British Educational Research Journal

Vol. 36, No. 2, April 2010, pp. 173–189

Contrasting concepts of learning and

contrasting research methodologies:
affinities and bias
Phil Hodkinsona* and Flora Macleodb
aUniversity of Leeds, UK; bUniversity of Exeter, UK
British
10.1080/01411920902780964
CBER_A_378266.sgm
0141-1926
Original
Taylor
02009
00
NG8
PhilHodkinson
000002009
[Link]@[Link]
and
&
Education
Article
Francis
(print)/1469-3518
FrancisResearch Journal
(online)

This paper analyses the conceptual significance of different methods of researching learning. Based
largely upon our own experiences, we briefly compare the use of mini-ethnography, life history,
cross-sectional surveys and existing panel survey data. We argue that there are strong affinities be-
tween each of these methods and significantly different ways of understanding the nature of learn-
ing: mini-ethnographies with learning as participation, life history with learning as construction and
both types of survey with learning as acquisition. Three things follow. The first is that decisions
about how to research learning are related to decisions about how to conceptualise and theorise
learning. The second is that there is no foolproof empirical way to adjudicate between different con-
ceptualisations of learning, though empirical evidence can and should play a significant part in in-
forming such decisions. The third is that though mixing methods (including mixing more than one
different qualitative approach) can bring advantages, the integration of the findings of mixed meth-
ods in relation to learning requires careful and sometimes difficult conceptual work.

Conceptualising learning
Our combined experience of researching learning using different methodologies
suggests that there are significant and strong affinities between different methods and
different ways of understanding learning. Phil Hodkinson has worked with a range of
qualitative methods, whilst Flora Macleod has worked mainly with quantitative meth-
ods. Underpinning these different experiences lie contrasting views on the nature of
research knowledge. Hodkinson’s views on research relativism are fairly well known,
but are not shared by Macleod, who strives for objectivity when working with quan-
titative data. We have approached learning from differing backgrounds. Macleod is a
psychologist, whereas Hodkinson draws upon more sociological literature.

*Corresponding author. Lifelong Learning Institute, E.C. Stoner Building, University of Leeds,
Leeds, LS2 9JT, UK. Email: [Link]@[Link]

ISSN 0141–1926 (print)/ISSN 1469–3518 (online)/10/020173–17

© 2010 British Educational Research Association
DOI: 10.1080/01411920902780964
174 P. Hodkinson and F. Macleod

In presenting these differences we1 are not trying to pigeonhole people (or
ourselves) conceptually according to their (our) methodologies. However, the
contrasts between us are relevant because one referee detected an inbuilt bias towards
constructivism in the paper (the other thought we were biased in favour of situated
theories) and because what follows could be read as simply another variation on
Hodkinson’s relativist position.
In what follows we attempt to analyse the central focus of different methodologies
we have used as a means of enriching understanding of the conceptual boundaries
inherent in studying learning at the time of writing. Can any methodology fully
capture something as illusive as learning when we impose meaning through the ways
data are collected and analysed? What does it mean when some methodologies make
great efforts to eliminate subjectivity and others do not? These are some of the ques-
tions that became apparent and have since preoccupied us as members of the same
mixed methodology project team. For us, these tensions have a much wider signifi-
cance for ongoing debates about research methodology and also for those about the
nature of learning. At the time of writing, neither of these debates is close to resolution
and protagonists in both often disagree even on the most fundamental issues. In open-
ing up these matters for discussion, we first say a little about the concept of learning.
Learning is a conceptual and linguistic construction that is widely used in many
societies and cultures, but with very different meanings, which are fiercely contested
and partly contradictory. Learning does not have a clear physical or reified identity in
the world. Rather, learning is a concept constructed and developed by people to label
and thus start to explain some complex processes that are important in our lives. For
many post-enlightenment writers, the ability to learn is what makes humans distinc-
tive. For such writers, learning entails purposeful activity by the learner and is thus
different from structural accounts of socialisation, the process whereby people are
moulded by their sociological context. On the other hand, behaviouristic studies of
learning assumed that learning by people was very similar to learning by animals, such
as mice in a maze. These and other different groups of writers each construct learning
in different ways. The differing concepts are often at least partly incompatible, so that
what learning is actually differs in different accounts, despite sharing the same label.
In using the term ‘learning’ in particular ways, we are constructing what learning is
and to whom (or what) it applies. Thus, from differing perspectives, organsations can
learn, families can learn, individual people can learn, animals can learn and maybe
computers can learn. One referee felt that this part of our argument was ambiguous.
Did we mean that concepts of learning differ, or that the activities to which those
concepts are applied differ? Our answer is both. For when different concepts of learn-
ing are used, the processes/activities they apply to also differ. If learning is an inher-
ently human, purposeful, cognitive activity, then whatever cats do when they
gradually discover that food arrives in a certain place and at a certain time is a differ-
ent sort of activity. It is not learning.
To substantiate this point we briefly discuss some currently common, often tacit,
root metaphors that are utilised or implied when learning is thought about. Sfard
(1998) drew our attention to the significance of metaphors in relation to learning
 Concepts of learning and research methodologies 175

when she analysed recent debates about the nature of learning as a contest between
the metaphor of acquisition and that of participation. These two metaphors, she
argued, present largely incommensurable views of what learning is. Acquisition
focuses on learning as a process whereby commodities of, say, knowledge are
acquired. It underpins many cognitive approaches to understanding learning (Mason,
2007). At its crudest, this amounts to little more than what Bereiter (2002) terms the
folk theory of learning—putting stuff (what is learned) into vessels (the human mind).
Even in its much more subtle versions, the acquisition metaphor focuses attention on
that which is learned and understands the process(es) of learning in relation to such
outcomes. It therefore tends to emphasise learning that is intentional, with an explicit
objective in view. This fits well with concerns about formal educational learning,
including how well students achieve the intended learning outcomes. Participation,
on the other hand, entails seeing learning as the undertaking of activities within a
social context, sometimes conceptualised as an activity system (Engeström, 2001,
2004), sometimes as a community of practice (Lave & Wenger, 1991; Wenger, 1998;
Wenger et al., 2002). From a participation perspective, much learning is informal—
simply an integral part of everyday activity. As such, the focus of attention switches
from cognitive processes within the person to the socio-cultural practices of living in
particular situations. Outcomes can, and frequently are, considered but as an integral
part of complex learning processes, not as an important end-point to be achieved.
Socio-cultural, participatory views of learning arose out of two perceived problems
with cognitive, acquisitional views of learning. The first was the realisation that many
people learned very effectively outside schooling (Brown et al., 1989). The second
was that there were socio-cultural reasons why some students failed to learn, say,
science or mathematics in schools (Engeström, 1984; Lave, 1988). It is important for
the argument we make here to remember that participatory views of learning have not
superceded acquisitional views. Rather, an intense debate between protagonists of
both views continues. Attempts to merge or blend these different views of learning
have all so far failed (Anderson et al., 1996, 1997; Greeno, 1997; Alexander, 2007;
Mason, 2007).
Despite the wide impact of the Sfard analysis, acquisition and participation are not
the only contenders as root metaphors of learning. There is a long tradition of seeing
learning as construction. In addition to the intellectual movement that is often termed
constructivism, which focuses mainly on the ways individual learners engage with new
knowledge, scaffolding it onto what is already known, Hager (2005) has recently
argued that workplace learning is best seen as embodied construction, arguing that
both acquisition and participation can imply learning as static, rather than as a process.
If we move beyond English as a language, other root metaphors become apparent. We
draw upon Dominicé (2000) to introduce the French concept formation ‘[which]
describes the alliance of formal and experiential learning that gives shape to an adult
life’ (p. 11). More recently, Hodkinson et al. (2008) and Hager and Hodkinson (in
press) have written about learning as becoming—a metaphor used to signify a blend-
ing of learning as participation with learning as embodied construction, in a broadly
Deweyan sense.
176 P. Hodkinson and F. Macleod

A central question for researchers is, therefore, the need to determine which of
these (or other) conflicting views of learning should be adopted. A conventional view,
drawn from science, is that any theory of learning can be tested against empirical
evidence. That is, empirical research should be used to eventually resolve the relative
merits of these competing views. For this to be achieved we would need to be able to
identify neutral, objective research methods. Our argument here is that this is not
possible, because different research methodologies each tend to favour one or other
of these competing ways of understanding learning. We support that argument by
analysing four different research approaches to this problem, two qualitative and two
quantitative. In three cases, we are reflecting upon personal experiences. In the fourth
we draw upon two recent studies published by others.

Mini-‘ethnographic’ approaches and participatory theories of learning

In the Transforming Learning Cultures in Further Education (TLC) project, Phil
Hodkinson worked with others to study learning within English further education
colleges (James & Biesta, 2007). The main investigatory method was a longitudinal
study of 17 ‘learning sites’ in four different colleges. The main tutor in each site was
an active partner in the research, who kept a dairy. In addition, there were repeated
observations of the site and repeated interviews with the tutor and a small sample of
students. We also collected questionnaire data from a wider sample of students within
these case study sites.
For shorthand purposes, we are using the term mini-ethnography to label research
like this, which focuses on a study of social life within a place or places. However, the
very term ethnography is used in many different ways. Two very common underlying
characteristics are the focus of cultural practices in a group or place and the method
of deep immersion by the researcher. This second focus leads, for example, to the use
of the term auto-ethnography, for research focused centrally upon the life of the
writer, who is, by definition, fully immersed. In this paper, we are centring on the
former characteristic—using research to search for what Wolcott (1999) terms a
cultural way of seeing, centred upon place. We use the term ‘mini-ethnographies’
partly to avoid confusion and because we know that many writers would not view the
TLC as ethnographic in a true sense.
We selected a mini-ethnographic approach in the TLC to help examine the
complex interrelationships between a wide range of factors that we predicted would
influence learning in the sites. They included:
● the positions, dispositions and actions of the students;
● the positions, dispositions and actions of the tutors;
● the location and resources of the learning site, which are not neutral, but enable
some approaches and attitudes and constrain or prevent others;
● the syllabus or course specification, the assessment and qualification specifications;
● the time tutors and students spend together, their interrelationships and the range
of other learning sites students are engaged with;
 Concepts of learning and research methodologies 177

● issues of college management and procedures, together with funding and inspec-
tion body procedures and regulations and government policy;
● wider vocational and academic cultures, of which any learning site is part; and
● wider social and cultural values and practices, for example around issues of social
class; gender and ethnicity; the nature of employment opportunities; social and
family life; and the perceived status of further education as a sector.
In making this methodological decision we were paralleling Flybjerg (2001), who
argues that case studies are the best way to study relational complexity. What may
already be obvious to readers was gradually noticed by us. By centring our investiga-
tion on the cultural practices of each learning site, we unintentionally decentred the
students. They were seen as an important part of something bigger and we had
neither time nor inclination to produce many individual student stories (but see
Davies and Tedder [2003] for one example). There were too many students in each
site for us to focus on many of them seriously and our focus was the learning site, of
which each student was part.
This centring of the study on locations has much in common with other work done
under what Sfard (1998) terms a participatory metaphor of learning. Here, the focus
is on the activities and practices of a situation where learning takes place. The lens
through which those practices are examined differs from researcher to researcher and
two common ones are the cultural and historical activity systems of Engeström
(2001) and the communities of practice of Lave and Wenger (1991) and Wenger
(1998). These studies also see learning as relational, though the nature of that rela-
tionality varies from one account to another. Lave and Wenger’s work is more
concerned with relationships between people, and a sense of belonging, whilst
Engeström is more concerned with structures, rules and procedures. In both cases,
learning is not seen as a separate process in its own right. As Lave and Wenger (1991)
write:
In our view, learning is not merely situated in practice—as if it were some independently
reifiable process that just happened to be located somewhere; learning is an integral part
of generative social practice in the lived-in world. (p. 35)

Because such accounts focus on the practices or activities through which people
participate, most see learning as embodied (Beckett & Hager, 2002). That is, engag-
ing with practice, and therefore learning, always entails more than cognitive activity.
People physically do things and react emotionally to things. Partly because of this,
and partly because of the fact that many of the participatory studies are located
outside educational institutions, there is also a major focus on what is often termed
‘informal learning’. Colley, Hodkinson et al. (2003) suggest that the often-assumed
division between formal and informal learning is unjustified. From their perspective,
one of the possible weaknesses in some participatory accounts is a devaluing of the
role of formal instruction (but see Billett, 2001) and of the value of learning in educa-
tional institutions (Brown et al., 1989; Lave & Wenger, 1991).
As the TLC work progressed, we engaged fairly deeply with this participatory liter-
ature, identifying several shortcomings in it and advancing some new thinking of our
178 P. Hodkinson and F. Macleod

own (Hodkinson et al., 2007, 2008). Our evidence suggested that it was important to
adopt participatory approaches to learning in college that did not see this as inevitably
second best to supposedly authentic everyday learning. Another weakness has already
been implied above, for very few studies of learning which adopt a participatory
perspective deal adequately with individual learners (Anderson et al., 1996). Whilst
we agreed with Cobb and Bowers (1999) that it is possible to do both, and recognis-
ing that there have been some recent attempts to do just that (Billet & Somerville,
2004; Hodkinson & Hodkinson, 2004a; Hodkinson et al., 2004), we felt that there
were some difficult conceptual problems to be overcome to pull off the full integration
of individuals into theories of learning focused on the situations where learning takes
place. Those conceptual issues are not the prime concern here. Rather, we (Hodkin-
son and Macleod) wish to point up a methodological problem.
This problem can be expressed very simply. It is difficult to give individual learners
close attention when the central focus of the study is on place. Hodkinson and
Hodkinson (2004a) partly overcame this in a study of schoolteachers’ workplace
learning, by focusing on very small school departments so that individual teacher
differences did not disappear. The decision to do this skewed their sample away from
larger departments (with more than five members of staff) and could not be replicated
in the larger TLC site studies. In the latter case, we were able to focus on the tutors
as individuals, because we had one main partner in each site. We could not do the
same for all the students. It is possible to study one or two students within their loca-
tional setting, but another problem remains. Both the Hodkinson and Hodkinson
study and the TLC marginalised individual learning outside the specific locations
being studied. This is a common problem with other participatory accounts. Lave and
Wenger (1991), for example, implicitly write about newcomers as if they arrived as
tabulae rasa.
Thus, there is a natural affinity between a mini-ethnographic, case study research
approach, focused on cultural practices in a place, and participatory ways of concep-
tualising learning, which also focus on cultural practices in a place. We will discuss
some of the implications of this affinity, after we have described three different exam-
ples of research approach towards learning.

Life history and learning as construction

Like mini-ethnographies, life histories are also case studies. But unlike mini-
ethnographies, they place the individual at the centre of investigation. Both authors
have recently been working together on the Learning Lives research project. This is
an ongoing four-year investigation of the relationships between learning, agency and
identity in people’s lives. A significant part of this project is concerned with
constructing life histories of over 100 adults, of differing ages and differing back-
ground and positions. In the Learning Lives study, as with many other life histories,
the prime unit of analysis is the individual stories of people’s lives, based upon what
they say to us in our interviews. These stories share with ethnographies the facility to
deal with complexity within people’s lives and in the patterns between different lives.
 Concepts of learning and research methodologies 179

In revealing that complexity, it is relatively easy to explore the significance of infor-

mal as well as formal learning, especially if the researcher is tuned in to look for
informal learning and progressively helps the research subjects to become aware of it
in their own lives. Another obvious strength of the method is that learning can be
seen and understood over very long time frames within the life course, whereas mini-
ethnographies tend to give a snapshot, taken over a relatively short time span.
Much life history is also good at placing the subject in a social, economic and
historical context (e.g. Antikainen et al., 1996). Indeed, for some it is this attention
to contexts that distinguishes life history from life stories (Goodson & Sykes, 2001).
However, the method is arguably less good than mini-ethnographies in unpicking and
understanding the significance of practices in particular locations for a person’s learn-
ing. That is, whereas with ethnographies, the lives of individuals outside the site tend
to disappear, or be backgrounded, so with life histories the significance of particular
situations and their learning cultures (Hodkinson et al., 2007) tend to disappear or be
backgrounded.
What this means is that there is a likely disjuncture between understanding learn-
ing as participation and a life history approach. Rather, there is an affinity between
life history and conceptualising learning as a form of construction or as formation in
Dominicé’s (2000) sense. This is because the construction metaphor centres upon
the ways in which people make sense of any learning experiences they have—the
ways in which they construct their own versions of what is being learned and/or
construct themselves through that learning. Furthermore, life histories are them-
selves constructed from the subjective perceptions of the subjects, as revealed in the
ways they tell stories about their lives and learning. As a direct consequence the
method is very good at revealing ways in which people learn through the process of
narrative construction and through reflection upon their own narrative accounts of
their lives. It is no coincidence that Alheit’s (1995; see also Alheit & Dausien, 2000)
work on biographicity (see also Tedder & Biesta, 2007) and Dominicé’s (2000) use
of research to promote self-learning, both come about through the adoption of life
history methods.
Much of the early constructivist literature on learning was inherently cognitive, but
this does not have to be the case. More recently, there has been a revival of interest
in construction from a more explicitly Deweyan perspective (Hager, 2005). Learning
is then seen as a form of embodied practice, or embodied judgement making, and life
history can be very effective in bringing out the practical, physical and emotional
dimensions of learning that are often omitted in more cognitive accounts.
Hodkinson has argued elsewhere (Hodkinson et al., 2008) that the view of learning
as embodied construction and that of learning as taking place through participation
in learning cultures are mutually complementary and that they can and should be
combined, using the root metaphor of learning as becoming. Whether this argument
is accepted or not, we would contend that both ethnography and life history entail
strong affinities with particular, but different understandings of learning: participa-
tion on the one hand and construction on the other. We now turn to our quantitative
examples.
180 P. Hodkinson and F. Macleod

Quantitative survey analyses and learning as acquisition

Alongside this life history work on the Learning Lives Project, Flora Macleod has
been using an ongoing longitudinal panel dataset, the British Household Panel
Survey (BHPS), to gain a deeper understanding of learning in the life course. Panel
surveys share with life history research a focus on longitudinal data collected from
individuals, rather than a focus on locations where learning takes place. The main
purpose of a social panel survey is to produce quantitative numerical descriptions
(statistics) about some aspects of the population that have been surveyed through
time in order to develop and test theory. Panel surveys have the capacity to follow the
same individuals through time and thus have a major role to play in understanding
social change.
Panel and cross-sectional survey research use a methodology that is set firmly
within the ambit and rules of science of which measurement is a fundamental part.
Thus, in a project that sets out to gain a deeper understanding of learning in the life
course using survey data, it is necessary to begin from a point where learning is theo-
rised, conceptualised and precisely defined so that it can be operationalised and made
quantifiable (i.e. made into a phenomenon that is measurable). Inevitably this quan-
tification of learning involves a process of reification whereby statistics (in this case
numerical descriptors of learning) become measures of a conceptualisation of learn-
ing that purports to have an objective reality. That is ‘an existence as real’ to the
extent that science is about providing ‘true’ descriptions of a real world or, at the very
least, descriptions that appear to work in practice.
Quantification of learning is possible because learning, in common with other
complex concepts, has potentially a whole range of directly observable characteristics,
which are taken to be indicative of a single underlying hypothetical (or ‘latent’) vari-
able. This means that although learning does not have a clear physical or reified iden-
tity in the world, it can be argued that at least some of its characteristics can be
observed directly and quantified. Examples include the number of times an individual
participates in formal learning events or the time spent in formal education or train-
ing. Also the ‘static’ products of learning such as the number and level of qualifica-
tions obtained can be said to be straightforward manifestations of learning that can
be counted or measured. These are the ‘factual’ material, which may safely be
collected by direct questioning. However, these directly observable characteristics of
learning are all indicative of seeing learning as acquisition, as Felstead et al. (2008)
point out. That is, they focus on either what has been learned or on identifying the
types and frequency of engagement with specific and explicit learning situations.
It follows then that when learning is conceptualised as acquisition, the limitations
imposed by survey research are not serious. But when learning is conceptualised as
participation or construction, it can neither be observed directly nor quantified very
easily as the emphasis is on subjective processes rather than objective products. The
social science solution to measuring hypothetical (or hidden) phenomena is to design
items (questions) on a survey instrument (questionnaire) that purport to be valid
measures of a phenomenon (in the context of the present discussion ‘learning’). This
 Concepts of learning and research methodologies 181

is a very difficult problem, not because surveying subjective phenomena is not possi-
ble, but because the more complex the phenomenon the more difficult it is to find
measures of demonstrable validity. One way of further illustrating the nature and
strength of the affinity between survey techniques and the acquisition view of learning
is to examine what happens when researchers use these techniques with alternative
conceptualisations of learning in mind. We do this next with two recent cross-
sectional surveys.

Moving cross-sectional questionnaire surveys beyond learning as acquisition

As has already been shown, informal learning is seen as much more significant within
participatory and some constructionist perspectives than in acquisitional views. Yet
identifying measures for informal learning is very difficult. Livingstone (2000) got
round the measurement problem by surveying what he described as ‘detectable infor-
mal learning’, which he distinguished from ‘everyday perceptions, general socializa-
tion and other tacit learning by peoples’ [sic] conscious identification of the activity
as significant learning’ (p. 2). His survey participants had to be able to recognise retro-
spectively that significant learning of one sort or another had taken place in their lives.
This meant that self-report of the learning had to be possible. But since much informal
learning would be impossible to distinguish in any discrete way from other activities
it could be neither identified by the person concerned nor measured by Livingstone’s
survey instrument. For example, learning that is tacit and taken-for-granted would
not be included in his ‘explicit informal learning’ category as it would not be recogn-
ised by his participants because it is so deeply embedded in other activities that occur
both routinely in their daily lives or at special occasions. Livingstone accepted that this
limitation would inevitably lead to ‘a very substantial underestimation of the amount
of informal learning in people’s lives’ but concluded that ‘to study informal learning
empirically, educators have to focus on those things people can identify for themselves
as actual learning’ (p. 3).
Tuschling and Engemann (2006) have argued that the very act of trying to measure
the informal and tacit is to change it from what it is to what it is not. Their argument
goes as follows: making informal learning tangible (measurable) is to bring it under
the auspices of the formal, where standardisation rather than diversity is valued as
comparison, can only be done through standardising what one is measuring. For
example, as with Livingstone’s survey instrument, in systems for accrediting prior
learning (APL) such as the APL system established in Britain since the early 1990s
and those found in France and Switzerland, the onus is on the individual to make
their capabilities visible through self-reporting a combination of their formal and
informal learning. Further, in APL systems not only does the individual have to
generate knowledge of themselves by collecting and documenting data so as to
make the invisible visible, the information they collate (profile) must also be made
recognisable and comparable across boundaries including national boundaries. This
inevitably requires universal agreement as to what constitutes the informal and tacit
rather than depending on the (subjective) position of the learner. Tuschling and
182 P. Hodkinson and F. Macleod

Engemann (2006) thus concluded that the very process of making learning visible,
recognisable and comparable is to make the local global as it involves generating tools
whereby the local has to conform to a (mutually agreed) norm.
In recent times much of the research into informal and tacit learning has been
conducted in the workplace. Explicitly in relation to workplace learning, Hodkinson
and Hodkinson (2004b) also argued that to measure informal learning was to
change its nature. In particular, they argued that much workplace learning was unin-
tended and focused on dealing with new and sometimes unforeseen situations. In
such circumstances it was not possible to predefine measures of what was learned
without changing that learning into something planned for and more formal, as in
worker development schemes, which parallel some of the effects which Tuschling
and Engemann (2006) found for APL.
Despite such problems, Felstead et al. (2008) made a valuable and interesting
attempt to survey learning as participation in the workplace. They did so by
constructing two survey instruments: one measuring acquisition learning and the
other participation learning. The latter was made up of questions about the extent to
which respondents found the activities ‘doing the job; being shown by others how to
do things; reflecting on one’s own performance; watching and listening to others; and
using trial and error on the job’ (p. 367) helped them to improve their work perfor-
mance. The researchers also developed a scale to measure the workplace environment
and the degree of employee influence and involvement at work.
Despite some clear and important advances in methodology and some valuable
findings, there are problems with this attempt to operationalise participatory learning.
The first is that it excludes much learning at work that does not directly contribute to
doing the job better, despite the fact that mini-ethnographies of learning frequently
record the contested nature of learning. Thus, for example, Colley, James et al. (2003)
show how nursery nurses not only learn how to do their job, but also wider issues of
gender and class stereotyping and inequalities. A second, more serious, problem is the
casting of acquisition and participation as different and implicitly complementary
types of learning. Thus, getting a qualification is seen as acquisitional, whereas learn-
ing through doing the job is seen as participatory. Yet from the perspective of partic-
ipation as a metaphor, learning to get a qualification is itself a participatory process,
whether in college or at work (Rogoff, 2003). In other words, in trying to measure
participatory learning, Felstead et al. (2008) have significantly changed normal views
of the participation root metaphor that is broadly seen as incommensurate with the
alternative metaphor of acquisition (Sfard, 1998; Saljö, 2003; Alexander, 2007) into
a new term for informal learning, seen as a type that is different from, but compatible
with, formal learning. The real benefit in what they have done is to quantify the extent
to which workers really value these informal learning processes. However, there
remain problems with this move. As Colley, Hodkinson et al. (2003) show, there is
no safe and generally agreed distinction between formal and informal learning.
Rather, they argue, the terms formal and informal are attributed to facets of learning
in numerous and sometimes contradictory ways. More seriously for our argument
here, in operationalising participatory learning Felstead et al. (2008) have partly lost
 Concepts of learning and research methodologies 183

sight of its fundamental distinguishing characteristic: that learning is seen not just as
a process of doing something, like a job, but of participating in a social activity system,
community of practice or learning culture. They have had to change the meaning of
participation in order to operationalise it. Neither Livingstone (2000) nor Felstead
et al. (2008) made any attempt to research learning as construction. Had they done
so, one central problem would have been the need to research learning as an ongoing
individual process—something that is very difficult in a cross-sectional survey but that
can be done using panel survey data.

Panel survey analysis and learning as acquisition

The BHPS collects data from the same panel of people at regular intervals. Thus,
such data does, at least in principle, raise the possibility of investigating learning as an
ongoing process over many years. However, when Flora Macleod worked with the
BHPS as part of the Learning Lives research project, the problem of capturing infor-
mal learning remained. Unlike the specifically designed questionnaires of Livingstone
and Felstead et al., we had to find good operationalisations of informal learning using
an existing dataset, the BHPS. This is necessary if the research is to engage with
essential aspects of either the construction or participatory root metaphors for learn-
ing. This raises the central problem of how far existing BHPS survey items can act as
proxies for informal learning? For instance, social networking could be seen as
providing the conditions for informal learning. However, in the panel dataset there
are eight items (questions) that could be proxy measures of the latent variable ‘social
networks’ the strength of which could, in turn, be taken to be an indicator of one
dimension of informal learning. The problem is that the sorts of social networks that
are more conducive to self-development and self-transformation are unevenly distrib-
uted across social groups. Thus, finding good measures of learning through social
networking for population sub-groups, a more nuanced and subtle approach involv-
ing the validation of subjective questions and testing their applicability to sub-groups
is necessary before their power as measures can be demonstrated. But all this takes
time and effort, which begs the question of whether breaking learning down into its
constituent dimensions and relentlessly pursuing proxy indicators for each of these is
the best use of time when weighed against what could be achieved by playing to the
strengths of panel survey data. Furthermore, we have already seen how researchers
using cross-sectional surveys were able to construct their own items to measure infor-
mal learning. With the BHPS, we could only use the items already in the survey.
Similarly, whereas life history work centres individual lives, the strengths of survey
data lie in seeing people as members of groups or clusters. Whilst an important
strength panels have over cross-sections is their ability to analyse individual change
over time, the research focus of all surveys (panel and cross-sections) is on teasing out
large-scale patterns and trends. Thus, although panels can reveal a great deal about
how each person changes over time, a more efficient way of using ‘person-period’ (the
same individuals over time) information in a large data set is to randomly select a sub-
sample of individuals (perhaps stratified into groups defined by the values of relevant
184 P. Hodkinson and F. Macleod

predictors) to conduct exploratory analyses. That is, though it is possible to use panel
data to tease out the stories of individual lives over time, such large scale highly
complex datasets deal better with the sort of information about people that enables
them to be described as members of population (sub)groups or be compared with
other groups.
For group descriptions and comparisons to be authoritative (credible) it is neces-
sary to have consistency and precision of the measures across respondents.
Differences between survey respondents must be based on differences in their views
and experiences rather than anything else. This means not only having valid and reli-
able measures, but also having standardisation at every level so that the data can be
relied upon. Survey questions must be designed in such a way that they mean the
same thing to all respondents, otherwise comparisons would be invalid. Answering
questions should be possible and easy for all respondents. To achieve this, the words
of the survey interview schedule need to be written in such a way that all interviewers
can follow the same script without deviating from it. They must, in short, conduct the
interview in a standardised way so that they avoid misleading or misdirecting the
interviewee and, as far as is humanly possible, ask the questions in the same way as
other interviewers conducting the same survey. Thus, in direct contrast to the life
history interview, the survey interview is deliberately made generic and not tailored
or customised to the individual interviewee and, as such, is not conducive to eliciting
the nuances and subtleties found in participatory and construction accounts of learn-
ing. Although probes (routed questions) are used extensively in large-scale surveys to
ferret out more information, they too have to be standardised.
From a participatory or construction perspective learning does not come about as
a result of simple causal relations among variables in the traditional sense. Instead,
learning reflects a process that is largely responsive to broader social and cultural
conditions and/or norms. This means that a focus on causal relations, in the tradi-
tional counterfactual sense, may hide lifelong learning as formation, or biography, or
narrative, shifting inquiry away from the whole towards its constituent parts as though
these were independent processes. Learning, conceptualised as participation or
construction, can involve the rational calculation and assessment of current situations
or circumstances. It can involve projection into the future that serves to orientate
present activities. It may involve the effects of past on the future but also the antici-
pated future on the present. These ‘reverse causations’ (Marini & Singer, 1984) mean
that attempts to unravel causes and effects of lifelong learning may do little to deepen
understandings of learning across the life course. It also means that an examination
of lifelong learning using participatory and formation conceptualisations is not well
accommodated with statistical models, which require the researcher to clearly delin-
eate independent and dependent variables, exogenous and endogenous variables,
causes and effects. It would thus seem that traditional statistical techniques have few
tools at their disposal to examine learning that involves a view of relationalism found
within participatory and formation accounts. Put differently, both panel survey and
cross-sectional survey approaches are most comfortable when adopting a broadly
acquisitional approach to learning.
 Concepts of learning and research methodologies 185

Consequences and implications

The core argument of this paper is that different research methodologies have fairly
strong affinities with particular conceptualisations of learning. That is, the strengths
and weaknesses of mini-ethnographies, as defined here, are closely related to partici-
patory views of learning; the strengths and weaknesses of life history are closely
related to views of learning as construction; and those of large scale survey research
are closely related to views of learning as acquisition. If this argument is accepted,
then, at least in relation to learning, it decentres some very common assumptions
about research. Put bluntly, it means that choice of a particular research methodology
is likely to skew the research into understanding learning in particular ways. No meth-
odology can act as a conceptually neutral lens, transparently revealing what learning
is. That is, in relation to decisions about how learning should be conceptualised,
research methods are all biased.2
The consequences of this are significant in some ways, but less so in others. If a
researcher has already decided, for whatever reasons, to adopt a particular view of
learning, and therefore asks research questions deriving from that conceptual posi-
tion, it is natural and sensible to adopt a methodology that can answer those questions
and is sympathetic to the understanding of learning being adopted. This is what
happened in the TLC project, as we had already identified the list of influences we
wanted to investigate before the research was done. In Learning Lives we had not as
a team thought through our stance on learning in detail, but our explicit focus on
agency and identity influenced our research questions and was arguably already push-
ing us towards various conceptions of learning as construction, of narrative learning
and learning as becoming and it was these same interests that also pushed us into
choosing a life history approach.
The affinities of different research methodologies with different conceptualisa-
tions of learning influence research but are not deterministic. It is quite possible to
hold a view of learning that is at odds with the selected methodology, but to do so
requires considerable effort and a clear awareness of the limitations of the methods
being used. The TLC mini-ethnography did not prevent the team from conceptu-
alising learning in ways that went beyond most participatory approaches, by theo-
retically blending participation in learning cultures with a cultural theory of
learning that owes much to Deweyan construction (Hodkinson et al., 2008).
Similarly, in Learning Lives, the panel survey work strove to go beyond an acquisi-
tional understanding of learning, as did Felstead et al. (2008) using a cross-
sectional survey.
Of greater significance is the fact that, if we are right in our analysis, empirical
research alone cannot adjudicate between different understandings of learning. It
can and does produce evidence that may support one version of learning and under-
mine another, but it always does so from a non-neutral position. It follows that
arguments about preferences for one view of learning as opposed to others, or
between those who strive for a unified understanding of learning as against those
(Hager, 2005) who argue that there are different types of learning or that we need
186 P. Hodkinson and F. Macleod

pluralistic ways of understanding learning (Sfard, 1998), are essentially conceptual

and philosophical arguments, where empirical evidence plays an important but
supportive role. This means that researchers into learning need to engage not only
with the implications of the strengths and weaknesses of the methodologies they
adopt, but also with the ongoing debates about how learning can and should be
conceptualised.
Our final point is a practical one. In the Learning Lives project we struggled to inte-
grate three methodologies: life history; longitudinal qualitative research conducted in
real time; and the panel survey. The first two blend together fairly easily, apart from
some tensions about interview approaches, which do not need discussion here. This
relatively easy fit is because both methods share the same affinity with understanding
learning as construction. However, it proved much more difficult to integrate the
panel survey findings with those from the qualitative research. This is partly because
the affinities of these methodologies with different conceptualisations of learning
mean that integration is not simply a matter of blending empirical data. Similar prob-
lems would occur if we were trying to combine life histories with ethnographies of
particular learning locations.
This difficulty points to the limitations of one obvious solution to the problems
identified here. At least in the UK, there is a strong movement favouring mixed
method research—a movement of which the Learning Lives project was a part. The
assumption is that because all research methods have different strengths and limita-
tions, we are more likely to discover a triangulated truth by using more than one.
There is a powerful appeal in that argument in relation to learning. Perhaps we need
to use mixed methods in order to overcome the inherent biases towards different ways
of understanding learning in different methodologies. However, the use of mixed
methods, including mixing different qualitative methods, cannot, of itself, solve the
difficult issues around the blending of different conceptions of learning. At root, the
problem is theoretical and conceptual, not an empirical matter than can be settled
through methodological triangulation.
For us, however, using mixed methods worked in a different way from that
normally envisaged. The contrasts between the qualitative and quantitative parts of
Learning Lives have sharpened our awareness of, and thinking about, how we can
best understand learning and about these issues of methodological bias in ways that
might not have happened had both approaches not been used. Because of the overall
concern with the relationships between learning, identity and agency, leading to
progressive adoption of a view of learning as construction or becoming, the panel
survey research team worked hard, within the limits of their methodology, to find
ways to use the survey data to further illuminate these issues. However, the difficulties
faced were large and our solutions were partial.
All researchers need to remain fully aware of the orientations towards learning that
are implicit in any chosen methodology. Furthermore, these differing conceptual
orientations mean that mixing methods in relation to learning is not just a technical
problem, but also a conceptual and theoretical struggle, which can be much more
difficult than most methodology texts recognise.
 Concepts of learning and research methodologies 187

Acknowledgements
Learning Lives: Learning, Identity and Agency in the Life-Course is funded by the
Economic and Social Research Council, Award Reference RES139250111 and is
part of the ESRC’s Teaching and Learning Research Programme. Learning Lives is
a collaborative project involving the University of Exeter (Gert Biesta, Flora Macleod,
Michael Tedder, Paul Lambe), the University of Brighton (Ivor Goodson, Norma
Adair), the University of Leeds (Phil Hodkinson, Heather Hodkinson, Geoff Ford,
Ruth Hawthorne) and the University of Stirling (John Field, Heather Lynch). For
further information see [Link]

Notes
1. Because this paper draws upon the authors’ separate experiences, sometimes working in col-
laboration with different groups of others, we have struggled to find the right use of pronouns.
Each of us has led in reflecting on our own research experiences and we have named the lead
author for each of these sections. In such sections, ‘we’ or ‘us’ refers to that lead author and the
other researchers working with her or him on that project.
2. We are using this term to mean ‘mental tendency or inclination’ towards a particular view of
learning, but not ‘irrational preference or prejudice’ (quotes from Collins Concise Dictionary,
1985).

References
Alexander, P. A. (2007) Bridging cognition and socioculturalism within conceptual change
research: unnecessary foray or unachievable feat? Educational Psychologist, 42(1), 67–73.
Alheit, P. (1995). Biographical learning: theoretical outline, challenges and contradictions of a
new approach in adult education, in: Alheit, P., Bron-Wojciechowska, A., Brugger, E. and
Dominicé, P. (Eds) The biographical approach in European adult education (Wien, Germany,
Verband Wiener Volksbildung).
Alheit, P. & Dausien, B. (2000) Biographicity as a basic resource of lifelong learning, in: Peter Alheit
(Ed.) Liefelong learning inside and outside schools (Roskilde, Roskilde University, University of
Bremen and University of Leeds).
Anderson, J. R., Reder, L. M. & Simon, H. A. (1996) Situated learning and education, Educational
Researcher, 25(4), 5–11.
Anderson, J. R., Reder, L. M., & Simon, H. A. (1997) Situated versus cognitive perspectives: form
versus substance, Educational Researcher, 26(1), 18–21.
Antikainen, A., Houtsonen, J. Kauppila, J. & Huotelin, H. (1996) Living in a learning society: life
histories, identities and education (London, Falmer).
Beckett, D. & Hager, P. (2002) Life, work and learning: practice in postmodernity (London, Routledge).
Bereiter, C. (2002) Education and the mind in the knowledge age (London, Lawrence Erlbaum).
Billett, S. (2001) Learning in the workplace: strategies for effective practice (Crow’s Nest, NSW, Allen
& Unwin).
Billett, S. & Somerville, M. (2004) Transformations at work: identity and learning, Studies in
Continuing Education, 26(2), 309–326.
Brown, J. S., Collins, A. & Duguid, P. (1989) Situated cognition and the culture of learning,
Educational Researcher, 18(1), 32–42.
Cobb, P. & Bowers, J. (1999) Cognitive and situated learning perspectives in theory and practice,
Educational Researcher, 28(2), 4–15.
188 P. Hodkinson and F. Macleod

Colley, H., Hodkinson, P. & Malcolm, J. (2003) Informality and formality in learning: a report for the
Learning and Skills Research Centre (London, Learning and Skills Research Centre).
Colley, H., James, D., Tedder, M. & Diment, K. (2003) Learning as becoming in vocational
education and training: class, gender and the role of habitus, Journal of Vocational Education
and Training, 55(4), 471–497.
Davies, J. & Tedder, M. (2003) Becoming vocational: insights from two vocational courses in a
further educational college, Journal of Vocational Education and Training, 55(4), 517–539.
Dominicé, P. (2000) Learning from our lives: using educational biographies with adults (San Francisco,
Josey Bass).
Engeström, Y. (1984) Students’ conceptions and textbook presentations of the movement of the
moon: a study in the manufacture of misconceptions, in: H. Nielsen & P. V. Thompsen (Eds)
Fysik i skolen: problemer og perspectiver (Aarhus, Aarhus Universitet. Det Frysiske Institut).
Engeström, Y. (2001) Expansive learning at work: towards an activity-theoretical reconceptualisa-
tion, Journal of Education and Work, 14(1), 133–156.
Engeström, Y. (2004) The new generation of expertise: seven theses, in: H. Rainbird, A. Fuller &
A. Munro (Eds) Workplace learning in context (London, Routledge).
Felstead, A., Fuller, A., Unwin, L., Ashton, D., Butler, P. & Lee, T. (2008) Surveying the scene:
learning metaphors, survey design and the workplace context, Journal of Education and Work,
18(4), 359–383.
Flyvbjerg, B. (2001) Making social science matter: why social inquiry fails and how it can succeed again
(Cambridge, Cambridge University Press).
Goodson, I. & Sykes, P. (2001) Life history research in educational settings: learning from lives
(Buckingham, Open University Press).
Greeno, J. G. (1997) On claims that answer the wrong questions, Educational Researcher, 26(1), 5–17.
Hager, P. (2005) Current theories of workplace learning: a critical assessment, in: N. Bascia, A.
Cumming, A. Datnow, K. Leithwood & D. Livingstone (Eds) International handbook of educa-
tional policy (Dordrecht, Springer), 829–846.
Hager, P. & Hodkinson, P. (in press) Moving beyond the metaphor of transfer of learning, British
Educational Research Journal.
Hodkinson, P. with Anderson, G., Colley, H., Davies, J., Diment, K., Scaife, T., Tedder, M.,
Wahlberg, M., & Wheeler, E. (2007) Learning cultures in further education, Educational
Review, 59(4), 399–413.
Hodkinson, P. & Hodkinson, H. (2004a) The significance of individuals’ dispositions in workplace
learning: a case study of two teachers, Journal of Education and Work, 17(2), 167–182.
Hodkinson, P. & Hodkinson, H. (2004b) The complexities of workplace learning: problems and
dangers in trying to measure attainment, in: H. Rainbird, A. Fuller & A. Munro (Eds) Workplace
learning in context (London, Routledge/Falmer).
Hodkinson, P., Biesta, G. & James, D. (2008) Understanding learning culturally: overcoming the
dualism between social and individual views of learning, Vocations and Learning, 1(1), 27–47.
Hodkinson, P., Hodkinson, H., Evans, K. & Kersch, N., with Fuller, A., Unwin, L. & Senker, P.
(2004) The significance of individual biography in workplace learning, Studies in the Education
of Adults, 36(1), 6–26.
James, D. & Biesta, G. J. J., with Colley, H., Davies, J., Gleeson, D., Hodkinson, P., Maull, W.,
Postlethwaite, K. & Whalberg, M. (2007) Improving learning cultures in further education
(London, Routledge).
Lave, J. (1988) Cognition in practice (Cambridge, Cambridge University Press).
Lave, J. & Wenger, E. (1991) Situated learning (Cambridge, Cambridge University Press).
Livingstone, D. (2000) Exploring the icebergs of adult learning: findings of the First Canadian
Survey of Informal Learning Practices. NALL, New Approaches to Lifelong Learning) Working
Paper 10 (Toronto, Ontario Institute of Studies in Education, University of Toronto).
Marini, B. & Singer, B. (1984) The order of events in the transition to adulthood, Sociology of
Education, 57, 63–84.
 Concepts of learning and research methodologies 189

Mason, L. (2007). Introduction. Bridging the cognitive and sociocultural approaches in research
on conceptual change: is it feasible? Educational Psychologist, 42(1), 1–7.
Rogoff, B. (2003) The cultural nature of human development (Oxford, Oxford University Press).
Saljö, R. (2003) From transfer to boundary-crossing, in: T. Tuomi-Gröhn & Y. Engeström (Eds)
Between school and work: new perspectives on transfer and boundary-crossing (Amsterdam, Elsevier
Science).
Sfard, A. (1998) On two metaphors for learning and the dangers of choosing just one, Educational
Researcher, 27(2), 4–13.
Tedder, M. & Biesta, G. (2007) Learning from life and learning for life: exploring the opportunities
for biographical learning in the lives of adults, paper presented at the ESREA Conference on
Life History and Biography, Roskilde University, Denmark, 1–4 March.
Tuschling, A. & Engemann, C. (2006). From education to lifelong learning: the emerging regime
of learning in the European Union, Educational Philosophy and Theory, 38(4), 451–469.
Wenger, E. (1998) Communities of practice: learning, meaning and identity (Cambridge, Cambridge
University Press).
Wenger, E., Mcdermott, R. & Snyder, W. M. (2002) Cultivating communities of practice (Boston,
MA, Harvard Business School Press).
Wolcott, H. (1999) Ethnography: a way of seeing (Walnut Creek, CA, AltaMira Press).