Data Analysis: Getting conceptual

Helen Scott, Ph.D.

Abstract

This paper will track my battle to ‘get conceptual’ in the
production of a Grounded Theory. It will discuss early attempts
at creating substantive codes through the process of open coding
which, despite my best efforts, merely produced descriptive codes.
It will illustrate the process by which these descriptive codes
became more conceptual, earning the title of substantive code and
how their presentation in essay form produced a perfect example
of ‘conceptual description’. It will then describe the slow dawning
of the purpose of ‘theoretical codes’ as organisers of substantive
codes and the emergence of a Grounded Theory.

Open Coding: The mechanics

The substantive population1 of my study is adult online
distance learners whose main concern (in descriptive terms) is
finding the time to study. The process which addresses this
concern is the ‘temporal integration of connected study into a
structured life
’ (Scott, 2007 a, b). An overview of the theory and
its structure is offered in the Appendix to this paper.

Participants were located all over the world, therefore most
of the data for the study was collected online using email or chat2.
Typically, the first emailed response from each person was the
most detailed response with perhaps one or two emails received
in reply to follow up questions. I would print and read the email
or chat transcript for an overview. If I felt that I could
understand what the participant was telling me, I started coding,
otherwise I waited until subsequent emails or chat sessions
improved my understanding. When open coding I had a piece of
paper in front of me which asked:

• What category does this incident indicate?
• What property of what category does this incident
indicate?
• What is the participant’s main concern? (Glaser, 1998, p.
140)

I asked these questions of every incident that I perceived and
I wrote the codes in the margin. In addition, I used coding cards
and wrote the indicators3 in full on the appropriate coding card(s)
and referenced the indicator both on the printed, coded document
and on the coding card. As a reference number, I used the initial
of the person plus the incident number e.g. J-10. Coding to cards
was cumbersome and time consuming but it helped me to get a
feel for the process and to feel in control of my data. Actually I
had too much control of my data; since I could record each and
every code, the number of codes soon spiralled out of control.
Thus the rhythms built into the method could not operate
allowing the undesirable state of ‘full coverage’ over parsimony.
Had I only coded in the margins, the relevant might have
emerged more quickly, by the process of forgetting that which did
not pattern out. Not yet understanding this, I would write the
name of the code at the top of the card and in the body of the card
write the reference number and the indicator. This was
reassuring; as the cards became fuller, I could compare incident
to incident easily. I could see how codes grew and became
saturated. I could compare codes with codes and indicators
between codes. I could see codes metamorphose into other codes
and see the dimensions of codes emerge either across cards or
within a card. For example, the coding card ‘Compliance’ listed
indicators of ‘high compliance’ and ‘minimal compliance’.
Indicators of ‘reducing compliance’ emerged, then ‘non
compliance’
and then there were degrees of ‘non compliance’.
Thus I realised that ‘non compliance’ was an aspect of
‘withdrawal’ which itself was ‘partial withdrawal’, ‘temporary
withdrawal’
and ‘permanent withdrawal’.

The practice of using coding cards misled me into thinking
that one allowed the relationships between codes to remain
unwritten and subject to preconscious processing and that one
sorted codes, whereas the stricture is to memo ideas about the
relationships between codes and to sort memos (Glaser, 1978, p.
83). When I finally realised this, the relief was enormous and led
to a flood of memos.

When coding, I was very much aware of having some
experience in the field of online learning as a student, designer
and facilitator. As advised, I interviewed myself (Glaser, 1998,
p.120) which helped enormously; setting down my experiences
and thoughts and coding them helped me to relax about what I
thought I knew. If it was relevant it would pattern out, if not it
would sink without trace. I wrote in a method memo4 of my
conscious effort “to follow the data absolutely. I am not coming
outside of it and investigating how much of this is obvious or
banal”. I was therefore not judging the data, merely working with
it. Thus in coding, I believe that I was successful in suspending
my professional concerns; however, I recognise that the way that
I understand the world determines how I interpret any given
incident, where I fracture the data and thus the codes that I
choose.

The Main Concern

After nine months of online discussions and open coding I
prepared for a Grounded Theory Seminar in October 2004. My
elaboration to the question ‘Have you identified your core
category?’ is shown in Textbox 1..

Textbox 1. Attempting to identify the core category

Have you identified your core category? If so please elaborate.
I think I have several potential cores as follows: (categories shown
in capitals).

Online Learning offers the opportunity of further study to part
time adult distance learners. The property of Online Learning 24/7
AVAILABILITY, where the learning environment is open for
business 24/7, means that I (as an online student) can develop the
solution: I ADAPT. Both of these things together, means that for
those with EXISTING COMMITMENTS to work and family,
further study becomes a viable option. There is a process where
the NEED for study is identified and (consciously or
unconsciously) EXPECTATIONS as to outcomes of studying are
formed. Where the outcome of the decision process leads to a
decision to undertake further study there is
a COMMITMENT2LEARNING (to various degrees). Committing
to further study gives rise to the problem of when do I DO THE
WORK? And the solution I ADAPT and find time in which to
work: This can mean TIMETABLING TIME i.e. planning time;
OPPORTUNISTIC USE OF TIME i.e. taking advantage of the
spare moment; JUGGLING COMMITMENTS to free up time to do
work; EXTENDING-THE-DAY i.e. working late/getting up early.

Equal to the opportunity offered by 24/7 AVAILABILITY is the
problem of ACCESS: How do I overcome the barriers and gain
ACCESS to the opportunities to learn? TECHNOLOGY is a barrier
to entering the learning environment including issues of: the right
to use the equipment (privately owned or publicly available), which
is of an appropriate specification, having the right software and
having access to an Internet connection which is fixed
link/wireless. SCATTERED ACCESS where access is spread
across machines, where learners use multiple machines, gives rise
to problems in managing software and work files.

There is a relationship between ACCESS: TECHNOLOGY and the
scope for ADAPTing.

Also: there are language barriers to access the opportunities to
learn because of the reliance on dense text; Dyslexia is a barrier to
access the opportunities to learn for the same reason. There are
financial barriers to the right to enter the environment i.e. the cost
of course

‘Doing’ THE WORK leads to an iterative ongoing process of
EVALUATION of the RELEVANCE OF WORK, i.e. its usefulness,
potential use or its inherent interest i.e. its VALUE OF WORK to
me. A positive VALUE OF WORK to me leads to CONTINUED
COMPLIANCE where work is undertaken. Low VALUE OF
WORK is of REDUCING VALUE and therefore results in
REDUCED COMPLIANCE or NON COMPLIANCE AND
WITHDRAWAL.

At this point I was searching for the point from which to ‘hang’
my theory. The sense of knowing but not knowing was
infuriatingly tantalising until at last I realised ‘Time is the
problem for all my people.’ (Textbox 2)

Textbox 2: Memo on time

Time is the problem for all my people. Time to develop
competencies: knowledge domain competence (time to explore
issues), technical competence (time to explore software, master
technology), language competence either as foreign language
speaker or as dyslexic (time to translate/understand, formulate
and express ideas). Finding time/making time/stealing time to
study (juggling existing commitments). Constantly evaluating
whether the time spent studying is well spent – is the work
relevant, valuable, useful. If yes continue, if no withdraw.
Access as an issue which eats time. 24/7 Availability of online
learning enabling people to ADAPT and find time, making study
possible. A tutor is someone who saves me time. Poor design (of
work or of environment) wastes time.
Memo dated November 10, 2004

Conceptualising the main concern during the seminar had
led to the suggestion of the ‘Tyranny of flex-time as integrated
into a structured life’5. This proved an extremely useful example
of how to conceptualise a problem and showed me how to move
forward though I was aware that it was not quite right.
Eventually, I realised that the ‘tyranny’ was experienced by and
captured the main concern of some of the participants of the
study but not of all of them. Some did not experience the tyranny
as experienced by others. The conceptualisation thus evolved over
the next few months into the main concern of ‘integrating study
into a structured life’, where the problem and its resolution
eventually became as one.

Selective Coding and Theoretical Sampling

At this point I had almost all I needed to create a theory
except the skill. My next steps were to selectively code incidents
that related to ‘time’, to saturate those codes and to theoretically
sample within the substantive population6 of adult online
distance learners for comparison groups. Glaser writes:

The general procedure of theoretical sampling, as we now
shall describe it, is to elicit codes from raw data from the
start of data collection through constant comparative
analysis as the data pour in. Then to use the codes to
direct further data collection, from which the codes are
further theoretically developed with respect to their
various properties and their connections with other codes
until saturated. (Glaser, 1978, p. 37)

I decided to begin selective coding by revisiting the data from the
initial interviews being certain that I had not noticed all that
there was to notice about the participants’ comments concerning
time. I conducted further discussions with students from two
other postgraduate courses run online. Glaser writes:

It [theoretical sampling] focuses questions more and more
on the direct emergence of the theory (thus showing
again, how interview schedules constrain theoretical
sampling). Questions constantly change with the
requirements of the emergent theory and theoretical
sampling. (Glaser, 1998, p. 157)

Here, I found a tension between the constraints of the online data
collection method and the Grounded Theory method. The
questions deduced from the induced codes, at this point were:

• Was time a big issue for you?
• How did you fit in work and personal life and study?
• How did you decide what to work on and what not to work
on?
• Why did you take this course?

I did ask these questions of one participant with whom I had
corresponded earlier and who had not replied to my original
questions but who had contacted me again. The response gained
was extremely useful though lacked the context of earlier replies.
I reasoned that since the original questions elicited useful
responses and that since the earlier participants seemed to
respond well to the approach of having their earlier comments
quoted back to them and being asked for more details, that I
would continue with this approach but code and follow up only on
issues relating to the main concern and it’s resolution. I made
further attempts at writing useful memos and I found that I used
memos to tease out thoughts about categories and that the
memos showed my thought process but not yet the relationships
between concepts.

In September 2004, I believed that I had the horrible amount
of over 130 open codes7 in which time was mentioned only three
times:

• Allocation of resources – time
• Designed work – similar time/discontinuity
• Making time

At that point, I saw time as a flow, as a resource and whilst I
had identified ‘similar time’ working, I had not yet recognised
time as structure. By December 2004, I had achieved a step
change in the way I thought about the design of a course and put
aside Wenger’s (1998) concepts of ‘designed work’ and ‘designed
environment’ for the moment and started to think about the
‘Time Design’ of a course. My observations had surfaced
assumptions about the pace of work achieved and the timing of
when work would be accomplished based on assumptions about
learners’ competencies, in particular language competence and
also about learners’ work/rest and wake/sleep patterns. My data
collection process had shown me that learners have a range of
competencies and different work/rest, wake/sleep patterns which
are further complicated by different time zones; thus that the
pacing and the timing of work is often different from that
assumed by the course designer. By March 2005, I believed the
categories of Personal Commitment Structure and Time Design to
be as shown in Textbox 3.

Textbox 3: Memo on categories

Personal Commitment Structure
Commitment type

o Work
o Family
o Social
o Learning
o Other

Committed time

• Structure points

o By this time
o At this time
o Organisers (lunch
break)
o Fixed (children pickup)

• To commitment types

o Work
o Family
o Timetabled study time
o Spare
o Sleep

Location time relative to base time
Patterns

• Wake/sleep pattern
• Work/rest pattern

Time Design

Assumed/implied typical learner
profile:

• Assumed/Implied Personal
Commitment Structure
• Assumed/Implied Personal
Competencies

Attendance requirement
Course period e.g. 10 weeks
Study hours e.g. 80 hours
Core Period e.g. one week
Assessment period e.g. 3 weeks
Base Time (of course relative to
UTC)
Focal Time (of local group or tutor)
Structure Points

• Start/end points
• Assessment points
• Organising points
• Emergent connection
points

Connection Design

• Same time connections
• Similar time connections

o Any time
connections

Conceptual Description

In preparation for a Grounded Theory seminar in March
2005, I wrote per Textbox 4:

Textbox 4: Memo on core category.

Have you identified your core category? If so please elaborate:
GT Summary:
The issue for part time adult online learners – or CONNNECTED
LEARNERS – is DEVELOPING COMPETENCE in the context of
CONNECTED LEARNING ONLINE. The main concern that
protagonists are constantly working to resolve is the
INTEGRATION of the TIME DESIGN of the learning opportunity
into their PERSONAL COMMITMENT STRUCTURES.
INTEGRATION causes TIME TENSION and for some learners
TIME TYRANNY.

The warning was in the phrase ‘GT Summary’. Instead of being
able to state my core category as ‘Integrating study into a
structured life’ my current understanding forced me to write a
paragraph. The concepts are there but I am going for ‘full
coverage’. I cannot let any of my concepts go, I am wedded to
them all. The working paper prepared for this seminar is a
perfectly crafted example of ‘conceptual description’ (Glaser,
2001) an excerpt of which is shown in Textbox 5:

Textbox 5: Conceptual description

A CONNECTED LEARNER will commonly have commitments to,
for example, family and employment and perhaps to other social
commitments e.g. Church or sport. A PERSONAL COMMITMENT
STRUCTURE will therefore comprise COMMITTED TIME to
work, family, social organisations, self, sleep and to timetabled
study time. Any time left over is ‘spare’ time. Thus Committed
time plus spare time = Wake time. Wake time plus sleep time = All
time.

A WORK/REST pattern relates to days, takes into account shift
working (e.g. one month on, six weeks off) and an example of which
is the 5 day Saturday to Wednesday working week and the 2 day
Thursday/Friday weekend of the United Arab Emirates. A learner
working with such a pattern will find it harder to work and
connect within a CORE period designed around a 5 day Monday to
Friday working week and the 2 day Saturday/Sunday weekend
where work for the core is released on Saturday.

A WAKE/SLEEP pattern is over 24 hours, takes into account shift
working and is relative to BASE TIME of the CONNECTED
LEARNING OPPORTUNITY and the FOCAL TIME of a group of
learners. A CONNECTION DESIGN which requires SAME TIME
working e.g. tutorials or next-stepping group organisation sessions
can either effectively exclude some learners or add to the TIME
TENSION experienced. Figure C.1 shows how day time workers in
the USA attending UK based courses are effectively excluded from
synchronous sessions by their WAKE/SLEEP pattern relative to
BASE TIME because most UK chat sessions are held when they
are asleep. Similarly, night shift workers in the UK can only easily
attend chat sessions held during their evening.

‘Theoretical coding’ is needed to rescue the theory, to enable the
theory to be brought into relief from the flatness of descriptive
codes where “…theoretical codes implicitly conceptualise how the
substantive codes will relate to each other as a modelled,
interrelated, multivariate, set of hypotheses in accounting for
resolving the main concern” (Glaser, 2005a, p. 11).

My memos show how I was desperately seeking the structure
of my theory as I drew bubble maps and decision trees to help me
see the patterns but they were one dimensional and I focused on
either what was in the middle of the bubble map or at the top of
the tree. The best that they could do was to capture my confusion
and illustrate my struggle to identify the structure as I sought to
understand how to model the theory.

Upping the Level of Conceptualisation

Two strands of thought collided. Firstly, ‘How many Time
Designs are there?’ Since there are an infinite number of
variations of timings of assessments, course duration etc., the
idea becomes useless. How can one possibly account for all the
Time Designs where the distinctions between each are all but
indistinguishable? It is this question that led me to leave behind
the descriptive properties of course period, study hours,
assessment period and to abstract the implicit; that is, to
recognise that there are start points and end points of courses
and assessment points. Thus I moved from the descriptive to the
conceptual. The second strand had to do with participants’
comments about structure; that structure is helpful and that the
lack of structure is problematic; that structure is linked to how
learners organise their lives and integrate study that the
beginning of the week is an organising point, where new work is
required to be done. Together these strands led to the realisation
that the descriptive properties of ‘Time Design’ (Textbox 3) were
based on insight drawn from my experience as a course designer
but that what really mattered to the learners were the structure
points and the degree to which the points were fixed or moveable.
Correspondingly, it was less relevant whether a commitment was
to family, work, social life etc. and more relevant as to whether
the structure point was fixed or moveable and thus that the
whole issue for learners was integrating their structure points
into one life.

It is at this late stage that I can label ‘Time Design’ as a
category, having a property ‘Structure Point’ where the
dimensions of that property relate to the degree to which a
structure point is fixed (or moveable). I can also label ‘Personal
Commitment Structure’ as a category having a property ‘Structure
Point’ having dimensions along a range of fixed to moveable.

Sorting and Theoretical Coding

For me, sorting and theoretical coding happened hand in
hand, where I understood theoretical coding to mean the
emergence of relevant theoretical codes as opposed to (as I had
first envisaged) the active labelling of substantive codes as
pertaining to a theoretical code, in the manner of open and
selective coding. I had first sorted my memos in preparation for
the working paper prepared for the seminar of March 2005. I
sorted by code and wrote about each. This was an exercise in
finding out what I knew and for me was a necessary part of the
process – part of finding out what not to do, of finding out that
this approach results in conceptual description and how a
conceptual description reads.

In April 2005 The Grounded Theory Perspective III:
Theoretical Coding
(Glaser, 2005a) was published and offered
invaluable guidance and discussion of ‘new’ theoretical codes. A
memo of April 2005 is shown in Textbox 6 where I notice that
several theoretical codes may be relevant.

Textbox 6: Memo on method

“I have too much; am blurring two stages. I don’t have the proper
‘story’ about how people absorb learning into one life. I do have:
juggling-integrating-evaluating.” I should have this sorted before I
start to identify theoretical codes. However, I think I need the
theoretical code to help me make sense of the substantive!

I can see – as I read TC 05 – that many different theoretical codes
might be relevant. Balancing, cycling, Basic Social Process
(becoming a student).

I am having tremendous difficulty in seeing the theoretical
patterns. I think I have 3 levels: strategic, operational,
implementation with 3 level looping and spiralling and may have
two different spirals one for the successful and one for the
unsuccessful.

But given that a TC is about the relationships between codes, I’m
not really at that stage of identifying, merely sensitising self to
same and playing with ideas.”

All the theoretical codes in Textbox 6 are found to be relevant
together with a few others and it will be helpful to define these. A
‘Basic Social Process’ (BSP) “processes a social or social
psychological problem from the point of view of continuing social
organisation. Irrespective of whether it solves the problem, to
some degree it processes it. (Glaser & Holton, 2005, p. 6)

There are two types of BSPs – basic social psychological
process (BSPP) and basic social structural process
(BSSP). A BSPP refers to social psychological processes
such as becoming, highlighting, personalising, heath
optimising awe inspiring and so forth. A BSSP refers to
social structure in process… (where a) BSSP abets,
facilitates or serves as the social structure within which
the BSPP processes. (Glaser & Holton, 2005, p. 11).

Cycling “refers to going over the same path over and over. It also
refers to going over and over the different paths in succession
whatever the unit action. It easily refers to people’s temporal
order of work, eating, sleeping etc.” (Glaser, 2005a, p. 24).
Balancing “is handling many variables at once in order to start
an action, keep an action going or achieve a resolution. One gets
an equilibrium between all the variables. One can achieve stasis
for a time.” (p. 29)

Having sensitised oneself to different theoretical codes, it is
then a matter of ‘trying on’ various codes while sorting memos to
see which ones fit. In June 2005, I notice that there are many
potential Basic Social Structural Processes appertaining to any
one learner (e.g. parenting, studying, working). I confuse the
theoretical code ‘Balancing’ with the substantive code ‘Juggling’
– a stage in ‘integrating study into a structured life’. This is
understandable since

Balancing is an abstract model that also can be seen
substantively or used as a substantive category e.g. the
professional-client balance in a doctor-patient
relationship. Balancing as such can also be used as a
BSP, when it is worked or occurs in stages such as
balancing out the factors in a divorce settlement or in
resistance to change in organisation. Thus balancing
provides it s own mix of TC and substantive categories.
(Glaser, 2005a, p. 29)

I also wonder if those who juggle and those who struggle are
defined by the integrating strategies they employ or the outcomes
of their efforts to integrate. I make my first attempt at expressing
the ‘homeostasis of motivation’ modelled on Thulesius’ (2003)
‘homeostasis of hope’. The homeostasis of hope has three
variables:

Existential hope (H) which is a function of the value of
every lived moment (V), and expected time left to live (T);
H=V x T. Existential hope is defined as the motivation
and well-being required to live a normal everyday life. In
the disclosure situation the expected time left to live (T)
goes down and this reduces the value of the lived moment
(V) and thus existential hope (H) drops very fast…. By
increasing V and T the patient and the caregivers are
trying to regain the homeostasis of hope. (Thulesius, 2003
p 158)

Thelusius captures beautifully the interrelationship of variables
and the impact that a change in one has upon the other. In my
study, at one point I had a huge and descriptive list of problems –
which interfered with the integration of study – the negative
effects of which were mitigated or exacerbated by the behaviours
of the learners. The greater the learners’ competence levels the
less the negative effect. It occurred to me that an algorithm such
as this might be helpful in expressing the complex interrelationships
between the variables in my study.8

At this point in the analysis, it felt as if all the categories
were suspended above me, waiting to be told where to land. I was
not threatened by them but there were a lot of them and they
were beginning to weigh heavy. I remembered and was comforted
by the comment: “Confusion? Rest in the confusion. Confusion is
a really good indicator of something emerging” (Glaser, 2005b). I
disentangled ‘balancing’ and ‘juggling’ realising that in this
study, ‘balancing’ is not part of the substantive code ‘integrating
study into a structured life’ but is a theoretical code, where the
substantive code ‘integrating study into a structured life’ is
modelled by the theoretical code, ‘balancing’. On July 8th, I sorted
my memos again and attempted to sort more intuitively. I had
papers spread over two tables, a desk, the kitchen work surface
and – dangerously – the cooker. I ended up with an enormous pile
of memos under the heading of ‘normal integration’ and two
smaller piles marked ‘integration: step change – new study’ and
‘integration step change – not study’.

By the end of July 2005 and in response to the question
‘What is failed integration?’ (see Textbox 7) I recognised the
theoretical code of type and that I have a typology of learners
where “types indicate a variation in the whole, based on a
combination of categories” (Glaser, 1978, p. 75).

Textbox 7: Memo on Theoretical Coding

Failed to Integrate

What does this mean? To what degree has someone failed to
integrate? Not consumed enough work. Enough work as planned
by self or time design? Failed to integrate today, this week, at all,
ever.

Integration is about the long term integrating of structure points.
On a day to day basis stuff gets squeezed out or squeezed in. It’s
at an Operational/ Implementation level.

Cumulative squeezing out …. And the relationship with
propensity to study?

“I have yet to work out a routine that doesn’t have me stressed out
come exam time”. The rest of his life is constantly tugging at his
sleeve. His wife has the family to support – no time for him. Two
people studying in one family. He is having time taken from him!!!
Study is squeezed out because relatively other stuff is more
important reducing his propensity to study – so he allows –
reluctantly and stressily, study to be squeezed out. But the costs
of failure are high. He has no life and suffers time
tension/tyranny. Come exam time, as a structure point
approaches, becomes an operational/strategic issue – P2S
(Propensity to study) increases and for a while he studies.

This makes him a struggler. How do you cope? It’s a struggle.
Strugglers experience pain. They may complete or they may fadeaway.

Passive Squeezing Out where study of low value and thus P2S is
low.

Operating in avoidance mode and displacement activities allowed
to intervene

Active intentional squeezing out is part of juggling and is
reorganising or reordering. (Jugglers and strugglers will do this).

Leavers decide to stop. Strategic decision.

Fade-aways not so decisive, they keep failing to integrate until the
plug is pulled. (Which is why there are few reliable drop out rates,
merely completion numbers – as people only become fade-aways
when a structure point – an end point – defines them as having
faded away. Stages of fading away: passive squeezing out,
temporary withdrawal, end point defines. Fade-aways have not
necessarily failed – may have developed competence to required
level and have no need of the validation.

This made it easy now to re-sort the huge pile of memos
regarding integration into piles pertaining to ‘jugglers’,
‘strugglers’, ‘fade-aways’ and ‘leavers’. Some of the memos had to
be cut up, for example where I had talked about each type on one
memo of integrating a step change into the personal commitment
structure. I also noticed that the variables relating to type are the
same variables that go into the evaluation calculation. The
variables are predictors of type – of whether or not the learners
will process their problem of integrating study into their
structured lives and the time tension and time tyranny that they
will be prepared to tolerate.

As one by one the theoretical codes brought order to a section
of chaos, the codes were also confirmed and less pertinent
properties dropped. This made it easier to see where other
smaller codes fit in e.g.’ catch up’ is a strategy that both ‘jugglers’
and ‘strugglers’ employ but probably not ‘fade-aways’ and
definitely not ‘leavers’. By October 2005, I have seven theoretical
codes: a Typology, two Basic Social Processes (BSP); i.e. a Basic
Social Structural Process (BSSP) and a Basic Social Psychological
Process (BSPP) also Balancing, Cycling, Amplifying Causal Loop
and Cutting Point. Amplifying causal looping is “.. an ordered,
calculated growth of increased size based on a set temporal path
(Glaser, 2005a, p. 24). For example ‘strugglers’ and ‘fade-aways’
fall further and further behind as they cycle through the basic
social processes: integrating and studying. The Cutting Point
family:

is a variation of the degree family. Degree focuses on the
full range, while here we focus on significant breaks or
cutting points on the range. Cutting points are very
important in theory generation, since they indicate where
the difference occurs which has differential effects”.
(Glaser, 1978, p. 76)

In this study, the Cutting Point is a step change of the Personal
Commitment Structure experienced by ‘Leavers’ e.g. the birth of a
child, the death of a close family member.

This part of the study was about seeking, noticing, exploring,
defining, testing and trying on, refining or rejecting, and rehanging
codes and relationships. During this process I changed
the way I thought about the descriptive codes (e.g. evaluating the
value of study) and descriptive relationships (e.g. leads to).
Where my focus had been principally on codes, my focus moved
on to the dynamic relationships between codes. In this study, this
is where the complexity lies and which is ultimately and
elegantly expressed in an algorithm and a set of propositions.
(Scott, 2007a, b)

Literature Review

Glaser (1998) is emphatic when he writes:

a) do not do a literature review in the substantive area
and related areas where the research is to be done, and b)
when the grounded theory is nearly completed during
sorting and writing up, then the literature search in the
substantive area can be accomplished and woven into the
theory as more data for constant comparison. (p.67)

Consistent with these strictures, I embarked on the literature
review only when I felt confident about the shape of my theory. It
was during the literature search that I came to appreciate much
more the role of sorting in organising the theory and defining the
relationships between categories; since it is the sorting and the
use of theoretical codes to organise my theory that separates my
work from the other qualitative studies reviewed. I can also see
why we are enjoined not to read the literature first. Had I
conducted the review before data collection and analysis, I would
have read widely and wastefully in the field of personality since
that was the field in which I had expected my study to be located.
I had certainly not envisioned a study to do with student attrition
and retention, or student persistence or withdrawal, as became
the case.

Secondly, if I had located that field in advance, my study
would have been abandoned. My horror at finding Kember’s
(1999) article “Integrating Part-time Study with Family, Work
and Social Obligations” was profound; my study was almost done
and my theory emergent. As I sat down to read the article I
struggled to see how I could add to the understanding of the field
particularly since very few of the categories I had identified in my
theory seemed to be new ideas. As it was, however, I quickly
came to see how I could add value.

The qualitative literature in the field of online distance
learning, fully describes the problems such that the mass of detail
is overwhelming (Dupin-Bryant, 2004). Since each issue is given
equal prominence, the burden on online distance learning
professionals is huge – every issue has to be addressed as there
are no clues as to where interventions might be made most
effectively. By offering an explanation of the main concern of
online learners and how the structural conditions impact upon
their experiences, it is possible to identify where interventions,
changes, might usefully be made. It is possible to design for
learner persistence.

Quantitative studies proved similarly unhelpful as
practitioners struggled to find relevant and related variables to
test and conjectured as to the meanings of what their statistics
meant. As I read I could see how the tools, the methods, with
which quantitative researchers analysed their data could not cope
with the complexity of the field. My theory makes it possible to
untangle the threads of description and add meaning to
conjecture; for example, it is now possible to identify separate
strands of temporal issues and academic issues, which matters
were often intertwined, maintaining confusion (e.g. Woodley
2004). This confusion was important since it meant that in some
studies, inappropriate variables were used to measure
persistence rendering the studies valueless. By regarding
persistence and withdrawal (i.e. dropout) as a temporal matter
and the process of academic performance and the outcome of that
process, academic achievement (success or failure) as academic
matters, those variables which do and do not measure learner
persistence can be identified and used or not used as appropriate.
Additionally, the algorithm which captures the dynamic
relationships between codes can be used to identify dependent
and independent variables.

During the review, I also came to appreciate that using the
literature as more data in developing one’s own Grounded Theory
is invaluable, both in finding new categories and particularly in
being able to theoretically sample and saturate existing
categories of the emerging theory. Not least, as I tussled with the
principle authors, I finally came to recognise what my study was
about and that the core category was ‘temporal integration’.

Summary

In this paper, I have exampled the experiential nature of the
process of producing a Grounded Theory (Glaser, 1998, pp.6,102);
detailing how my understanding of the method developed as I
engaged with it. I have also illustrated the power of the method
and in particular, theoretical coding, by showing how the
potentially overwhelming complexity of data is made manageable
by organising theory using ‘theoretical codes’. In doing so I have
illustrated how the call for axial coding and the use of one
theoretical code as suggested by Strauss and Corbin (1998), is
restrictive in the production of a Grounded Theory.

Appendix: Structure of Theory

This is a complex Grounded Theory and for completeness it
may be helpful to outline the structure of the theory, i.e. how the
categories are inter-related. In descriptive terms, the main
concern of learners is fitting study into their lives on an ongoing
basis. In conceptual terms, the basic social psychological process
which processes this concern is temporal integration. This is the
process by which the structure points of the time design of a
connected learning opportunity are combined into the personal
commitment structure of the connected learner
. Thus two related
categories are of import to this theory; the connected learning
opportunity
and the connected learner. The category connected
learning opportunity
has properties of the knowledge domain of
the course, the language of the course, the technology of the course

and a sub-category of time design. The time design has properties
comprising structure points of start points, end points, organising
points
, assessment points and connection points.

The category connected learner has properties of five
personal competencies of the knowledge domain of the course, the
language of the course, technical skills, integration skills and
online learning skills. Each of these properties has dimensions
ranging from high to low, i.e. from high levels of personal
competence to low levels of personal competence. In addition, the
related category connected learning has properties of need for
learning and satisfaction (with learning opportunity) and cost of
failure, all of which have dimensions from high to low. This
category also has a sub-category of personal commitment
structure which has structure points having dimensions of being
more or less fixed.

The process of temporal integration has three stages;
juggling which has properties of scoping, prioritising and
scheduling; engaging having dimensions ranging from full
engagement, reducing by degrees to partial withdrawal,
temporary withdrawal down to full withdrawal (or
disengagement); and evaluation. The evaluation stage involves
the assessment of the benefits of engaging in study, (‘what’s in it
for me?’
) and the costs (‘is it worth it?’). The outcome of the
evaluation is expressed as the propensity to study which forms a
feedback loop to the juggling stage. The balancing algorithm,
captures the relationships between the dimensions of a connected
learner and the structural conditions under which temporal
integration takes place and how they co-vary during the temporal
integration process to impact upon the assessment of the benefits
and costs of engaging in and the propensity to study. Lastly a
learners’ type is defined in the first instance by the two
categories, personal commitment structure and personal
competencies
; by the value of study in the second instance; and by
the cost of failure in the third. See also textbox 8.

Textbox 8: Substantive and theoretical codes of the study ‘The
temporal integration of connected study into a structured life: A
Grounded Theory’

Temporal integration – core category and BSPP with stages of:
Juggling
Engaging
Evaluating

Connected learners – related category
Personal commitment structures – sub category and structural
condition; having properties (e.g. structure points)
personal competencies – sub category having properties (of
knowledge domain of the course, the language of the course,
technical skills, integration skills and online learning skills.) with
dimensions (high or low levels of competence)
value of study – property of connected learner having dimensions
satisfaction with study – property of connected learner having
dimensions
the propensity to study’ – property of connected learner having
dimensions.
cost of failure – property of connected learner having dimensions

 Juggler, Struggler, Fade-away and Leaver – typology of connected
learners defined by the interrelationships between the learner’s
personal commitment structure and his/her personal
competencies.

Connected learning – related category and structural condition
Time design – sub category having properties (e.g. structure
points)

Studying: BSSP relevant stage – ‘Doing the study’

Theoretical codes which organise the substantive codes are: a
Typology, two Basic Social Processes (BSP); i.e. a Basic Social
Structural Process (BSSP) (studying) and a Basic Social
Psychological Process (BSPP) (temporal integration) where the
codes Balancing, Cycling, Amplifying Causal Loop and Cutting
Point organise the movement and flow of the process.

Author

Helen Scott, Ph.D.
Lead Consultant
Grounded Theory Online
Email: helen@groundedtheoryonline.com

1 The area of interest is online distance learning for adults (from the perspective
of adult online learners). Here I use the wording substantive population to mean
the people in/of the substantive area.
2 Collecting data online for a Grounded Theory study has its own issues which
are discussed in a separate paper (in preparation).
3 I understand an incident as a section of data in a source document/field notes.
As an act in the process of coding, I label the incident. A label with only one
incident does not earn the status of substantive code. It is only if other incidents
join this incident under the same label that a code emerges. At this point of
emergence, incidents earn the status of being indicators of a code. Thus not all
that I label becomes a code. This matters later when I think I am overwhelmed by
codes – many are not, they are simply labels!
4 I used ‘method memos’ to record my tussles with learning and applying the
Grounded Theory method.
5 I am indebted to Judith Holton for this suggestion
6 This meant talking to undergraduates studying online as well as postgraduates
and to people studying on differing online vocational courses. I also sampled
outside of the substantive area and collected data in person and in literature
review on on-campus online learners, correspondence distance learners, and parttime
face-to-face learners. For the purpose of my thesis the area of interest was
limited to online distance learning although the resultant theory has relevance for
all adult part time vocational learning.
7 However, many of these were not codes but were one incident labels.
8 Eventually, the homeostasis of motivation emerged – as the propensity to study
– to become the feedback loop of the BSPP ‘Integrating study into a structured
life’.

References

Dupin-Bryant, P. (2004). Pre-entry Variables Related to
Retention in Online Distance Education. The American
Journal of Distance Education 18(4), 199 -206.

Glaser, B. (1978). Theoretical Sensitivity. Mill Valley: Sociology
Press

Glaser, B. (1998). Doing Grounded Theory: Issues and
Discussions. Mill Valley: Sociology Press.

Glaser, B. (2005a). The Grounded Theory Perspective III:
Theoretical Coding. Mill Valley: Sociology Press.

Glaser, B. (2005b). A Teachable Moment. Grounded Theory
Research Seminar, Stockholm

Glaser, B., & Holton, J. (2005). Basic Social Processes. The
Grounded Theory Review, 4(3), 1 – 27.

Kember, D. (1999). Integrating Part-time Study with Family,
Work and Social Obligations. Studies in Higher
Education, 24(1), 109 – 124

Scott, H. (2007a). The Temporal Integration of Connected Study
into a Structured Life: A Grounded Theory. In B. G.
Glaser & J. A. Holton (Eds.), The Grounded Theory
Seminar Reader 2007. Mill Valley, California: Sociology
Press.

Scott, H. (2007b). The Temporal Integration of Connected Study
into a Structured Life: A Grounded Theory. University of
Portsmouth, Portsmouth.

Scott, H. (in preparation) Data Collection Online

Shin, N., & Kim, J. (1999). An exploration of learner progress and
drop-out in Korea National Open University. Distance
Education, 20(81-95).

Strauss, A. & Corbin, J (1998) Basics of Qualitative Research:
Techniques and Procedures for Developing Grounded
Theory’ London: Sage Publications Ltd.

Thulesius, H. (2003). Balancing Cancer from a primary care
perspective. Diagnosis, postraumatic stress and end-of-life
care. Unpublished Doctoral thesis, Lund University,
Lund.

Wenger, E. (1998). Communities of Practice: Learning, Meaning,
and Identity Cambridge: Cambridge University Press.

Woodley, A. (2004). Conceptualizing student dropout in part-time
distance education: pathologizing the normal? Open
Learning, 19(1), 47-64.

Facebooktwittergoogle_plusredditpinterestlinkedinmail