible, the practices themselves are only accessible through observation; and
observation may enable the researcher to find out how something actually works
(Flick, 2009). Observational methods can be classified along five dimensions: (1)
covert versus overt, (2) non-participant versus participant, (3) systematic versus
unsystematic, (4) natural versus artificial situations, and (5) self-observation versus
observing others (Flick, 2009).
(1) Strictly speaking, my research can be considered as covert, as I did not reveal that
I observed the instructors as well as the participants. Ethical implications of this
approach are discussed in section 3.7. (2) I practiced participant observation, as I
certainly became a – very minor (amongst literally thousands of participants) – active
part of the observed field. Applying Gold’s (1958) typology of participant roles to my
research, it can be said that I was “the complete participant”. (3) The observation
undertaken would fall more into the “unsystematic” category, as I remained flexible
and responsive to the processes. (4) In qualitative research, data are generally
collected in natural situations. The MOOCs that I observed certainly also fall into
that category. (5) Finally, while I observed the professors who taught the various
MOOCs and also other participants’ behaviour, the reflections that I share within this
project are largely based on my own reflexive self-observation.
Participant observation was a minor method in my research, and the observations that
I am using for the purpose of this project, were also reflected in half of my interviews.
My approach to observation was thus not highly systematic – it was more of an
afterthought, as it seemed ludicrous to write about a topic that I had not experienced
first-hand.
88
Various phases of observation have been discerned. For the purposes in the context of
this project, only three appear relevant: (1) the selection of a setting, (2) initial and
general descriptive observations, and (3) focused observations that concentrate on
aspects that are relevant to the research question (Flick, 2009). Settings selected
included courses on topics where I have prior knowledge such as business strategy
and philosophy, and I selected different providers, namely Coursera and EdX, thus
choosing both for-profit and non-profit providers. I took some notes on things that I
had previously been unaware of or which struck me as interesting. While I was
interviewing experts, they highlighted certain aspects of MOOCs, which also gave me
the opportunity to reflect on some of my observations. In this research, the participant
observation obviously did not occur in a physical environment, but on the Internet
that has become a part of many people’s everyday life (Flick, 2009).
3.6 Assessment of the quality of qualitative research
Qualitative research is unsuitable for the quality standards of quantitative research,
and how to assess quality of qualitative research appears to be a rather vexing
problem (Flick, 2009). The interweaving of illustrative quotations (which will be
made use of plentiful in chapter 4) is labeled as selective plausibilisation and is not a
solution to the problem. There are many suggestions for quality assessment in
qualitative research, but in Flick’s view, “none of them solves the problem of
adequate quality assessment” (Flick, 2009, p. 398). Below, I discuss qualitative
approaches to reliability and validity and some approaches that are specific to
qualitative research.
Reliability appears to be partially applicable in the form of procedural reliability, as
“the quality of recording and documenting data becomes a central basis for assessing
their reliability and that of succeeding interpretations” (Flick, 2009, p. 386). Also, in
case there are several interviewers (which was not the case in this research project),
reliability could possibly increase via interview training. Generally, reliability could
be enhanced if it were possible to check the original statement of the subject (via
89
access to the full transcripts) vis-à-vis the researcher’s interpretations. The more
detailed the research process is documented, the higher the reliability (Flick, 2009).
Validity is another key concept against which the quality of quantitative research is
routinely assessed. From a constructivist perspective, validation is a social
construction of knowledge whereby “the production of the data becomes one starting
point for judging their validity; and their presentation of phenomena and of the
inferences drawn from them, becomes another one” (Flick, 2009, p. 388). Validation
has been reformulated as the social discourse through which the trustworthiness of the
interpretations is established (Mishler, 1990).
A particular approach to qualitative research validation is communicative validation.
While in some contexts, ethical problems may arise “in the confrontation of
interviewees with interpretations of their statements” (Flick, 2009, p. 399),
communicative validation may be particularly suitable for expert interviews. I did
this in several stages in my research. After the checking of the transcriptions, I
emailed them to the experts, and made sure that they continued to be fine being
quoted for the purposes of the project. One expert, Downes, edited the interview and
then published it on his website. Also, after I had completed a final draft of the thesis,
I emailed it to the eight experts and asked for feedback, with special reference to
whether I may have misunderstood any of their remarks in chapter 4. Again, this
provided some additional validation.
Generally, there is a shift from validity to validation and towards “increasing the
transparency of the research process as a whole” (Flick, 2009, p. 391). Flick (2009, p.
403) argues that there is “no one right method to use in qualitative research” and that
the quality of qualitative research cannot be assessed by applying criteria. Rather,
quality in qualitative research applies to the whole research process.
Triangulation is a specific approach to quality. Amongst other things, it refers to
methodological triangulation, and more specifically, may refer to using two
qualitative methods (Flick, 2009). I do not wish to overstate the aspect of my
research in which I conducted some participant observation while sampling MOOCs,
but especially if that aspect of my research had been extended, this would have been a
90
case of methodological triangulation. Flick (2009) argues that triangulation is
justifiable when its results are “different in focus and level, which means that they
complement each other, or even contradict at first sight” (p. 450). In my research, the
sampling of MOOCS, in addition to the expert interviews, was certainly fruitful as it
produced additional insights.
3.7 Ethical considerations
Although researchers normally face ethical issues in every stage of the research
process (Flick, 2009), ethical considerations did not feature prominently in my
research. For the main method of convergent expert interviews, I dealt with high-
powered experts who were often in the public limelight. Moreover, half of the
interviewees are prominent proponents of openness and were consequently more than
happy to be mentioned by name.
It was not necessary to anonymise most experts’ contributions, and it was also
unnecessary to keep them confidential. I of course offered anonymity and
confidentiality as options in line with the ethics application that was approved by the
University of Adelaide ethics committee (H-2014-064). Only one of the experts
(“Professor A”) requested anonymity, and only in terms of her not to be identified by
name with regard to direct references to the interview.
On the topic of anonymity and confidentiality, I had some refreshing responses during
the interview, for instance from Sir John: “I don’t do confidentially [interviewer
chuckles] and anonymity, (…) I’m an open education resource kind of guy
[interviewer chuckles], so what I say can be quoted wherever you like and I’ll be fine
with whatever, so don’t worry about any of that” (Sir John, interview). Professor
Siemens’s response to being offered anonymity and confidentiality at the beginning of
the interview was comparable to Sir John’s: “Sure, yeah, absolutely, there is no need
for anonymity in this interview as far as I’m concerned” (George Siemens, interview).
When I asked again via email, Professor Siemens’s emailed response was similar:
“Sure – go ahead and mention my name. I don't need anonymity/confidentiality!”
91
The usual consideration of informed consent (participants are informed of the
potential risks and benefits before giving their consent) was applied. To use a
Hokkien term, a kiasu (literally: ‘afraid to lose’) approach to informed consent was
employed, thus reconfirming the consent of the experts not only once, but at several
stages of the research. The permission to use the experts’ names was given by them
initially in writing. During the interview, I also asked for permission to use their
names. As an additional safeguard, I provided them with the transcripts of the
interviews. Again, I provided the options of anonymity and confidentiality. Also, the
project, at this stage, is not meant for publication, and the experts will be provided
with the final copies at the time of thesis submission. If publication wholly or in parts
were to be considered at a later stage, I would certainly ask again for permission.
Other applicable ethical principles include non-maleficence (the avoidance of
harming participants), non-invasion of privacy, non-deception about the research’s
aims and beneficence (there should be identifiable benefits of the research – Murphy
Cross, J. (2011, January 28). Those who forget the mistakes of history. Internet Time
Blog. Retrieved February 9, from http://www.internettime.com/2011/01/those-
who-forget-the-mistakes-of-history/
Currie, P. & Arbour, V. (2014). Dino 101. Dinosaur Paeobiology. Coursera MOOC
from University of Alberta. Retrieved from
https://www.coursera.org/course/dino101.
Czerkawski, B., & Lieberman, M. (2012). Building Collaborative International
Learning Experiences using Social Networking Services. In T. Bastiaens & G.
Marks (Eds.), Proceedings of World Conference on E-Learning in Corporate,
Government, Healthcare, and Higher Education 2012 (pp. 2022-2025).
Chesapeake, VA: AACE. Retrieved from http://0-
www.editlib.org.prospero.murdoch.edu.au/p/41900.
Daniel, John (2014). Bios. Retrieved from http://sirjohn.ca/wordpress/?page_id=14 Daniel, J. (2012). Making sense of MOOCs: Musings in a maze of myth, paradox and
possibility. Journal of Interactive Media in Education, 3, 1-20. Retrieved from