Q-Methodology as a Phenomenological Research Method
Pnina Shinebourne and Martin Adams
Abstract
This paper aims to introduce Q-methodology in phenomenological research. We
discuss the process of using the methodology and demonstrate Q-
methodology's distinctive features and its particular approach to
researching the range and diversity of subjective understandings, beliefs
and experiences. Q methodology is particularly suitable for identifying
commonality and diversity and has a powerful capacity for thematic
identification and analysis. In our view, Q-methodology can make a
contribution to expanding the repertoire of phenomenological research
methods available to psychotherapists. In the final section of the paper we
provide an example of a research project using Q-methodology. The project
aims to increase our understanding of therapists' subjective attitudes,
beliefs and experiences in working with clients with problems of addiction.
Using Q-methodology it was possible to construct a unique configuration of
each participant's engagement with the topic, and at the same time to
identify four distinctive factors, indicating patterns of divergent
understandings of addiction. Each factor is shown to be a multifaceted
construct which does not conform to a priori conceptualizations.
Keywords
Q-methodology, phenomenological research methods, qualitative research
methods, therapists' understandings of addiction.
Introduction
Q methodology was invented by the British physicist-psychologist William
Stephenson in the 1930s (for a historical review see Brown: 1980).
Stephenson was interested in providing means for exploring subjectivity in
a variety of situations (such as appraisals of health care, experience of
bereavement or political attitudes) from the standpoint of the lived
experience of the person (Brown: 1996). He encountered difficulties as his
approach ran counter to the positivist viewpoint which dominated psychology
at the time. Q methodology was later developed in the United States by
Brown, and with the "turn to subjective experience in social science
research" (Watts & Stenner: 2005a) and the development of social
constructionist studies, found a more sympathetic environment for
applications.
In this paper we present the case for Q-methodology as a phenomenological
research method. Q methodology (Brown, 1980, Stainton Rogers: 1995, Watts &
Stenner, 2005) shares with other phenomenological methods the principles of
seeking meaning through exploring subjective accounts of phenomena from
participants' perspectives, attempting to identify broad categories and
common themes and a commitment to a collaborative engagement with
participants[i]. There are, however, some significant variations which we
discuss below. Q methodology is considered particularly suitable for
researching the range and diversity of subjective understandings, beliefs
and experiences "addressing the critical kind of research questions which
are concerned to hear 'many voices'" (Stainton Rogers1995:183). At the same
time, it facilitates the identification of similarities, the construction
of broad categories or dimensions of the phenomenon being investigated and
the exploration of patterns and relationships within and between these
dimensions.
Q methodology has been used in various fields[ii], for example, studies
of health and illness (e.g. Stenner et al.: 2000, Stainton Rogers: 1991),
feminist studies (e.g. Kitzinger: 1987, Snelling: 1999), self-harming
behaviour (Rayner & Warner: 2003), exploration of emotions such as jealousy
(Stenner & Stainton Rogers: 1998) and love (Watts & Stenner: 2005a). Yalom
and his colleagues used Q-method in their studies of the therapeutic
factors in group therapy (a detailed account appears in Yalom & Leszcz:
2005), McLeod & Ryan (1993) used Yalom's set of statements in their study
of therapeutic factors in an out-patient therapy group for older women.
Overview of Q-methodology
Participants in a study using Q-methodology are asked to sort a set of
statements representing a broad diversity of opinions and perspectives on
the phenomenon being investigated. Items for the Q set can be gathered from
a variety of sources. For example, direct quotes and themes from interviews
with participants (Kitzinger: 1987); statements originating from academic
literature and popular media in addition to interviews (Stainton Rogers:
1991); statements based on the literature (Snelling: 1999) or a complete
set of scale items (from previous research) to create a ready-made Q set
(McLeod & Ryan: 1993). In preparation for the sorting task, each item is
numbered and written on a separate card. A set of between 40 and 80
statements is considered satisfactory (Stainton Rogers: 1995, Watts &
Stenner: 2005).
Pilot studies are recommended (Watts & Stenner: 2005, Stainton Rogers:
1991) to ensure that the set of items reflects the broad range of ideas,
opinions and arguments about the topic. Piloting participants are asked to
identify additional statements to ensure that the Q set provides adequate
coverage and to provide general comments on the existing statements.
However, As Watts & Stenner (2005) point out, the Q set can never be
"complete", as there is always something else which may potentially be said
about any topic. The selection of participants is guided by the aims of the
study. Although a single Q case study is possible (Watts & Stenner: 2005),
aiming to demonstrate a wider range of viewpoints will require more
participants. Pilot studies require small numbers, perhaps selected
strategically (bearing the research question in mind) to include
participants who can provide helpful comments and additional statements
from a variety of perspectives.
Participants sort the cards according to the instructions given by the
researcher. For example, an instruction could be to sort the cards
initially into three piles, 'agrees', 'disagrees' or 'neither agrees, nor
disagrees (neutral)' with the statement. They then continue to sort the
cards within each broad pile, according to the number of possible positions
in the sorting template (Figure 1). Beginning with the 'agree' pile,
participants select the two items they agree with most, then the three
items with slightly less degree of agreement, and proceed until all the
items in the 'agree' pile have been allocated. The process is repeated with
the 'disagree' pile and continues with the 'neutral' pile until all cards
have been sorted. Participants than write all the statement numbers in the
appropriate box in the sorting template. A post sorting interview is
important, asking participants to comment on specific items, to suggest
additional items that might be included, to point out items that are not
clear, etc. Such open-ended comments aid the interpretations of the sorting
configuration. Further follow-up interviews with participants can take
place at a later stage (see below).
Figure 1 – sorting template
The next task is to identify patterns of similarities and divergence
using a statistical package such as SPSS or a dedicated Q package[iii]. The
program will correlate each Q-sort (i.e. the overall configurations
produced by each participant) with each other Q-sort, to identify a small
number of factors or themes which can represent the shared forms of
understandings among participants. Following a number of statistical
procedures[iv] a number of factors will emerge. The Q-sorts of all
participants who loaded significantly on a factor are merged to produce a
single configuration which serves as a factor array, or factor exemplar,
used as the basis for preparing narrative accounts of each factor and in
factor interpretation.
A table of the complete ranking assigned to each statement in each factor
array is prepared as a starting point for factor interpretation. (see
Appendix One). Factor interpretation is based on the examination of the
ranking assigned to each statement in the factor array together with
participants' comments from the post-sorting interview. Factor
interpretation is seen as a hermeneutic process (Stenner et al.:2000),
engaging the interpretative perspective of the researcher. Yet it is
constrained by the subjective input of the participants as expressed in the
sorting templates and in follow-up interviews.
Q-methodology as a phenomenological research method
Because of its use of numerical data and statistical analysis Q-
methodology has sometimes been assumed to be a quantitative method (and
criticized for not meeting quantitative research standards). However,
unlike quantitative methods "Q-sorts really measure nothing if they are
correctly applied, and their meaning-conferring and intentionality only
emerges by way of factor analysis of several Q-sorts" [emphasis in
original] (Stephenson: 1988: 211). Another potential area of confusion is
that the statistical technique of Q-sort has been used in various
applications to compare participants' sorts with an a priori constructed
ideal Q-sort[v], rather than using factor analysis to let the factors
emerge from the individual sorts. Stephenson contends that "Husserl's
phenomenology and Q-methodology are on the same footing, except for
denial of consciousness as substantive and its replacement by
communicability in Q"[vi] (ibid: 216). Communicability in Q refers to the
flows of ideas, statements, commentaries, beliefs, etc, surrounding any
topic (Brown: 1980). The set of statements is envisaged as a "launch pad
for an investigation, an entré into a phenomenon" (ibid: 39), the
researcher's "best initial guess" (ibid) to start the process and engage
in a collaborative manner in a dialogue with participants. Brown is in
fact describing the Q set of statements as an entry point into a
"hermeneutic circle".
As mentioned above, Q-methodology shares with other phenomenological
methods the principles of seeking meaning through exploring subjective
accounts of phenomena from participants' perspectives, attempting to
identify broad categories and common themes and a commitment to a
collaborative engagement with participants. There are however, some
significant differences. It can be argued (see Cordingley et al.: 1997)
that fixing the number of items to be assigned to each ranking position may
seem counter to the objective of Q-methodology, that is, that participants
are able to indicate their own viewpoint without restraints. Countering
this point, Brown (1980) demonstrated that the effect of having either free
or forced distributions is minimal on the pattern of factors which
emerge[vii].
Another difference is that In Q-methodology the participants control the
classification process and the factors are derived statistically from the
results of the sorting activities of the participants, rather than arrived
at through the researcher's process of analysis and classification of
themes as in other phenomenological research which privileges the
researcher's conceptualizations. For this reason "the analysis involved in
Q-methodology may incorporate less 'researcher bias' than other
interpretive techniques" (Cordingley et al.: 1997). However, the activity
of selecting the statements for the Q sample privileges the researcher at
an earlier stage of the process. Participants are constrained to engaging
with the selected statements, in contrast to other phenomenological
approaches in which participants' accounts in their own words are at the
heart of the enquiry. Clearly another phenomenological method would be more
appropriate for an in-depth idiographic research (Smith et al.: 1995) in
which a detailed account of the lived experience of a particular person is
paramount. However, to a certain extent participants' own words can be
incorporated in a Q set by deriving statements directly from interviews
(Kitzinger: 1987, Rayner & Warner: 2003), or by constructing a set of
statements using direct quotes and themes from interviews and individual
accounts, mixed with materials originating from other sources, such as
academic literature and popular media (Stainton Rogers: 1991) or online
discussion groups.
In our view, by deriving the statements from multiple discourses, the
process of constructing a subjective meaning enabled by Q-methodology makes
explicit how individual accounts are intertwined with, and inevitably
mediated through our engagement with the world which is situated, informed
and constrained by social, cultural, professional (and other) discourses
available to us. This multidimensional aspect may not be readily apparent
in descriptive accounts in traditional phenomenological research, thus
placing particular responsibility on the researcher to position the initial
description in relation to a wider social, cultural or theoretical context
and provide a critical and conceptual commentary (Larkin et al.: 2006), a
hermeneutic undertaking shared by Q-methodology.
In the spectrum of phenomenological methods ranging in emphasis from
uniqueness to universality, Q-methodology could be seen as encompassing
both dimensions, but with particular strengths in identifying
commonalities. In Q research the unique configuration of each participant's
engagement with the topic is retained in the individual's sort matrix that
can be used to explore further aspects of the individual position in
relation to other relevant data (e.g. biographical or professional). At the
same time, similarities between these subjective accounts are identified
through factor analysis to produce a limited number of factors sharing
common understandings. Considering that a set of, for example, 60 cards is
sorted by perhaps 20 or more participants in numerous different ways, in
probabilistic terms the likelihood that distinct patterns will occur is
indeed very small. Therefore, any factors that do emerge suggest that
participants are responding to the sorting task in significant degrees of
shared understandings, albeit with different loadings (indicating different
degree of agreement with the exemplar account of each factor).
In our view, Q-methodology can make a contribution to expanding the
repertoire of phenomenological research methods available to
psychotherapists. Q-methodology is particularly suitable for identifying
commonality and diversity and has a powerful capacity for thematic
identification and analysis. Another phenomenological method (e.g.
Interpretative Phenomenological Analysis (IPA), Smith et al.:1999) would be
suitable for a different type of phenomenological enquiry, e.g. an in-depth
exploration of the lived experience of a specific individual, or a detailed
narrative account unfolding over a period of time, as a Q study can provide
a snapshot at a given time only. The strengths of both approaches could be
combined in a more comprehensive phenomenological study, using Q-
methodology in combination with IPA to explore more fully both the
uniqueness and the commonalities of human experience. For example, a Q
study with a (relatively) small number of participants could be used as a
starting point to identify a small number of groups through factor
analysis. The individual whose Q-sort best exemplifies each factor (highest
significant loading) could then be a subject for an in-depth
phenomenological case study.
Using Q-methodology in phenomenological research
In the final section of this paper we briefly present a possible use of Q-
methodology in phenomenological research. The example is taken from a
pilot study exploring therapists' subjective beliefs, values and
experiences in working with clients with problems of addiction. The set of
statements for the study was drawn from diverse sources, ranging from
personal experience and reflection on working with clients with problems of
addiction, collecting statements from colleagues working with clients with
problems of addiction, as well as from a detailed survey of relevant
literature (see examples of statements in Appendix One). The Q-set
contained sixty statements and was sorted by thirteen participants,
selected to reflect a range of theoretical orientations, variety of
agencies, years of work experience and supervisory experience. Following
the sorting task participants were asked to provide and suggest additional
statements that might be included. The resulting Q-sorts provide a unique
configuration of each participant's engagement with the topic.
The next task was to identify a small number of factors or themes which
can represent the shared forms of understandings among participants, using
a dedicated Q package (PCQ, Stricklin & Almeida, 2001). Four distinctive
factors were identified as a result of the statistical analysis, as
explained above. Factor interpretation was based on the examination of the
ranking assigned to each statement in the factor array (see Appendix A)
combined with participants' comments from the post-sorting interview. A
detailed narrative summary of each factor was prepared (see Appendix Two
for an example of a narrative summary[viii]) and each factor was given a
descriptive title (Factor A – acceptance, Factor B – challenge, Factor C –
ambivalence, Factor D – rejection).
Each factor is shown to be a distinctive multifaceted construct which
does not conform to a priori conceptualizations. The shared understandings
expressed by participants who loaded significantly on a specific factor do
not correspond to any one theory or therapeutic orientation, but rather
represent a perspective which is a composite of divergent understandings,
beliefs and experiences. For example, factor A broadly corresponds to a
humanistic, non-directive approach, while factor B corresponds to a more
directive, challenging approach with cognitive behavioural elements.
However, at the same time, issues of denial, attachment and the importance
of gaining insight, traditionally associated with a psychoanalytical
perspective, are very significant in factor B account. These multifaceted
perspectives were confirmed in the post-sorting comments with most
participants indicating that more than one theoretical approach impacts on
their understanding and their practice despite different conceptualizations
of addiction and different views of the therapeutic process. At the same
time it was noted that incorporating practices and views from various
approaches may also result in difficulties for therapists when their
understanding and approach to therapy conflicts with the philosophy or
policy of the agency where they work (as presented in factor A account).
The factors also display divergent perspectives of therapists' feelings
and attitudes towards clients, ranging from acceptance of "difficult"
behaviour (factor A) and feelings of compassion towards clients struggling
to overcome problems with drugs/alcohol (factor A), to feelings of being
manipulated, frustrated and anxious (factors B & C). Negative feelings
about clients are also expressed in relief when some "difficult" clients
drop out of therapy and doubting competence as a therapist (factor C) or in
a lack of compassion towards clients (factor B). Indeed, this range of
attitudes to clients with problems of addiction correspond to accounts in
the literature suggesting that they are perceived as more "difficult" than
other clients. They present challenges to therapists' competence (Najavits
et al., 2000; Najavits et al., 1995; Wheeler & Turner, 1997) and they may
evoke difficult countertransference (Vannicelli, 2001; Rodríguez de la
Sierra, 2002). At the same time, a degree of positive feelings of empathy
and caring is also indicated (Najavits et al., 1995).
Several participants commented on the multiple possibilities for
understanding and interpretation of some statements, with terms sometimes
standing for ideological positions. For example, one participant commented
on the term "relapse". Although she accepts it is used in a specific sense,
for her it entails a negative judgemental attitude towards client's
behaviour. Similarly, the term "gratified" evoked a number of possible
interpretations, feeling pleased for a client, taking credit for the
client's work or taking on too much responsibility for the process of
therapy. It appears that statements which were perceived as open to several
meanings could be considered particularly productive for further
reflection. Several participants saw the possibilities in an extended
discussion following the sorting and suggested that the study might be
developed as a tool for staff training in their agencies.
Using Q-methodology it was possible to construct a unique configuration
of each participant's engagement with the topic (as expressed in the
individual's sort matrix), and at the same time to identify commonalities
and diversity of viewpoints which do not conform to a priori
conceptualizations.
Conclusion
This paper has aimed to introduce Q-methodology in phenomenological
research. It has attempted to demonstrate Q-methodology's distinctive
features and its particular approach to researching the range and
diversity of subjective understandings, beliefs and experiences. In our
view, Q-methodology can make a contribution to expanding the repertoire of
phenomenological research methods available to psychotherapists and we
hope our paper will encourage other researchers to explore Q-methodology.
Appendix One - Extracts from Table of ranking of statements by factor array
Reading the table by column reveals the comparative ranking of items within
each factor array.
Reading the table by row reveals the comparative ranking of a particular
item across all factors.
Factor A – Acceptance, Factor B – Challenge, Factor C – Ambivalence, Factor
D – Rejection.
Statements:( Factors :( A B C D
01. Addictive behaviour is developed, in part, by repeating +2
0 0 -3
patterns learned in the social environment (e.g. family/peers)
02. There is a genetic predisposition to addiction -1 -6
-2 -3
03. Religious or spiritual engagement supports recovery from +2 -1
0 +1
addiction
05. Attachment to the addiction is a substitute to other -2 +6
+5 +3
relational attachments
06. Denial is a typical feature of addictive behaviour -1 +6
0 +2
07. Medication is helpful in supporting recovery from +1 0
+1 +2
alcohol/drug addiction
08. Thinking about addiction as a disease helps to remove -1
-1 -4 +4
stigma or blame
17. A harm-reduction approach is suitable for working +4 -1
-1 0
with alcohol / drug users
18. I recognize addictive elements in my own behaviour +4 0
+6 0
20. Group therapy is helpful for people with alcohol/drug +2
0 +3 +1
problems
21. I find that working with clients with problems of addiction -3
-5 +3 -2
leads me to doubt my competence as a therapist more
often than working with other clients
22. I encourage clients to explore their ambivalence -5
+5 +6 -6
about change
23. People with problems of addiction are often unable to -1
-2 +3 +3
control their behaviour
25. People use alcohol/drugs as "self medication" +1 +4
0 -2
32. I feel manipulated by some clients -3 +3 -2
+4
42. I feel relieved when some "difficult" clients drop -4 -3
+4 -4
out of therapy
Appendix Two – extracts from Factor A – Acceptance
[Note: the numbers in brackets represent the statement number followed by
the rank given to the statement as shown in Appendix One. For example in
the account of factor A, the numbers in brackets (17:+4) indicate that
statement number 17 is ranked at +4].
Factor A has seven significantly loading participants. Participants who
share this perspective accept that a client may not aim for abstinence
(statement no. 58 was ranked +6, as shown in Appendix Three, hence =
58:+6). In this view, a harm-reduction approach is seen as suitable for
working with alcohol/drug users (17:+4). Controlled drinking is not
considered a myth (14:-4) and abstinence is not seen as crucial to recovery
(11:-3). Participants experience difficulties when their understanding and
approach to therapy conflicts with the philosophy or policy of the agency
where they work with regards to abstinence or harm reduction (26:+5), as
illustrated by participant 13 comment 'I walked out of an agency which
insisted on abstinence in week 5'. In factor A account addicted people are
not seen as powerless (50:-6), and there is some disagreement with the view
that people are unable to control their behaviour (23:-1) or are
unaccountable (29:-2) for behaviour under the influence of the addictive
substances.
In this account, addiction often occurs with other mental health
problems (30:+4), but is not seen as an incurable disease (34:-2). Thinking
about addiction as a disease may encourage addicted people to believe that
someone, a doctor or a health care professional, will deal with the problem
for them (47:+3). However, there is some agreement that medication is
helpful in supporting recovery from alcohol/drug addiction (07:+1) and that
people use alcohol/drugs as "self-medication" (25:+1). In this view,
addiction has meanings and functions in the life of the person (54:+5) and
participant 11 comments that trying to understand these meanings is a good
way to enter the client's world. It is suggested that alcohol/drug use can
help coping with negative emotional states, such as anger, frustration, low
self-esteem, anxiety or loneliness (56:+3) and with stressful life events,
such as unemployment or divorce (37:+3). To a lesser degree, it is
suggested that alcohol/drug use brings about positive emotional states,
such as pleasure, freedom, sexual enhancement or relaxation (45:+1). There
is some agreement that addiction can be a manifestation of underlying
psychological (19:+2) or social (24:+2) problems and that addictive
behaviour can be developed by repeating patterns learned in the family or
peer environment (01:+2). There is some suggestion that addiction can be a
manifestation of traumatic events or abuse in childhood (49:+1). However,
genetic predisposition is not seen as a factor in developing addiction (02:-
1)
There is a strong feeling of compassion for a client struggling to
overcome problems with drugs/alcohol (46:+6) and a belief that many people
with problems of addiction recover eventually (38:+5). Relapse can be
understood as a part of the process of recovery which could take place over
a long period of time (59:+3). There seems to be no negative feelings
regarding clients' behaviour: no frustration with clients' attempting to
conceal drug/alcohol use (40:-4), not feeling manipulated by clients (32:-
3), not feeling let down by a client's frequent non-attendance (16:-4) or
relieved when a "difficult" client drops out of therapy (42:-4). There is
some support for the view that personal experience of addiction helps the
therapist to understand and work more effectively with clients with
problems of addiction (13:+1). Working with addicted clients does not lead
to doubting competence as a therapist more often than working with other
clients (21:-3). There is some support for specific training in the field
of addiction to help therapists to understand and work more effectively
with this client group (41:+2).
Pnina Shinebourne completed her UKCP training at the New School of
Psychotherapy and Counselling. She is currently a PhD research student at
Birkbeck, University of London.
Correspondence:
[email protected]
Martin Adams is a psychotherapist and supervisor in private practice and a
visiting lecturer at Regents College and the New School of Psychotherapy
and Counselling, both in London.
Correspondence:
[email protected]
Notes
-----------------------
[i] There are, however, significant variations between different approaches
to phenomenology. See for example, Giorgi (1997), Moustakas (1994), van
Manen (1990), Smith et al (1999), Benner (1994).
[ii] PsychINFO lists 206 references (searched 1/3/3006) to Q methodology in
various areas of its coverage of psychology and related disciplines.
[iii] A freeware package PQMethod-2.11d, is available from: http://www.lrz-
muenchen.de/~schmolck/qmethod/ Alternatively, PCQ (Stricklin & Almeida,
2001) is a commercially available program.
[iv] The factors are rotated to obtain a pattern of loadings on each factor
that is as diverse as possible. The Varimax procedure is suggested by
Watts & Stenner (2005). For a factor to be interpretable one requirement
is eigenvalue greater than 1.0 (eigenvalue is sum of squared loadings for
a factor). It conceptually represents the amount of variance accounted for
by a factor. A second requirement is that a factor must have at least two
sorts that load significantly upon it alone. A third requirement is to set
a significance level to determine the minimum threshold required to become
a 'member' of a factor. The significance level can be calculated by the
program in relation to the overall number of statements in the set.
Setting the level lower means easier membership, setting the level higher
means more restricted membership.
[v] For example, Rogers and Dymond (1954) used the Q-sort technique
(without using factor analysis), to measure mental health by comparing the
Q-sorts of mental patients against a Q template of the 'ideally healthy
person'.
[vi] Husserl's understanding of consciousness is contested here. A
discussion of the approach to phenomenology implicated in Stephenson's
view is beyond the scope of this paper.
[vii] In Brown's example, for a set of 33 items with a -4 to +4
distribution, there are above 11,000 times more ways to sort the
statements than "there are people in the world".
[viii] A full account of each factor is inevitably lengthy and beyond the
scope of this paper (full details can be provided on request).
References
Benner, P. (Ed). (1994). Interpretive phenomenology. Thousand Oaks: Sage.
Brown, S. (1980). Political subjectivity: applications of Q methodology in
political science. New Haven: Yale University Press.
Brown, S. (1996). Q methodology and qualitative research, Qualitative
Health Research, 6:4:561-567.
Cordingley, L., Webb, C. & Hillier, V. (1997). Q methodology, Nurse
Researcher, 3:3:3 –45.
Giorgi, A. (1997). The theory, practice, and evaluation of the
phenomenological method as a qualitative research procedure, Journal of
Phenomenological Psychology, 28;2:
Kitzinger, C. (1987). The social construction of lesbianism. London: Sage.
McLeod, J. & Ryan A. (1993). Therapeutic factors experienced by members of
an out-patient therapy group for older women. British Journal of Guidance
& Counselling, 21, 1, 64-72.
Van Manen, M. (1990). Researching lived experience. Albany: State
University of New York Press.
Moustakas, C. (1994). Phenomenological research methods. Thousand Oaks:
Sage.
Najavits, L. et al. (1995). Therapists' emotional reactions to substance
abusers: a new questionnaire and initial findings, Psychotherapy: Theory,
Research, Practice, Training, 32, 4, 669-677.
Najavits, L., Crits-Christoph, P. & Dierberger, A. (2000). Clinicians
impact on the quality of substance use disorder treatment. Substance Use &
Misuse, 35, 2161-2190.
Rayner, G. & Warner, S. (2003). Self-harming behaviour: from lay
perceptions to clinical practice. Counselling Psychology Quarterly, 16,
4, 305-329.
Rodríguez de la Sierra, L. (2002). Countertransference: our difficulties in
the treatment of substance abuse. In: Weegmann, M. & Cohen, R. (Eds). The
Psychodynamics of addiction. London: Whurr.
Rogers, C., Dymond, R. (Eds). (1954). Psychotherapy and personality change.
Chicago:University of Chicago Press.
Smith, J. A., Harré, R. & Van Langenhove, L. (1995). Idiography and the
case study. In Smith, J.A, Harré, R. & Van Langenhove, L. (eds.).
Rethinking methods in psychology. London: Sage.
Smith, J. A., Jarman, M. & Osborn, M. (1999). Doing interpretative
phenomenological analysis. In: Murray, M. & Chamberlain, K. (Eds.),
Qualitative health psychology: Theories and methods (pp.219-240). London:
Sage.
Snelling, S. (1999). Women's perspectives on feminism: a Q-methodological
study,Psychology of Women Quarterly, 23, 2, 247-266.
Stainton Rogers, R. (1995). Q methodology. In Smith, J., Harré, R. & Van
Langenhove, L.(Eds.). Rethinking methods in psychology. London: Sage.
Stainton Rogers, W. (1991). Explaining health and illness: an exploration
of diversity. London: Harvester Wheatsheaf.
Stenner, P., Dancey, C. & Watts, S. (2000). The understanding of their
illness amongst people with irritable bowel syndrome: a Q methodological
study. Social Science & Medicine, 51, 439-452.
Stenner, P. & Stainton Rogers, R. (1998). Jealousy as a manifold of
divergent understandings: a Q methodological investigation. British
Journal of Social Psychology, 28, 71-94.
Stephenson, W. (1988). William James, Niels Bohr, and complementarity: V –
phenomenology of subjectitivy. The Psychological Record, 38:203-219.
Stricklin, M. & Almeida, J. (2001). PCQ: analysis software for Q-technique
(Academic edition revised) [computer software].
Vannicelli, M. (2001). Leader dilemmas and countertransference
considerations in group psychotherapy with substance abusers.
International Journal of Group Psychotherapy, 51, 43-62.
Watts, S. & Stenner, P. (2005). Doing Q methodology: theory, method and
interpretation, Qualitative Research in Psychology. 2, 67-91.
Watts, S. & Stenner, P. (2005a). The subjective experience of partnership
love: a Q-methodology study. British Journal of Social Psychology. 44,
85-107
Wheeler, S. & Turner, L. (1997). Counselling problem drinkers: the realm of
specialists, Alcoholics Anonymous or generic counsellors? British Journal
of Guidance & Counselling, 25, 3, 313-326.
Yalom, I. & Leszcz, M. (2005). The theory and practice of group
psychotherapy, 5th Edition. New York: Basic Books
-----------------------
-6 -5 -4 -3 -2 -1 0
+1 +2 +3 +4 +5 +6
