Methodological Learning-by-doing: Challenges, lessons learned and rewards

Pernilla Pergert, RN, Ph.D.

Abstract

The experience of minus mentoring in learning classic grounded
theory (CGT) is shared by many people over the world. The aim of
this article is to share experiences of learning and using CGT.
Data for the article included methodological discussions in the
author’s thesis and articles, as well as memos. Consequences of
learning grounded theory by doing are presented in the form of
challenges and lessons learned but also some rewards.
Challenges and lessons learned include sampling-confusion,
delimiting-disregarding, judging saturation and conceptual
language-struggling. Rewards include trusting the method,
insider-researcher and expert-resourcing. Presented rewards
could be seen as advice and inspiration for novice GT researchers.

Introduction

Grounded theory (GT) is an inductive method, useful and
suitable for qualitative data. It is highly appropriate for nursing
research (Nathaniel & Andrews, 2007; Schreiber & Stern, 2001)
and aims to discover a main concern of participants and how they
manage and resolve such concern (Glaser, 1978). GT was
formulated by Glaser & Strauss (1967) and elaborated by Glaser
(1978, 1998), Strauss and Corbin (1998), and others. The method
elaborated by Glaser is often called classic grounded theory
(CGT). Researchers need to choose not only what method to use
but also what approach (Heath & Cowley, 2004), remodeling
(Glaser & Holton, 2004) or even synthesis of approaches (Chen &
Boore, 2009).

The aim of postgraduate studies is to get a deeper
understanding of both the subject and scientific methodology
(Karolinska Institutet, 2007). A situation in which no expert is
present to teach and guide in GT methodology is known as minus
mentoring (Glaser, 1998, p. 5; Stern, 1994). Experience of such a
situation is shared by many people over the world. One challenge
with minus mentoring is that informed formative feedback, given
during the process in order to enhance learning (Biggs & Tang,
2007), may be lacking.

When my research education started, I did not know much
about CGT. One of my supervisors had supervised an earlier
thesis using a “grounded theory approach” (Baarnhielm, 2003, p.
47 ); the other two supervisors had no experience in using GT,
though their attitude to the method was positive. In choosing the
CGT method, my main concern was to perform good research
while learning-by-doing.

The aim of this article is to share experiences of learning and
using CGT. Memos as well as methodological discussions in my
thesis and articles have been used as the basis for this discussion.
The various categories, presented in the text below, are further
illustrated with examples from my experience. The examples are
taken from the my thesis (Pergert, 2008) and the four studies
included there, referred to throughout this article by their Roman
numerals I – IV.

Methodological Learning-by-Doing

This refers to the capability to acquire methodological skills
and understanding while using the method and doing research.
Consequences of learning grounded theory by doing include
challenges and lessons learned but also rewards.

Challenges and lessons learned

In this section, some challenges and lessons learned, from
my experience in using GT and learning-by-doing, will be
presented, including sampling-confusion, delimiting-disregarding,
judging saturation and conceptual language-struggling.

Sampling-confusion

The initial decisions for sampling in GT are based on the
general subject area (Glaser & Strauss, 1967, p. 45). This is
similar to purposive sampling in the sense that it aims to include
people who are knowledgeable about the subject being studied
(Polit & Hungler, 1999). In GT, this initial sampling should be
followed by theoretical sampling of comparative groups and
literature. “Theoretical sampling is the process of data collection
for generating theory whereby the analyst jointly collects, codes,
and analyzes his data and decides what data to collect next and
where to find them, in order to develop his theory as it emerges.
This process of data collection is controlled by the emerging
theory” (Glaser & Strauss, 1967, p. 45). Since theoretical
sampling proceeds in tandem with CGT’s joint processes of data
collection and analysis, it could easily be confusing. A study that
does not theoretically sample to sample other groups would still
use the joint procedures for data collection and analysis;
analyzing data between interviews to influence questions in
subsequent interviews to further elaborate the emerging
categories. Furthermore, the researcher needs to decide how large
the sample should be from the initial group. If the study is a part
of a dissertation project, the initial group might be sufficient for
one study, so there would be no sampling from other groups.
However, full use of theoretical sampling is important in
developing the theory. For example, my third study (Pergert,
Ekblad, Enskar, & Bjork, 2008b) is the one where I theoretically
sampled beyond the initial group and it is by far the best GT
study in the thesis.

Delimiting-disregarding

GT has delimiting tools for data collection. Some of these
delimiting tools may be disregarded in studies, thus contributing
to data-wallowing, which needs to be dealt with. In CGT, field
notes are recommended in data collection rather than recording
since the latter will undermine delimiting (Glaser, 1998). There
could be several reasons for not following this recommendation.
One is that recording is more acceptable to the scientific
community, since field notes are often viewed as selective and
biased (Glaser, 1998). Other reasons for recording interviews
could be to have transcriptions for comparative analysis,
quotations for illustrating various points as well as to enhance
transparency in the supervision of the scientific work. However,
this decision will contribute to data-wallowing, which can lead to
premature closure of data collection and a lack of theoretical
sampling, leading to a lack of conceptual depth.

Another methodological choice is the data collection method.
In GT, common methods are interviews and observations. Focus
group interviews, which I used in studies I & II (Pergert, Ekblad,
Enskar, & Bjork, 2007, 2008a), may not be the preferred method
for data collection in CGT, but it was argued that “the process of
generating theory is independent of the kind of data used”
(Glaser & Strauss, 1967, p. 18). Focus group interviews were
found to be highly relevant for collecting data in the beginning of
a GT but not as good later in the process, since theoretical
focusing and delimiting can be hard. One problem is the large
amount of data from a few focus groups, which can delay the start
of theoretical sampling and lead to premature closure of data
collection, leading to a lack of conceptual depth.

As delimiting-disregarding will result in data-wallowing, it
needs to be dealt with somehow, often by using different software
programs for handling qualitative data. In the thesis, the
software program QSR NVivo 2.0 was used as a tool (QSR
International, 2002) to manage the data. Being relatively
computer literate, I was soon engrossed in learning the software.
However, becoming too enthusiastic, I started to use it for
organizing code trees and creating models (Bazeley & Richards,
2000), which is not consistent with CGT. The main concern did
emerge but the use of code trees and models may have
preconceived an outline rather than letting the integration
emerge in later sorting of memos (Glaser, 1998). In subsequent
studies I used the software only as a coding tool, to deal with
data-wallowing in a more GT congruent way. Even though the
ambition was to use full GT procedures in every study of the
thesis, different conceptual levels were reached in the analysis.
This could be related to my GT learning curve and a lack of
theoretical focusing and delimiting relating to delimitingdisregarding.

Judging saturation

There is no such thing as an ideal sample size in GT;
instead, size is based on saturation (Glaser, 1998; Glaser &
Strauss, 1967). That is, sample size is based on a judgment, in
coding and analyzing, of theoretical saturation of categories,
which implies that “no new properties emerge and the same
properties continually emerge” (Glaser, 1978, p. 53) and that gaps
in major categories are more or less filled (Glaser & Strauss,
1967). Saturation is always a subjective judgment and the
decision to stop theoretical sampling, using the methodological
guidelines, is always influenced by the scope of the research
project, particularly in terms of time and resources. This
judgment is a real challenge and the outcome could always be
different; further theoretical sampling can usually be motivated.
For example, in study I (Pergert, et al., 2007) it was decided not
to theoretically sample for, and saturate, the subcategory of trustbuilding,
since it could probably be a whole theory in itself. In
study IV (Pergert, Ekblad, Bjork, Enskar, & Andrews, n.d),
theoretical sampling could have been carried out among Swedishborn
parents but it was decided to use literature for comparative
analysis, which is consistent with GT methodology (Glaser, 1998).
Furthermore, only one father was included in the individual
interviews with foreign-born parents. This did not emerge as
relevant but could be seen as a limitation, as support-seeking, one
of the subcategories, is used more frequently among women
(Norberg, Lindblad, & Boman, 2006). Theoretical sampling was
delimited by a judgment about the scope of the dissertation
project; other comparative groups could always be sampled. In all
of the studies, further theoretical sampling, bringing in new
relevant data from new fields, would undoubtedly lead back to
theoretical non-saturation (Glaser, 1978) and modification, but
then again, a GT is always modifiable.

Conceptual language-struggling

Conceptualization is central in GT. The name of the core
category should have “grab” (capture attention) and often takes
the form of a gerund (ending in -ing) to bring out its nature of
explaining a behavior; managing and resolving the main concern
(Glaser, 1978, 1998). Naming a category with grab in a language
that is not one’s mother tongue is a challenge, as nuances and
subtle meanings are easily missed. Often categories need to be
named in two languages and sometimes translation can be a
problem. For example, I named categories in the analysis in both
English and Swedish, the latter being my native tongue.
However, the use of gerund verbs is characteristic of GT but the
Swedish language lacks the gerund verb form, so the core is often
named as an infinitive, for instance “to bridge” instead of
“bridging”. In the search for the best possible names of categories,
I discussed my choices with English text editors; this was a great
help but these text editors were not acquainted with the method
and seldom enthusiastic about new conceptual gerund names.

Rewards

Rewards from using and learning GT will be presented
below. These rewards, which could be seen as advice and
inspiration for novice GT researchers, include: trusting the
method, insider-researching and expert-resourcing.

Trusting the method

In CGT, emergence of theory is central (Glaser, 1998). The
concept of emergence may sound unscientific and strange but it is
simply a matter of trusting that what is going on in the empirical
field will emerge from the data (Glaser, 1998). The aim of CGT is
to let the participants’ main concern emerge, instead of focusing
on what Glaser (1998) calls “professional concern” (p. 99) or a
“professionally preconceived problem” (p. 118). The inductive
emergence will guarantee a good fit; that is, the theory will
adequately express what is happening in the empirical situation,
and be highly relevant. For example, when I began my first study
(Pergert, et al., 2007) my preconceived notion was that the study
would be concerned with medical information; this was also
healthcare staff’s spontaneous answer to what was the biggest
challenge in transcultural care. As data were analyzed, using GT,
what emerged instead as the core was bridging obstacles to
transcultural caring relationships. Bridging is what Glaser (1998,
p. 5) calls the ‘latent pattern’ of behavior, of which participants
are not necessarily aware. The preconceived notion of giving
medical information was something that healthcare staff was
aware of, whereas obstacles to transcultural caring relationships
were actually a major concern. This major concern would
probably not have been identified with a method that focuses
more on predefined problems, testing hypotheses, and using
preset and narrower questions. For example, Strauss and Corbin
(1998) proposed that professional experience and suggestions
could be used to identify the research problem. Experiencing
emergence is a most rewarding moment; after working in
complete confusion with masses of data; you finally discern the
pattern or the core. It is a great advantage to use and trust a
well-tried methodology, especially when doubting one’s own
capacity.

Insider-researching

Glaser (1998) holds that research is easier to do where you
know nothing about the substantive area under study; on the
other hand, doing research in a familiar area leads to motivation
and more variables to deal with. Insider research is common in
the qualitative field in the context of nursing (Asselin, 2003;
Cudmore & Sondermeyer, 2007) and could be seen as an
advantage, as the double role may enhance trust in the
interviewer (Glaser & Strauss, 1967); positively influence the
relationship (Asselin, 2003) and consequently also the data. This
was, for me, a rewarding experience in working as an “insider
researcher”. Further, my motivation and drive as an insider were
far more important than doing easy research.

Expert-resourcing

Gathering resources to compensate for minus mentoring
includes information-seeking by using the literature on CGT and
feedback-seeking while learning by using available resources
such as research groups, experienced grounded theorists and
reviewer feedback on manuscripts – not always fun to get but
sometimes very helpful indeed. When I had written my first
manuscript, I heard of a GT troubleshooting seminar arranged by
the Grounded Theory Institute (2005a) in Stockholm. To prepare
for the seminar, I read three books on GT (Glaser, 1978, 1998;
Glaser & Strauss, 1967) and realized that I had been on the
wrong track. Following the seminar, I reworked the manuscript,
now with a new core category and an attempt to write more
conceptually. Since then, I have had the opportunity to
participate in several international GT seminars (Grounded
Theory Institute, 2005b, 2006, 2007) given by GT experts,
providing great assistance in naming core categories, taking the
analysis several steps further and contributing immensely to the
GT learning process.

Discussion

In this article, challenges and lessons learned as well as
rewards in learning GT by doing have been presented. Expertresourcing
and feedback-seeking are recommended in this article,
in accordance with educational and methodological literature that
states that feedback is necessary to enhance learning (Biggs &
Tang, 2007; Glaser, 1998). Feedback-seeking could include
feedback from reviewers of manuscripts, using errors for learning
(Biggs & Tang, 2007). However, it is crucial to consider how the
feedback is given and received, especially if there is a discrepancy
in the understanding of GT as held by the reviewer and the
author. This is often the case when CGT is confused with other
approaches referred to as GT.

In regard to learning-by-doing, one could argue that research
education should be organized differently, with much more
methodological study before starting the research project.
Equally, it could be argued that the only way to learn a method is
by using it and in doing so, develop one’s skill. Learning-by-doing
is similar, or rather has an ingredient of just-in-time learning,
which is about learning something when one needs to (Biggs &
Tang, 2007), thus giving motivation and relevancy.
Methodological studies that are not needed for the task at hand
are often perceived as boring and irrelevant. In the words of
Glaser, “Just do it” (1998, p. 254) and “Trust Grounded theory, it
works!” (1998, p. 254).

Acknowledgements

I would like to thank supervisors, colleagues in different
research groups, and GT experts who have contributed to my GT
learning; and Patrick Hort for revising the English text. The
research was funded by the Swedish Children’s Cancer
Foundation.

Author

Pernilla Pergert, RN, Ph.D.
Department of Woman and Child Health
Karolinska Institutet
Stockholm, Sweden
E-mail: pernilla.pergert@ki.se

References

Asselin, M. E. (2003). Insider research: Issues to consider when
doing qualitative research in your own setting. Journal for
Nurses in Staff Development, 19(2), 99-103.

Baarnhielm, S. (2003). Clinical encounters with different illness
realities: Qualitative studies of illness meaning and restructuring
of illness meaning among two cultural groups of female patients
in a multicultural area of Stockholm. (Dissertation) Stockholm:
Karolinska Institutet.

Bazeley, P., & Richards, L. (2000). The NVivo qualitative project
book. London; Thousand Oaks, CA: SAGE Publications.

Biggs, J., & Tang, C. (2007). Teaching for quality learning at
university (3. Ed.). Two Penn Plaza, New York, NY: McGraw-
Hill/Society for Research into Higher Education, Open University
Press.

Chen, H. Y., & Boore, J. R. (2009). Using a synthesised technique
for grounded theory in nursing research. Journal of Clinical
Nursing.

Cudmore, H., & Sondermeyer, J. (2007). Through the looking
glass: Being a critical ethnographic researcher in a familiar
nursing context. Nurse Researcher, 14(3), 25-35.

Glaser, B. G. (1978). Theoretical sensitivity: Advances in the
methodology of grounded theory. Mill Valley, CA: Sociology Press.

Glaser, B. G. (1998). Doing grounded theory: Issues and
discussions. Mill Valley, CA: Sociology Press.

Glaser, B. G., & Holton, J. (2004). Remodeling Grounded Theory
[80 paragraphs]. Forum: Qualitative Social Research [On-line
Journal], 5(2), Art.4. Retrieved May 13, 2009,
from:http://www.qualitativeresearch.
net/index.php/fqs/article/view/2607

Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded
theory: Strategies for qualitative research. New York: Aldine.

Grounded Theory Institute (2005a). Grounded theory seminar.
July 4-5, Stockholm, Sweden.

Grounded Theory Institute (2005b). Grounded theory seminar.
Oct. 25-27, New York, USA.

Grounded Theory Institute (2006). Grounded theory seminar.
April 25-26, London, UK.

Grounded Theory Institute (2007). Grounded theory seminar.
April 24-26, London, UK.

Heath, H., & Cowley, S. (2004). Developing a grounded theory
approach: A comparison of Glaser and Strauss. International
Journal of Nursing Studies, 41(2), 141-150.

Karolinska Institutet (2007). Rules for postgraduate-level
education at Karolinska Institutet Retrieved 11 May, 2009, from
http://ki.se/content/1/c4/84/60/Rules%20070628_080930.pdf

Nathaniel, A. K., & Andrews, T. (2007). How grounded theory can
improve nursing care quality. Journal of Nursing Care Quality,
22(4), 350-357.

Norberg, A. L., Lindblad, F., & Boman, K. K. (2006). Supportseeking,
perceived support, and anxiety in mothers and fathers
after children’s cancer treatment. Psycho-Oncology, 15(4), 335-
343.

Pergert, P. (2008). Façading in transcultural caring relationships:
Healthcare staff and foreign-born parents in childhood cancer
care. (Dissertation) Stockholm: Karolinska Institutet.

Pergert, P., Ekblad, S., Bjork, O., Enskar, K., & Andrews, T. (n.d).
Struggling on: A grounded theory from interviews with foreignborn
parents in childhood cancer care. Submitted for publication.

Pergert, P., Ekblad, S., Enskar, K., & Bjork, O. (2007). Obstacles
to transcultural caring relationships: Experiences of health care
staff in pediatric oncology. Journal of Pediatric Oncology Nursing,
24(6), 314-328.

Pergert, P., Ekblad, S., Enskar, K., & Bjork, O. (2008a). Bridging
obstacles to transcultural caring relationships—Tools discovered
through interviews with staff in pediatric oncology care.
European Journal of Oncology Nursing, 12(1), 35-43.

Pergert, P., Ekblad, S., Enskar, K., & Bjork, O. (2008b).
Protecting Professional Composure in Transcultural Pediatric
Nursing. Qualitative Health Research, 18(5), 647-657.

Polit, D. F., & Hungler, B. P. (1999). Nursing research : principles
and methods (6. Ed.). Philadelphia: Lippincott Williams &
Wilkins.

QSR International (2002). QSR NVivo 2.0 New generation
software for qualitative data analysis. Doncaster, Australia: ©
QSR International Pty Ltd.

Schreiber, R. S., & Stern, P. N. (2001). Using grounded theory in
nursing. New York: Springer.

Stern, P. N. (1994). Eroding grounded theory. In J. M. Morse
(Ed.), Critical issues in qualitative research methods (pp. 212-
223). Thousand Oaks, CC: Sage.

Strauss, A. L., & Corbin, J. M. (1998). Basics of qualitative
research: Techniques and procedures for developing grounded
theory (2. Ed.). Thousand Oaks, CA: SAGE.