methodological issues in grounded theory

9
Methodological issues in grounded theory John R. Cutcliffe RMN RGN BSc(Hons) Doctoral Student, Lecturer in Mental Health Nursing and Practice Development Co-ordinator, Sheffield University, Sheffield, and RCN Institute, Oxford, England Accepted for publication 9 December 1999 CUTCLIFFE CUTCLIFFE J.R. (2000) (2000) Journal of Advanced Nursing 31(6), 1476–1484 Methodological issues in grounded theory Examination of the qualitative methodological literature shows that there appear to be conflicting opinions and unresolved issues regarding the nature and process of grounded theory. Researchers proposing to utilize this method would therefore be wise to consider these conflicting opinions. This paper therefore identifies and attempts to address four key issues, namely, sampling, creativity and reflexivity, the use of literature, and precision within grounded theory. The following recommendations are made. When utilizing a grounded method researchers need to consider their research question, clarify what level of theory is likely to be induced from their study, and then decide when they intend to access and introduce the second body of literature. They should acknowledge that in the early stages of data collection, some purposeful sampling appears to occur. In their search for conceptually dense theory, grounded theory researchers may wish to free themselves from the constraints that limit their use of creativity and tacit knowledge. Furthermore, the interests of researchers might be served by attention to issues of precision including, avoiding method slurring, ensuring theoretical coding occurs, and using predominantly one method of grounded theory while explaining and describing any deviation away from this chosen method. Such mindfulness and the resulting methodological rigour is likely to increase the overall quality of the inquiry and enhance the credibility of the findings. Keywords: grounded theory, methodology, rigour, nursing, sampling, creativity, reflexivity, precision INTRODUCTION Since Glaser and Strauss’ discovery in 1967, and its application within sociological study, grounded theory has been used in many other fields including anthropol- ogy and nursing. Many authors have written about the method, and scrutiny of this literature shows that there appear to be conflicting opinions and unresolved issues regarding the nature and process of grounded theory. Researchers proposing to utilize this method would therefore be wise to consider these conflicting opinions. This paper begins with a brief overview of grounded theory in order to identify the rudiments of the method. Then it identifies and addresses four key issues, namely, sampling, creativity and reflexivity, the use of literature, and precision within grounded theory. Brief overview of grounded theory A grounded theory is a theory that is induced from the data rather than preceding them (Lincoln & Guba 1985). Glaser & Strauss (1967 p. 3) provide an initial definition of grounded theory. They state that it is a theory that will: Correspondence: John R. Cutcliffe, 11 Blackthorne Close, Kilburn, Derbyshire DE56 0LF, England. E-mail: john.cutcliffe@sheffield.ac.uk 1 Journal of Advanced Nursing, 2000, 31(6), 1476–1484 Methodological issues in nursing research 1476 Ó 2000 Blackwell Science Ltd

Upload: stikes-merangin-jambi

Post on 12-Nov-2014

2.730 views

Category:

Documents


1 download

DESCRIPTION

 

TRANSCRIPT

Page 1: Methodological issues in grounded theory

Methodological issues in grounded theory

John R. Cutcliffe RMN RGN BSc(Hons)

Doctoral Student, Lecturer in Mental Health Nursing

and Practice Development Co-ordinator, Shef®eld University,

Shef®eld, and RCN Institute, Oxford, England

Accepted for publication 9 December 1999

CUTCLIFFECUTCLIFFE JJ.RR. (2000)(2000) Journal of Advanced Nursing 31(6), 1476±1484

Methodological issues in grounded theory

Examination of the qualitative methodological literature shows that there

appear to be con¯icting opinions and unresolved issues regarding the nature

and process of grounded theory. Researchers proposing to utilize this method

would therefore be wise to consider these con¯icting opinions. This paper

therefore identi®es and attempts to address four key issues, namely, sampling,

creativity and re¯exivity, the use of literature, and precision within grounded

theory. The following recommendations are made. When utilizing a grounded

method researchers need to consider their research question, clarify what level

of theory is likely to be induced from their study, and then decide when they

intend to access and introduce the second body of literature. They should

acknowledge that in the early stages of data collection, some purposeful

sampling appears to occur. In their search for conceptually dense theory,

grounded theory researchers may wish to free themselves from the constraints

that limit their use of creativity and tacit knowledge. Furthermore, the interests

of researchers might be served by attention to issues of precision including,

avoiding method slurring, ensuring theoretical coding occurs, and using

predominantly one method of grounded theory while explaining and describing

any deviation away from this chosen method. Such mindfulness and the

resulting methodological rigour is likely to increase the overall quality of the

inquiry and enhance the credibility of the ®ndings.

Keywords: grounded theory, methodology, rigour, nursing, sampling, creativity,

re¯exivity, precision

INTRODUCTION

Since Glaser and Strauss' discovery in 1967, and its

application within sociological study, grounded theory

has been used in many other ®elds including anthropol-

ogy and nursing. Many authors have written about the

method, and scrutiny of this literature shows that there

appear to be con¯icting opinions and unresolved issues

regarding the nature and process of grounded theory.

Researchers proposing to utilize this method would

therefore be wise to consider these con¯icting opinions.

This paper begins with a brief overview of grounded

theory in order to identify the rudiments of the method.

Then it identi®es and addresses four key issues, namely,

sampling, creativity and re¯exivity, the use of literature,

and precision within grounded theory.

Brief overview of grounded theory

A grounded theory is a theory that is induced from the

data rather than preceding them (Lincoln & Guba 1985).

Glaser & Strauss (1967 p. 3) provide an initial de®nition of

grounded theory. They state that it is a theory that will:Correspondence: John R. Cutcliffe, 11 Blackthorne Close, Kilburn,

Derbyshire DE56 0LF, England. E-mail: john.cutcliffe@shef®eld.ac.uk1

Journal of Advanced Nursing, 2000, 31(6), 1476±1484 Methodological issues in nursing research

1476 Ó 2000 Blackwell Science Ltd

Page 2: Methodological issues in grounded theory

¼ ®t the situation being researched and work when put into use.

By ®t we mean that the categories must be readily (not forcibly)

applicable to and indicated by the data under study; by work we

mean that they must be meaningfully relevant and be able to

explain the behaviour under study.

It is rooted in symbolic interactionism, wherein the

researcher attempts to determine what symbolic mean-

ings, artifacts, clothing, gestures and words have for

groups of people as they interact with one another.

Symbolic interactionists stress that people construct their

realities from the symbols around them through interac-

tion, therefore individuals are active participants in

creating meaning in a situation (Morse & Field 1995).

Grounded theory both describes and explains the system

or behaviour under study and consequently is a method-

ology for developing theory that is grounded in data

systematically gathered and analysed (Strauss & Corbin

1994). Consequently, grounded theorists search for social

processes present in human interaction (Hutchinson

1993). They aim to discover patterns and processes and

understand how a group of people de®ne, via their social

interactions, their reality (Stern et al. 1982).

A central feature of grounded theory is its method of

constant comparative analysis (Glaser & Strauss 1967), in

that data collection and analysis occur simultaneously

and each item of data is compared with every other item of

data. The theory induced is conceptually dense (Strauss &

Corbin 1994), that is theory with many conceptual rela-

tionships, and these relationships are embedded in a

context of descriptive and conceptual writing.

SOME RESEARCH DESIGNS

Sampling in grounded theory

Grounded theory uses non-probability sampling. In order

for concepts and categories to emerge during the data

analysis, the need for sampling of speci®c data sources

continues until each category is saturated. Therefore, at

the beginning of the study, there are no limits set on the

number of the participants, interviewees or data sources.

The researcher continues selecting interviewees until they

are saying nothing new about the concepts being explored.

Thus the selection of participants (and other sources of

data) is a function of the emerging hypothesis/hypotheses

and the sample size a function of the theoretical complete-

ness (Baker et al. 1992).

Sampling within grounded theory is therefore described

as `theoretical' rather than purposeful (Glaser & Strauss

1967, Glaser 1978, Becker 1993) in that it is driven by the

emerging theory. However, other authors of qualitative

research methods do not make such a distinction (Lincoln

& Guba 1985, Morse 1991b). Indeed, they suggest that the

terms theoretical and purposeful sampling are inter-

changeable. Patton (1990) argues that all types of sampling

in qualitative research can be termed purposeful

sampling. Interestingly, in his list of 15 different sampling

strategies Patton does not list or de®ne theoretical

sampling. This apparent confusion only serves to confuse

neophyte qualitative researchers (Coyne 1997). However,

if the researcher can describe his/her sampling strategy in

suf®cient detail, this should minimize any confusion

regarding sampling (Morse 1991b), improve the quality of

the research (Coyne 1997), avoid method slurring (Baker

et al. 1992) and provide some clari®cation of the use of

theoretical sampling in nursing research.

Glaser (1978), Sandelowski et al. (1992), Becker (1993)

and Coyne (1997) each delineated theoretical sampling

from purposeful/selective sampling, in as much that,

purposeful sampling involves the calculated decision to

sample a speci®c locale according to a preconceived but

reasonable initial set of dimensions. In contrast, theoret-

ical sampling has no such initial calculated decisions. The

grounded theory researcher seeks further interviewees/

sources of data in order to add to the fullness of the

understanding of the concept. Hence, theoretical sampling

is an integral part of the process of grounded theory.

However, it should be noted that before the researcher has

begun to collect and analyse data, the researcher has no

evolving theory which can act as a guide for further

theoretical sampling.

Baker et al. (1992) maintained that the researcher using

grounded theory initiates the sampling process by inter-

viewing signi®cant individuals. Perhaps it is these signi-

®cant individuals that Morse (1991b) is referring to when

she describes a good informant as one who has the

knowledge and the experience the researcher requires, has

the ability to re¯ect, is articulate, has the time to be

interviewed and is willing to participate in the study.

Glaser (1978) asserted that in the initial stages of theoret-

ical sampling, decisions for collection of data are based

only on a general sociological perspective and on a general

subject or problem area. Morse (1991b) submitted that the

researcher initially chooses interviewees with a broad

general knowledge of the topic.

These positions and arguments thus appear to indicate

that individuals are chosen initially who can provide a

relevant source of data, and this relevance is determined by

the requirements for generating and delimiting the theoret-

ical codes (Hutchinson 1993). Therefore, when a grounded

theorist is commencing his/her data collection, it appears

that they do enter into a process of purposeful sampling,

which is then superseded by theoretical sampling as the

data/theory highlight the direction which further sampling

needs to follow. This argument is supported by Sande-

lowski et al. (1992) and Coyne (1997 p. 625) who states:

¼ theoretical sampling does involve the purposeful selection of a

sample in the initial stages of the study.

Methodological issues in nursing research Methodological issues in grounded theory

Ó 2000 Blackwell Science Ltd, Journal of Advanced Nursing, 31(6), 1476±1484 1477

Page 3: Methodological issues in grounded theory

A second issue in sampling that warrants attention is that

of the choice between a wide and diverse sample or a more

`focused', narrow, concentrated sample. It is reasonable to

say that the literature on this issue is confusing and

con¯icting. Hutchinson (1993) argues in favour of a wide,

diverse sample in that this ensures extensive data that cover

the wide ranges of behaviour in varied situations. Lincoln &

Guba (1985) make similar arguments. They advocate, since

the purpose of sampling will most often be to include as

much information as possible, maximum variation

sampling to be the usual sampling mode of choice.

Another argument may be constructed that reasons in

favour of a more narrow or focused sample, rather than

maximum variation. Since the researcher in grounded

theory is concerned with uncovering the situated, contex-

tual, core and subsidiary social processes, the social

processes need to be shared and experienced by the

individuals who make up the researched group. Other-

wise, if an individual has no experience of the social or

psychosocial process, how can they comment on it?

Consequently, grounded theorists using a more narrow

or focused sample seek out participants who have experi-

ence, the most experience, in the topic of interest (Morse

1998). Indeed, Lincoln & Guba (1985) point out that

grounded theory has been termed `local theory' as it brings

together and systematizes isolated, individual theory. It is

an aggregate of local understandings. Selection of a sample

of participants who have only a limited experience of the

social process, or put another way, a sample that isn't

local, is thus likely to provide data and a subsequent

theory that has a partial or limited understanding of the

process being studied. Glaser & Strauss (1967) highlight

how the choice between sampling narrow or wider

substantive groups is directed by the conceptual level of

the theory that the researcher intends to induce. They

indicated that if the researcher intends to induce a

substantive theory that is applicable to one substantive

group, then the researcher needs to sample groups of the

same substantive type (e.g. a narrow sample). A more

general, or wider substantive theory would thus be

induced by sampling wider substantive groups, and if

the researcher is concerned with inducing a formal theory,

he/she will select dissimilar substantive groups from the

larger class, and thus increase the theory's scope.

Lincoln & Guba (1985) argue that the purpose of

maximum variation within theoretical sampling is best

achieved by selecting each unit of the sample only after

the previous unit has been taped and analysed. Conse-

quently the ®rst unit of the sample (®rst interviewee) often

acts as a `gatekeeper'. This ®rst set of data and subsequent

analysis can set the `tone' or highlight the direction for

further theoretical sampling. The notion of a `gatekeeper'

again raises the sampling issues identi®ed in this paper. If

the ®rst unit of a sample only has a limited experience of

the social process being studied, one could argue that the

subsequent theoretical sampling would also re¯ect the

limited experience. The data obtained from such an

individual, when analysed and coded, is unlikely to

indicate the lines of inquiry, the necessary sources of data,

which would then lead ultimately to the `fullest' or most

complete understanding of the social process. It should be

noted that it is possible that the fullness of the phenom-

enon may be uncovered during subsequent interviews

(Hutchinson 1993). However, this argument does appear

to highlight the importance of selecting an appropriate

gatekeeper, and therefore also indicates a degree of

a-priori sample selection.

Yet Lincoln & Guba (1985) deliberate that there can be

no a-priori speci®cation of the sample and that initially

any sample unit will do as well as any other. Nevertheless,

Morse (1991b p. 129) highlighted that selection of an

adequate and appropriate sample is critical in qualitative

research and that the eventual quality of the research is

contingent upon the appropriateness and adequacy of the

sample. Furthermore she states:

¼ it is essential for the researcher to discover who will be the

most appropriate informant before beginning interviews and that

informants must be carefully selected or carefully chosen

according to speci®c qualities.

She sums up her arguments regarding sample selection

by stating:

¼ informants must be knowledgeable about the topic and experts

by their virtue of their involvement in speci®c life events and/or

associations.

The author of this current paper is not advocating strict

adherence to sample criteria set prior to commencing data

collection. Such rigidity is only likely to limit theoretical

sampling (Morse 1991b). Yet, it appears to be logical for

the researcher to consider criteria for sample selection

prior to starting to collect data. Therefore the author

argues that this purposeful sampling should be considered

for the ®rst interview (especially as this individual occu-

pies the role of `gatekeeper') and possibly the second

interview. Following this theoretical sampling to guide

sample selection would be more appropriate.

Creativity and re¯exivity in grounded theory

A further issue that warrants consideration is that of

creativity and re¯exivity. Few would dispute that qualit-

ative methods invariably involve interaction between the

researcher and the data. Turner (1981) reasoned that in

social inquiry there is an interaction between the

researcher and the world that they are studying. Indeed,

Lipson (1991) expounded that re¯exivity refers to

researchers being part of, rather than separate from, the

data. Altheide & Johnson (1994) argue that theories

induced from qualitative methods always include some-

J.R. Cutcliffe

1478 Ó 2000 Blackwell Science Ltd, Journal of Advanced Nursing, 31(6), 1476±1484

Page 4: Methodological issues in grounded theory

thing of the researcher. Hutchinson (1993 p. 187) made

speci®c remarks concerning re¯exivity and grounded

theory:

¼ because grounded theory research requires interpersonal

interaction, the researcher is inevitably part of his or her daily

observations.

However, while there appears to be little argument that

grounded theory inevitably involves interaction between

the researcher and the world they are studying, how this

interaction affects the emerging theory remains a matter

for debate. Morse (1994) contends that qualitative

methods (including grounded theory) have been plagued

with con¯icting advice concerning the application of prior

knowledge (including the researcher's previous experi-

ence and knowledge which they bring to the study).

Berger & Kellner (1981) and Hutchinson (1993) advocate

that the qualitative researcher needs to become aware of

their own personal preconceptions, values and beliefs and

then hold them in abeyance. These authors add that if

these processes are not carried out, the scienti®c enter-

prise collapses and the researcher will perceive a mirror

image of hopes/fears and not the social reality.

A vigorous counter-argument exists that posits it is the

re¯exivity and the researcher's creativity within this

re¯exivity that makes grounded theory valuable. Turner

(1981 p. 227) stated:

¼ an advantage of grounded theory is that it directs the researcher

immediately to the creative core of the research process, and

facilitates the direct application of both the intellect and the

imagination to the demanding process of interpreting research

data.

Stern (1994 p. 217) supports this viewpoint:

¼ it is the creativity in the act that brings the real truth of a social

situation into being, and following grounded theory techniques is

one way to approach this creative process.

Morse (1994) encourages qualitative researchers to take

more risks in their theory development. It is worth

considering whether or not such a movement could

involve transition from a position where concerns for

holding prior knowledge and beliefs in abeyance predom-

inate, to an alternative position more concerned with

creativity. Perhaps consideration of why some researchers

advocate the need to acknowledge and discount any prior

knowledge and beliefs may shed some light on this issue.

It is reasonable to say that for many years qualitative

research methods have been regarded by many scientists

who use quantitative or positivistic methods as a `poor

relation' of quantitative methods. Consequently, some

qualitative researchers have been anxious to be seen as

credible in the eyes of their such scientists. Evidence of

this can be seen in some of the language used to describe

and explain early qualitative methods (Morse 1994). It can

be seen in attempts to establish the credibility of qualit-

ative ®ndings by using criteria developed for establishing

the credibility of quantitative ®ndings (Hammersley 1992).

Also, it can be seen in attempts to translate these quan-

titative criteria into terms more in keeping with the

qualitative paradigm (Cutcliffe & McKenna 1999).

Attempts to discount the researcher's values, know-

ledge, beliefs and experiences could also be regarded as an

attempt to gain credibility with scientists who use quan-

titative or positivistic methods by decreasing any chance

of personal bias. Such endeavours appear to be upholding

the philosophical position that there is one true reality,

and that personal values, knowledge and experience

would only serve to contaminate the researcher's repre-

sentation of this reality.

However, the philosophical position of qualitative

methods is dissimilar to that of quantitative methods.

Qualitative researchers believe that reality is constructed

from human perspectives, shared (social) and individual

interactions and meanings of given situations and

phenomena. To strive to attain more credibility according

to an alternative philosophical standpoint appears to be at

best inappropriate and at worst, a distraction from the

potential that creativity can bring.

That is not to say that the grounded theorist has license

to invent concepts, categories and posit these as a theory

that represents the meanings that a group of individuals

ascribe to their shared interactions and social world.

However, what it does is legitimize the researcher's

creativity as an integral part of the grounded theory

inductive process; liberating the restrictions on the

researcher's tacit knowledge that discounting such know-

ledge creates. Turner (1981 p. 228) stated:

¼ competent development of grounded theory rests in part upon

a sensitivity to these often tacit processes of perceiving and

understanding, and upon a willingness and an ability to bring

them into the open for discussion.

Lincoln & Guba (1985 p. 208) constructed similar

arguments when they stated:

¼ admitting tacit knowledge not only widens the investigator's

ability to apprehend and adjust to phenomenon in context, it also

enables the emergence of theory that could not otherwise have

been articulated.

Hence there is a need for the grounded theory researcher

to acknowledge his/her prior knowledge and tacit know-

ledge, to bring such knowledge into the open, to discuss

how it has affected the theory development (Turner 1981)

and allow the interplay between the researcher's know-

ledge, values and beliefs and the data to occur; to allow

the researcher's creativity to explore and articulate

theoretical links.

The choices of which facts and lines of inquiry to

follow and which not to follow are guided, to some

Methodological issues in nursing research Methodological issues in grounded theory

Ó 2000 Blackwell Science Ltd, Journal of Advanced Nursing, 31(6), 1476±1484 1479

Page 5: Methodological issues in grounded theory

extent, by the subconscious perceptual and intellectual

processes of the researcher's mind (Turner 1981). Conse-

quently, to deny a researcher who is using grounded

theory access to this knowledge and to restrict the

creativity necessary to utilize it, is likely to limit the

depth of understanding of the phenomenon and impose

unnecessary, rigid structures. As a consequence, the

researcher would be left questioning themselves each

time they draw upon their tacit knowledge, or when they

experience a moment of insight into the world they are

investigating. The researcher would be left asking them-

selves:

Does that thought originate from my knowledge, experience or

beliefs or does it belong to the interviewees?

Yet, importantly, the mechanism for checking the

authenticity or representativeness of such knowledge

and insight exists within the grounded theory method,

whereby such trustworthiness is achieved by exploring

the possible or emerging concepts/categories in further

interviews. If the hunch belongs solely to the researcher,

and is not a part of the world being investigated, this will

have no meaning for the interviewees and can be

discarded in due course.

Literature reviews in grounded theory

It is well documented that when utilizing a grounded

theory method the researcher should avoid conducting a

literature review prior to commencing data collection and

analysis (Stern 1980, Stern et al. 1982, Stern & Allen 1984,

Lincoln & Guba 1985, Stern 1994, Strauss & Corbin 1994,

Hickey 1997). By avoiding a literature review at the

beginning of the study it is more likely that the emergent

theory will be grounded in the data.

Another view is proffered by Hutchinson (1993), who

suggests that a literature review should proceed data

collection and analysis in grounded theory. In that, it is

the review of the literature that can identify the current

gaps in knowledge, or help provide a rationale for the

proposed research. Perhaps the apparent disagreement

between these two positions can be explained if one

considers the longitudinal nature of the generation of

knowledge, and the different positions that these argu-

ments appear to occupy on this longitudinal continuum.

Hutchinson's (1993) arguments can be located in a

position that has a starting point: `What do we know about

this phenomenon?' Therefore, at this point, she has not

begun to consider what is the most appropriate method-

ology. That will be indicated by the current extent and

depth of knowledge available regarding the phenomenon,

whereas, Lincoln & Guba's (1985) and Stern's (1994)

argument can be located in a position that has the starting

point, `We already recognize that there is a distinct dearth

or even absence of knowledge concerning the phenom-

enon, and therefore have already decided that a grounded

theory method would be suitable'. Consequently Lincoln

& Guba's (1985) and Stern's (1994) arguments appear to

occupy a position further along the continuum of know-

ledge generation. Given these considerations both argu-

ments appear to be cogent and not necessarily

contradictory of one another.

Many proposed research questions require conceptual

clarity. For example, within the author's doctoral study,

having identi®ed that there is an absence of literature that

explains if or how hope is inspired in bereavement

counselling and thus a grounded theory method would

be indicated, it may still be appropriate and indeed

prudent, to review the available literature that focuses on

hope and the literature that focuses on bereavement

counselling. Such a review may help provide a sense of

the key elements of hope that are implicit in the literature,

it may help provide some conceptual clarity of the nature

of hope and the nature and practice of bereavement

counselling and this examination of the relevant literature

would help the researcher to differentiate hope from

similar and related concepts.

If there is a need for a review of the literature in order to

clarify concepts and de®ne terms, the key questions that

need to be asked appear to be, how rigorous and thorough

should this literature review be and at what point in the

theory induction should this literature review occur?

Smith & Biley (1997 p. 20) acknowledged that a detailed

and comprehensive literature review is not the ®rst stage

in grounded theory. However, they go on to point out that

some reading may occur prior to data collection. They

state:

General reading of the literature maybe carried out to obtain a feel

for the issues at work in the subject area, and identify any gaps to

be ®lled in using grounded theory¼ but it is important that the

reading is not too extensive.

Their justi®cation for this technique is that the

researcher then approaches the subject area with some

background knowledge. It is the opinion of the author of

the current paper that such statements may confuse

potential researchers who are contemplating using

grounded theory. Just how much reading is `extensive'

and similarly `too extensive'? To advocate that the

researcher approaches the ®eld of study with this back-

ground knowledge may produce the situation where the

researcher has already begun to form tentative conceptual

and theoretical links. This, as discussed earlier, is inap-

propriate for grounded theory.

However, no potential researcher is an empty vessel, a

person with no history or background. Further, as it is

common for many researchers to pursue a particular

theme throughout their research activity, they may already

possess some background knowledge of the substantive

area they intend to study. Indeed, the researcher and all

J.R. Cutcliffe

1480 Ó 2000 Blackwell Science Ltd, Journal of Advanced Nursing, 31(6), 1476±1484

Page 6: Methodological issues in grounded theory

his/her knowledge and prior experience is bound up with

the interactive processes of data collection and analysis.

As Strauss & Corbin (1994 p. 278) indicated, the analyst is

also a crucially signi®cant interactant and Glaser (1978)

offered similar remarks when he argued that everything is

data. Thus, the researcher would perhaps be unwise to

carry out reading that provides him/her with anything

more than, as Glaser & Strauss (1967) term, a partial

framework of local concepts, which designate a few

principal or gross features of the situations that he/she

will study. As such knowledge becomes part of the

researcher and consequently becomes part of the inter-

active process. Instead of allowing the theory to emerge

from the data, the researcher albeit implicitly, is likely to

enter into a deductive process. Comparing the data

provided by the interviewees with the knowledge (and

pre-formed conceptual frameworks) already present in

his/her mind. Therefore, as indicated earlier, there may be

value in reading literature that assists in concept clari®-

cation. To draw upon Glaser & Strauss (1967) explanation

of these principal or gross features of the area of study,

they suggest that if a researcher intends studying hospi-

tals, he knows there will be doctors, nurses, and admis-

sion/discharge procedures. Indeed, having this initial

conceptual clarity prior to entering the study, perhaps

helps the researcher to reach conceptual density, enhance

the richness of concept development (Strauss & Corbin

1994) and subsequently the process of theory development.

To return to the example of hope inspiration in bereave-

ment counselling. If the researcher wishes to investigate if

the concepts are linked, and if so, how? it might be

worthwhile to begin by highlighting the current concep-

tualization of both hope and bereavement counselling

within the relevant empirical literature. As there is an

absence of substantive or formal theory that indicates how

such concepts may be related, neither of the reviews of

literature provide the researcher with an implicit theory

which could be tested during the data collection.

However, what it can do is provide an understanding of

the concepts and thus provide ®rm conceptual clarity and

an understanding upon which the rest of the emergent

theory can be built.

This process may not be appropriate for certain types of

research question. For example, factor isolating questions

such as `What is caring?'. What appears to become evident

is that the decision whether or not to conduct a review of

literature to help conceptual clarity may depend upon

what the research question is, but more importantly the

level of theory (Dickoff & James 19682 ) to be induced.

Thus, if the researcher is concerned with inducing factor

isolating theory, it may be disadvantageous to carry out

such a review of the literature. In contrast if the researcher

is concerned with factor relating theory, it could be

advantageous to carry out the literature review in order to

clarify the concepts, before going on to induce a theory that

explicates and explains how they relate to one another.

The other issue regarding literature reviewing in

grounded theory is: When should the second review of

the literature occur? Hutchinson (1993) argued that

because the preliminary literature review can sensitize

concepts (i.e. add to the conceptual clarity) and increase

an awareness of the gaps in the knowledge, this second

review turns to an entirely new body of literature. Stern

(1980), Stern & Allen (1984) and Strauss & Corbin (1994)

argue that selective sampling of the second body of

literature should be woven into the emerging theory

during their third stage on grounded theory induction,

the stage they term concept development.

However, in contrast to these arguments, Glaser (1978)

asserts that the researcher should refrain from accessing

this second body of literature until the theory has emerged

from the data. Therefore it occurs at a later stage. What this

difference of opinion indicates is that there are funda-

mental differences between Glaser's and Strauss' version

of grounded theory (and these are explored in the next

section of this paper). Consequently, the stage at which

the researcher begins to weave in the second body of

literature appears to depend upon which version of

grounded theory is being used.

Precision in grounded theory

There appear to be several issues regarding precision and

clarity in grounded theory and each of these warrants

consideration. These issues can be described as method

slurring with other similar yet different qualitative meth-

odologies, e.g. phenomenology (Baker et al. 1992). The

absence of theoretical coding in some studies which

propound to be grounded theory (Strauss & Corbin 1994,

Melia 1996). Additionally, does the term grounded theory

represent but one single method or, alternatively encom-

pass several methods (May 1996, Melia 1996).

Stern (1994) af®rmed that although there may be simi-

larities in all interpretative methods, the frameworks

underlying the methodologies differ. Baker et al. (1992)

constructed similar arguments and further reasoned that

failure to explicate qualitative methodologies is resulting

in a body of nursing research that is mislabelled. Morse

(1991a p. 15) warned of this mixing of methodologies and

stressed that:

¼ the product is not good science; the product is a sloppy

mishmash.

Thus by paying attention to the resolution or precision

of qualitative research methodology the researcher is

endeavouring to ensure rigour (Baker et al. 1992, Cutcliffe

1997). Such rigorous studies should `stand up' better to

critique by enabling the reader to examine whether or not

Methodological issues in nursing research Methodological issues in grounded theory

Ó 2000 Blackwell Science Ltd, Journal of Advanced Nursing, 31(6), 1476±1484 1481

Page 7: Methodological issues in grounded theory

the chosen methodology was appropriate to the nature of

the research study.

However, Stern (1994) acknowledged that she is

comfortable with the evolution of a methodology, and

that she does not regard researchers `tinkering with' a

given method as problematic. Further, she highlights the

advantages of combining different qualitative methodolo-

gies. Therefore the dif®culty does not lie with blending

one qualitative methodology with another, indeed the

resulting methodology may produce a more thorough,

multidimensional understanding of the phenomenon. In

support of this position, Wilson & Hutchinson (1991)

propose the triangulation of Heideggarian hermeneutics

and grounded theory. For example, these methods could

be combined in order to examine the nature of hope

inspiration in bereavement counselling. The grounded

theory would enable the researcher to ask, `How does this

inspiration occur?', `What if ``X'' were to happen here?',

`What symbolic meanings, gestures, words or behaviours

explain the social reality of the process of hope inspiration

in bereavement counseling?'. The Heideggarian phenom-

enology would enable the researcher to ask the questions,

`What is the lived experience of receiving such hope in a

bereavement counselling setting?', `What meanings do both

the counsellors and the client ascribe to this experience?'.

The crucial issue then, even when using combined

methodologies, is that there is still precision. The method-

ologies are combined purposefully and with intention, and

the researcher subsequently makes explicit what she/he

has done and why. The mixing of methodologies does not

occur by accident. Similarly, the researcher does not call

the product of this mixing grounded theory or phenome-

nology, but tells the researcher the methodology is some-

thing different, e.g. a combination of two methodologies.

The second issue regarding precision in grounded

theory is that of substantive coding and theoretical coding.

Strauss & Corbin (1994 p. 277) were adamant that some

researchers who:

¼ think they are doing grounded theory studies often seem to

concentrate on substantive coding as the methodology's chief and

almost exclusive feature, but do not do theoretical coding.

Few would dispute that substantive coding is an integ-

ral part of data analysis within grounded theory, but if the

intellectual rigour halts at substantive coding, then it is

debatable that the researcher used a grounded theory

methodology. The author of the current paper would argue

not. Glaser (1978) argues that it is the theoretical coding,

the conceptualization of how the substantive codes may

relate to each other as hypotheses, which enables the

substantive codes to be integrated into a theory. It is this

theoretical coding that can provide the full and rich

understanding of the social processes and human interac-

tions which are being studied. The author of this current

paper suggests that theoretical coding perhaps places the

most demand upon the grounded theorist's creativity.

Further, it is perhaps theoretical coding and the postu-

lating of previously undiscovered or unarticulated links

that enables the development of the theory.

If the grounded theorist concentrates on substantive

coding as the chief and almost exclusive feature (Strauss &

Corbin 1994) then it is possible that this limits the

researcher to inducing factor isolating theory, in as much

as substantive coding facilitates the researcher in asking

the questions, `What is this?', `What are the components of

this social process?'. Introducing theoretical coding

enables the researcher to induce factor relating theory.

Theoretical coding facilitates the researcher in asking the

questions, `What is happening here?', `How do the

substantive codes relate to each other as hypotheses?'.

To ignore the central feature of the methodology (Strauss &

Corbin 1994) and then subsequently call the methodology

grounded theory is a clear indication of imprecision.

Another issue of precision is whether or not the term

grounded theory encompasses more than one methodo-

logy. If so, is it appropriate to call a methodology

grounded theory or would another term have to be used?

As indicated in the introduction grounded theory was

discovered by Glaser & Strauss (1967), both of whom had

distinctly different academic backgrounds. It maybe no

surprise then that the subsequent development of the

methodology since then has taken a different path for each

author.

According to Stern (1994), Glaser, who has a back-

ground in statistical analysis, insists on allowing the

theory to emerge, whereas Strauss, whose sociology was

®rmly rooted in the Chicago school (i.e. a school of

sociology which has its roots in the symbolic interac-

tionist tradition, Robrecht 1995), prefers a method that is

tightly prescriptive. Stern (1994) and Melia (1996) also

suggest that Glaserian grounded theory would be expected

to be immediately applicable to individuals and groups

who shared the problem under study and would be

expected to be testable. However, theory produced using

Strauss' version of grounded theory has its applicability

downplayed (Stern 1994). The crux of the dichotomy is,

according to Glaser (1992) the fundamental difference

between emerging and forcing. Stern (1994) argued a key

difference is the questions each author asks of the data. As

Strauss examines the data, he stops at each word and asks,

`What if ?', whereas Glaser keeps his attention on the data

and asks, `What do we have here?'.

According to Stern (1994 p. 20):

Strauss brings to bear every possible contingency that could relate

to the data, whether it appears in the data or not. Glaser focuses

his attention on the data to allow the data to tell their own story.

Glaser (1978) went as far as to claim that Strauss'

evolution is a departure from the original methodology

and represents an erosion of grounded theory (Melia

J.R. Cutcliffe

1482 Ó 2000 Blackwell Science Ltd, Journal of Advanced Nursing, 31(6), 1476±1484

Page 8: Methodological issues in grounded theory

1996). Indeed he asserts that the two methodologies

should have different names, with Strauss' version being

termed, full conceptual description.

If the crux is in the different questions asked of the data,

what happens if the researcher asks both questions?

Further, if the researcher asks another question, is he/

she using another different version of grounded theory?

For example Turner (1981 p. 232) asked:

¼ what categories, concepts or labels do we need in order to

describe or account for the phenomena discussed in this para-

graph?

Stern (1980 p. 281) proposed that the researcher exam-

ines the data, line by line, and produce substantive codes:

These codes are called substantive codes, because they codify the

substance of the data, and often use the very words used by the

actors themselves.

If the nature of the grounded theory is determined in

part by the questions asked of the data, then does the

researcher have to predetermine the questions she/he will

ask and then ensure that this is the only question they ask

of the data? The author of this current paper suggests that

this might be somewhat restrictive and disabling. If the

researcher has to pause and ask himself, `What questions

am I asking of the data and consequently does this differ

from the methodology I chose originally?', then the

researcher may be becoming more concerned with process

questions rather than creative, interpretative questions.

Consequently, substantive coding and theoretical coding

could be impeded.

Perhaps a combination of questions of the data, a

combination of `What if?, `What do we have here?' and

`What categories, concepts or labels do we need to account

for the phenomena?', would provide a richer and more

complete understanding. Providing the researcher makes

explicit what questions they have asked, then some issues

of precision have been addressed. Consequently, the

researcher can predominantly use either Glaser's or

Strauss' version of grounded theory and then augment

their method by including additional questions. The

author of the current paper posits that providing the

researcher explains what she/he has done and how she/he

did it, straying outside of the boundaries of one particular

version is less of an issue than limiting the potential depth

of understanding that strict adherence to one version

would produce.

CONCLUSION

Researchers intent on utilizing a grounded theory meth-

odology should pay attention to methodological issues,

including those addressed in this paper; sampling,

creativity and re¯exivity, the use of literature, and preci-

sion. The author concludes by making the following

recommendations.

When utilizing a grounded method researchers should

consider their research question and clarify what level of

theory is likely to be induced and then decide when they

intend to access and introduce the second body of litera-

ture. They should acknowledge that in the early stages of

data collection, some purposeful sampling appears to

occur. They may wish to free themselves from the restraints

that limit their use of creativity and tacit knowledge. The

researchers might be served by attention to issues of

precision, including avoiding method slurring, ensuring

theoretical coding occurs, and predominantly using one

method of grounded theory while explaining and

describing any deviation away from this chosen method.

Such mindfulness and the resulting methodological

rigour is likely to increase the overall quality of the

inquiry and enhance the credibility of the ®ndings.

References

Altheide D.L. & Johnson J.M. (1994) Criteria for assessing inter-

pretative validity in qualitative research. In Handbook of

Qualitative Research (Denzin N. & Lincoln Y.S. eds), Sage

London, pp. 485±499.

Baker C., Wuest J. & Stern P.N. (1992) Method slurring: the

grounded theory/phenomenology example. Journal of

Advanced Nursing 17, 1355±1360.

Becker P.H. (1993) Common pitfalls in published grounded

theory research. Qualitative Health Research 3, 254±260.

Berger P. & Kellner H. (1981) Sociology Reinterpreted. Anchor

Books, New York.

Coyne I.T. (1997) Sampling in qualitative research: purposeful

and theoretical sampling; merging or clear boundaries? Journal

of Advanced Nursing 26, 623±630.

Cutcliffe J.R. (1997) Qualitative nursing research: a quest for

quality. British Journal of Nursing 6, 969.

Cutcliffe J.R. & McKenna H.P. (1999) Establishing the credibility

of qualitative research ®ndings: the plot thickens. Journal of

Advanced Nursing 31, 374±3803 .

Dickoff J. & James P. (1968) A theory of theories: a position paper.

Nursing Research 17, 3.

Glaser B. & Strauss A.L. (1967) The Discovery of Grounded

Theory: Strategies for Qualitative Research. Aldine, Chicago.

Glaser B.G. (1978) Theoretical Sensitivity. Advances in the

Methodology of Grounded Theory. The Sociology Press, Mill

Valley, California.

Glaser B.G. (1992) Basics of Grounded Theory Analysis: Emer-

gence vs. Forcing. Sociology Press, Mill Valley, California.

Hammersley M. (1992) What's Wrong with Ethnography? Routl-

edge, London.

Hickey G. (1997) The use of literature in grounded theory.

Nursing Times Research 2, 371±378.

Hutchinson S.A. (1993) Grounded theory: the method. In Nursing

Research: A Qualitative Perspective 2nd edn (Munhall P.L. &

Boyd C.A. eds), National League for Nursing Press, New York,

pp. 180±212.

Methodological issues in nursing research Methodological issues in grounded theory

Ó 2000 Blackwell Science Ltd, Journal of Advanced Nursing, 31(6), 1476±1484 1483

Page 9: Methodological issues in grounded theory

Lincoln Y.S. & Guba E.G. (1985) Naturalistic Enquiry. Sage,

London.

Lipson J. (1991) The use of self in ethnographic research. In

Qualitative Nursing Research: A Contemporary Dialogue

(Morse J.M. ed.), Sage, London, pp. 73±89.

May K.A. (1996) Diffusion, diffusion or distillation? The case of

grounded theory method. Qualitative Health Research 6,

309±311.

Melia K.M. (1996) Rediscovering Glaser. Qualitative Health

Research 6, 368±378.

Morse J.M. (1991a) Qualitative nursing research: a free for all. In

Qualitative Nursing Research: A Contemporary Dialogue

(Morse J.M. ed.), Sage, London, pp. 14±22.

Morse J.M. (1991b) Strategies for Sampling. In Qualitative

Nursing Research: A Contemporary Dialogue (Morse J.M. ed.),

Sage, London, pp. 126±145.

Morse J.M. (1994) Emerging from the data: the cognitive processes

of analysis in qualitative enquiry. In Critical Issues in Quali-

tative Research Methods (Morse J.M. ed.), Sage, London,

pp. 23±43.

Morse J.M. (1998) What's wrong with random selection? Quali-

tative Health Research 8, 733±735.

Morse J.M. & Field P.A. (1995) Qualitative Research Methods for

Health Professionals 2nd edn. Sage, London.

Patton M.Q. (1990) Qualitative Evaluation and Research Methods

2nd edn. Sage, London.

Robrecht L.C. (1995) Grounded theory: evolving methods. Qual-

itative Health Research 5, 169±177.

Sandelowski M., Holditch-Davis D. & Harris B.G. (1992) Using

qualitative and quantitative methods: the transition to parent-

hood of infertile couples. In Qualitative Methods in Family

Research (Gilgum J.F., Daly K., Handel G. eds), Sage, London,

pp. 301±323.

Smith K. & Biley F. (1997) Understanding grounded theory:

principles and evaluation. Nurse Researcher 4, 17±30.

Stern P. (1980) Grounded theory methodology its uses and

applications. Image 12, 20±23.

Stern P.N. (1994) Eroding grounded theory. In Critical Issues in

Qualitative Research Methods (Morse J.M. ed.), Sage, London,

pp. 210±223.

Stern P. & Allen L. (1984) Qualitative research Ð the nurse as a

grounded theorist. Health Care for Women International 5,

371±385.

Stern P.N., Allen L. & Moxley P. (1982) The nurse as a grounded

theorist: history, process and uses. Review Journal of Philos-

ophy and Social Science 7, 200±215.

Strauss A. & Corbin J. (1994) Grounded theory methodology: an

overview. In Handbook of Qualitative Research (Denzin N.K. &

Lincoln Y.S. eds), Sage, London, pp. 273±285.

Turner B. (1981) Some practical aspects of qualitative data

analysis: one way of organising the cognitive processes associ-

ated with the generation of grounded theory. Quality and

Control 15, 225±245.

Wilson H.K. & Hutchinson S.A. (1991) Triangulation of qualita-

tive methods: Heideggerian hermeneutics and grounded theory

Qualitative Health Research 1, 263±276.

J.R. Cutcliffe

1484 Ó 2000 Blackwell Science Ltd, Journal of Advanced Nursing, 31(6), 1476±1484