Professional Documents
Culture Documents
Practice of ResearcherClinicians?
Jeremy D. Safran, Iris Abreu, Joseph Ogilvie, and Anthony DeMaria, New School for Social
Research
This survey evaluated the impact of empirical research most widely cited surveys examining the degree to
findings on the clinical practice of a sample of psycho- which clinicians utilize the findings of psychotherapy
therapists who are also actively engaged in the con- research in their practices. They surveyed members of
sumption, critical evaluation, and production of the American Psychological Associations (APA) Divi-
psychotherapy research. On the one hand, we found
sion of Psychotherapy (Division 29) and found that a
marginal percentage (10%) of a total of 279 therapists
that even this sample of researcherclinicians perceives
reported that psychotherapy research articles or presen-
empirical research to be less helpful to them as clini-
tations were their primary source of guidance as clini-
cians than a variety of other sources of information,
cians. The most highly ranked sources of information
including their ongoing experience working with clients.
were ongoing experience with clients (48%), fol-
On the other, we found that they do value empirical lowed by theoretical publications and presentations
research, but consider it as one source of information (17%). Morrow-Bradley and Elliott (1986) also asked
among others to be incorporated into the clinical deci- their sample to rate the usefulness of seven different
sion-making process. Implications of the findings are therapy research areas (e.g., outcome research that
discussed with respect to a number of dimensions, compares a treatment with a control group and or
including the nature of professional expertise. other treatment approaches, process-outcome research
Key words: professional expertise, research and that links the process of therapy to differential out-
practice, researchpractice gap, research utilization. comes). They found that the research areas endorsed as
[Clin Psychol Sci Prac 18: 357371, 2011] most useful focused on topics such as the process of
change and the therapeutic alliance. Their respondents
Over the past several decades, a large literature has expressed less interest in research comparing the out-
accumulated demonstrating that psychotherapy of vari- comes of different forms of psychotherapy or in studies
ous types is an effective treatment for a range of psy- evaluating the effectiveness of a specific form of ther-
chological disorders (Lambert & Ogles, 2004; Lipsey & apy relative to a control group.
Wilson, 1993; Smith, Glass, & Miller, 1980). At the In that same year, Cohen, Sargent, and Sechrest
same time, research has consistently shown that practic- (1986) conducted in-depth interviews with 30 practic-
ing clinicians do not find psychotherapy research to be ing clinicians and found that empirical research
particularly relevant to clinical practice. Morrow-Brad- received the lowest clinical usefulness ratings of a
ley and Elliott (1986) conducted one of the earliest and number of different information sources, including dis-
cussions with clinical colleagues, workshops on clinical
practice, theoretical books on clinical practice, and
Address correspondence to Jeremy D. Safran, Department of
both how-to books and articles on clinical practice.
Psychology, New School for Social Research, New York,
NY 10011. E-mail: safranj@newschool.edu. Since that time, numerous studies using a variety of
2011 American Psychological Association. Published by Wiley Periodicals, Inc., on behalf of the American Psychological Association.
All rights reserved. For permissions, please email: permissionsuk@wiley.com 357
methodologies have continued to find that clinicians itation system, recently launched by the Academy of
perceive psychotherapy research as having limited rele- Clinical Science, which only accredits clinical programs
vance to their clinical practices (e.g., Beutler, Williams, that are considered science based in nature.
Wakefield, & Entwistle, 1995; Lucock, Hall, & Noble, In contrast to Baker and colleagues (2009) perspec-
2006; Mussell et al., 2000; Nunez, Poole, & Memon, tive, various critics over the years have raised questions
2003). In a more recent study, Stewart and Chambless about the relevance of much existing psychotherapy
(2007) surveyed APAs Division of Independent Prac- research to clinicians. Many criticisms fall into the gen-
tice (Division 42). They found that respondents rated eral category of poor external validity. Critics argue
their own clinical experience as significantly more rele- that randomized clinical trials (RCTs), which have
vant to making typical treatment decisions than either become the gold standard of psychotherapy research,
current research on treatment outcome, experiences in have limited clinical relevance for a number of reasons.
personal therapy, or colleagues advice. They point out, for example, that RCTs have limited
This continuing evidence of a gap between research external validity because patients are randomly assigned
and practice has been the topic of countless publica- to treatments rather than choosing their own therapists
tions over the years that bemoan its existence and sug- (e.g., Seligman, 1995). Critics also argue that the types
gest various alternative solutions (e.g., Bergin & of patients typically seen in real-world practice are
Strupp, 1972; Goldfried & Wolfe, 1996, 1998; Persons often screened out of RCTs because of comorbid diag-
& Silberschatz, 1998; Rice & Greenberg, 1984; Safran, noses. Moreover, they argue, clinicians in real-world
Greenberg, & Rice, 1988; Safran & Muran, 1994; practice find manualized treatments too inflexible, feel
Westen, Novotny, & Thompson-Brenner, 2004). It that both treatment lengths and follow-up intervals are
has also been a source of frustration for researchers. For too short in RCTs, and find that RCTs often fail to
example, a widely cited article by Baker, McFall, and capture change in ways that are clinically meaningful
Shoham (2009) compares clinical psychology to the (Goldfried & Wolfe, 1996, 1998; Persons &
prescientific state of American medicine at the time of Silberschatz, 1998; Westen et al., 2004). Other critiques
the Flexner report in the early 20th century. Citing focus on issues such as the lack of relevance of RCTs for
evidence indicating that many clinicians give more the clinician who is confronted with the question of
weight to their personal experiences than to science in how to intervene with a specific patient in a particular
making decisions about intervention, Baker et al. moment (e.g., Greenberg, 1984, 1986; Rice &
(2009) argue that although it is patent that impres- Greenberg, 1984; Safran & Muran, 1994; Safran et al.,
sionistic, clinical judgments are prey to numerous 1988).
biasesclinicians continue to use the former and Over the years, attempts to develop research
eschew the latter. In this vein, they continue, The approaches that are more relevant to the practicing clini-
upshot is that the person seeking psychological services cian have led to the development and implementation of
from a clinical psychologist cannot assume that his or a variety of different alternatives. These include system-
her treatment will be informed by the fruits of atic, research-informed case study approaches (Fishman,
the inferential deductive discipline known as science 1999; Strupp, 2001); a focus on the mechanisms of
(p. 83). change (Gendlin, 1986; Greenberg, 1986; Rice &
From Baker and colleagues (2009, p. 83) perspec- Greenberg, 1984; Safran & Muran,1994; Safran et al.,
tive, many practicing clinicians display what the authors 1988), qualitative research methods (e.g., Elliott, 1984;
regard as an insouciance in the face of the strong Hill et al., 2005; Rennie, Phillips, & Quartaro, 1988;
evidence base for the efficacy and cost-effectiveness of Stiles, 1993); and approaches such as task analysis, which
specific forms of therapy. According to them, this is integrate qualitative and quantitative research methods
harming the field and marginalizing clinical psychology (e.g., Greenberg, 1986; Rice & Greenberg, 1984).
within the health care system. To solve this problem, Advocates for the use of these methods argue that
Baker et al. (2009) advocate for the more widespread clinical trials that use group data to test hypotheses about
acceptance of an alternative to the current APA accred- the effectiveness of different forms of therapy fail to
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V18 N4, DECEMBER 2011 358
capture the complexity and subtle nuances of clinical therapy research than the average clinician. The
process and tend to emphasize hypothesis testing at the objective of this study was thus to survey respondents
expense of the discovery-oriented aspects of research (e.g., who are actively involved both as clinicians and as psy-
Elliott, 1984; Greenberg, 1986; Hill et al., 2005; Rennie chotherapy researchers. We were interested in the
et al., 1988; Stiles, 1993). They also argue that qualita- question of whether such respondents would be more
tive methods and approaches that incorporate qualitative likely than their less research-oriented colleagues to
methods can make use of the researchers capacity for see research as clinically relevant. We also wondered
rigorously observing the process of change, or what if a sample of clinicianresearchers who are particu-
Rice and Greenberg (1984) have referred to as patterns of larly likely to be familiar with some of the recent
change that recur across multiple cases, and therefore the developments in psychotherapy research described ear-
opportunity for discovery of phenomena that are not lier would be more likely to find research to be clini-
already known (e.g., Elliott, 1984; Greenberg, 1986; cally useful.
Hill et al., 2005; Rennie et al., 1988; Stiles, 1993). For these reasons, we decided to survey members
of the Society for Psychotherapy Research (SPR).
RATIONALE FOR THIS SURVEY Founded in 1969 as an international multidisciplinary
Most surveys on psychotherapy research utilization scientific organization, SPR has from its inception
have been conducted with therapists whose professional emphasized the importance of integrating research and
time is heavily devoted to the practice of psychother- practice (Orlinsky, 1995). There is a long-standing
apy, and who are not actively engaged in reading and tradition of SPR members being actively involved both
critically evaluating the psychotherapy research litera- as researchers and consumers of research and as prac-
ture or conducting research themselves. For example, ticing clinicians themselves (Angus et al., 2010). Mem-
82% of Morrow-Bradley and Elliotts (1986) sample bers of SPR have played a significant role in
reported spending more time in practicing, supervising, developing alternative research approaches designed to
or consulting about psychotherapy than in any other produce findings that are more meaningful to the cli-
professional activity. Seventy-three percent were nician and in conducting this type of research them-
employed in private practice or other primarily clinical selves (see, e.g., the recently published collection of
settings, with only 19% working in university or medi- essays about prominent SPR members, Bringing Psycho-
cal school settings. Modal number of research articles therapy Research to Life: Understanding Change Through
read in the last month, research conferences attended the Work of Leading Clinical Researchers; Castonguay
in the last year, and total number of research publica- et al., 2010). Moreover, Psychotherapy Research, the
tions or presentations produced were all 0. Sixty-two official SPR journal, is an important publication outlet
percent reported not conducting any research. Stewart for research that has been influenced by methodologi-
and Chambless (2007) did not ask their sample to indi- cal innovations designed to increase the clinical utility
cate how much research they read or produced, but of psychotherapy research. Finally, in contrast to most
88% of their sample identified private practice as their samples previously surveyed on the topic of research
primary employment site. It seems reasonable to infer utilization, SPR is an international organization. Sur-
that like many full-time clinicians, respondents in these veying members of SPR would thus allow us to
surveys do not work in environments that encourage examine the attitudes of researcherclinicians who are
an ongoing interest in research, nor are they likely to more representative of the international community
be familiar with the latest developments in psychother- than has been the case with many psychotherapy
apy research. Moreover, they are not immersed in pro- research utilization surveys.
fessional cultures that place a high value on
psychotherapy research. METHOD
We wondered if clinicians who are more immersed Participants, Procedure, and Survey
in a psychotherapy research culture might have differ- All members of the Society for Psychotherapy
ent attitudes regarding the clinical utility of psycho- Research (SPR) were contacted via e-mail and invited
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V18 N4, DECEMBER 2011 360
psychotherapy researchers, and this distinguishes them years of involvement as researchers, years of clinical
from those respondents usually surveyed in clinical uti- experience, and total hours of clinical practice per
lization studies, who tend to fall more toward the clini- week) was then compared to the distribution of the lar-
cian end of the scientist-practitioner spectrum (e.g., ger sample with respect to these characteristics. As we
Beutler et al., 1995; Cohen et al., 1986; Lucock et al., could not assume independence of the two samples, we
2006; Morrow-Bradley & Elliott, 1986; Mussell et al., treated the data obtained from our brief survey of mem-
2000; Stewart & Chambless, 2007). bership characteristics (the second survey) as our best
Respondents represented a wide range of theoretical estimate of the population (the SPR membership as a
orientations, with psychodynamic being the most whole). We then calculated confidence intervals for
commonly endorsed (35%). Ten percent designated proportions for respondent characteristics in our first
themselves as cognitive-behavioral, 27% designated survey and evaluated the extent to which the values
themselves as being either eclectic or integrative, 15% obtained in our second survey were included in the rel-
designated themselves as humanistic, and 13% evant confidence intervals for the first (Brown, Cai, &
designated themselves as other. On this dimension, Dasgupta, 2001; Clopper & Pearson, 1934). The confi-
there were both similarities and differences between dence level was set at 95% (i.e., p < .05). The only sig-
our sample and Morrow-Bradley and Elliotts (1986) nificant differences that emerged were on the
sample, in which 20% designated themselves as characteristics of age (where we appear to have
cognitive-behavioral, 36% designated themselves as undersampled members in the 6675 years category),
eclectic, 36% psychodynamic, and 8% other.1 years of clinical experience (where we undersampled
Finally, in contrast to many research utilization stud- the more than 30 years category), clinical hours per
ies that have been published, approximately 40% of week (where we undersampled the 1115 hour cate-
our sample came from countries outside of North gory), and theoretical orientation (where we undersam-
America. Over 20 different countries were represented pled psychodynamically oriented clinicians and
in our sample. Almost 60% of respondents came from oversampled those designating themselves as eclectic).
North AmericaUnited States (52%), Canada We thus have some evidence regarding our survey sam-
(7.3%)with approximately 40% of the sample repre- ples representativeness of the larger SPR membership.
senting 20 different countries in Europe and Latin
America. The following five countries were most Questions and Rating Endorsement Percentages
highly represented in our sample: United States (52%), The first survey item asked respondents to agree or dis-
Canada (7.3%), Italy (4.9%), United Kingdom (4.1%), agree with the following statement: Psychotherapy
and Sweden (4.1%.) research has had an important impact on my clinical
Because our return rate (12%) was relatively low, practice. The respondents answered accordingly:
questions arise as to the samples representativeness of strongly disagree (scale point 1: 2%), disagree (scale
the larger SPR membership on important characteris- point 2: 7%), neutral (scale point 3: 7%), agree (scale
tics. As we were not able to find data of this type, we point 4: 45%), and strongly agree (scale point 5: 39%).
conducted a second web site survey of the SPR mem- As a reference point, Table 1 compares the current
bership to obtain it. To maximize responsiveness, com- samples response to this question to the Morrow-
pletion time for this survey was designed to be Bradley and Elliott (1986) samples response to a similar
extremely brief (23 min maximum). This brief survey question.
was e-mailed to everyone on the SPR membership list It should be noted that Morrow-Bradley and Elliott
at the time (a total of 1,173 members). This time we (1986) phrased their question slightly differently (i.e.,
obtained a return rate of 73%, which is considered Please rate the extent to which the results from psy-
acceptable (Dillman, 2006). chotherapy research influence your clinical practice),
The distribution of sample respondents from our used a 6-point rating scale, and used different anchor
research utilization survey within each of the relevant points (see Table 1). These differences make it some-
characteristic categories (e.g., age, gender, work setting, what difficult to directly compare the responses of the
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V18 N4, DECEMBER 2011 362
.003 (.05 15) indicate that respondents found both respondents to indicate the source of information that
experience with clients and supervision and consul- they found to be most useful to their clinical practice,
tation more helpful than conducting psychotherapy caution must be exercised in comparing our results to
research, theoretical publications, experience of theirs. Bearing this caveat in mind, however, we will
being a client, and research publications (all p-val- tentatively draw some inferences regarding differences
ues < .004). Cohens d effect sizes were computed, in the attitudes of the two samples in the discussion
with strengths ranging from 0.89 (difference between section of this article.
experiences with clients and research publica-
tions presentations) to 0.39 (difference between Theoretical Orientation and Research Utilization
supervision and consultation with others and con- As both Morrow-Bradley and Elliott (1986) and Stewart
ducting psychotherapy research). and Chambless (2007) found that psychodynamically ori-
Table 2 provides information regarding the percent- ented clinicians were less likely to find research to be
age of respondents who endorsed the two highest ends clinically meaningful than clinicians of other orienta-
of the ratings scale (5 = helpful and 4 = somewhat tions, we were curious whether this tendency would
helpful) and compares these findings with Morrow- persist even among our sample of psychotherapy
Bradley and Elliotts (1986) rank-order findings. researchers. We compared the responses of psychody-
If we restrict ourselves to examining responses to namically oriented therapists (n = 44) to those of respon-
the highest scale point in our survey, we see that dents identifying with all other theoretical orientations
research publications presentations are perceived as combined (n = 79) using unpaired t-tests (two-tailed).
the least helpful source of information, whereas ongo- Psychodynamic therapists rated the importance of psy-
ing experience with clients is perceived as the most chotherapy research with regard to their clinical practice
helpful experience. Morrow-Bradley and Elliotts as significantly lower than other theoretical orientations,
(1986) sample also endorsed ongoing experience with t(122) = 0.83, p < .05. Given the evidence that psycho-
clients as the most useful source of information. Inter- dynamically oriented therapists tend to find research in
estingly, our sample endorsed conducting psychother- general less relevant to their clinical practice, we were
apy research as the third most helpful source of curious as to whether they might find qualitative
information, compared to Morrow-Bradley and research more relevant than quantitative research. Using
Elliotts (1986) sample, which rated this activity as the an independent samples t-test (two-tailed), the relation-
least useful source of information. ship between theoretical orientation and the reported
When our samples responses to the two highest impact of qualitative research was explored, showing no
scale points (helpful and somewhat helpful) are significant differences, t(122) = 0.25, p >.05. We did,
summed (Table 2, column 3), we see that approxi- however, find that psychodynamically oriented therapists
mately 90% of all respondents endorsed all sources of rated the relevance of ongoing experience with clients
information as either helpful or somewhat helpful. significantly higher than other therapists, t(122) = 1.17,
Because Morrow-Bradley and Elliott (1986) only asked p < .05.
Information source Helpful Somewhat helpful Helpful or Somewhat helpful Most Useful Helpful
Ongoing experience w clients 82 11 93 48
Supervision consultation w others 68 24 92 7
Conducting psychotherapy research 46 44 90 3
Experience of being a client 43 53 96 8
Theoretical publications presentations 43 53 96 17
Research publications presentations 29 61 90 10
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V18 N4, DECEMBER 2011 364
Table 3. Most common categories of research findings (with examples and approximate percentage of total)a
Research on the therapeutic alliance and ruptures in the Research on the effectiveness of behavioral and cognitive behavioral
therapeutic alliance (24%) interventions (17%)
Early alliance predicts outcome and retention Cognitive therapy for relapse prevention in depression
Recognizing and addressing alliance ruptures Efficacy of exposure-based treatments for anxiety
Alliance repair methods can work and improve outcome Dialectical behavior therapy for borderlines
Psychodynamically oriented research (15%) Research on psychotherapy process and or mechanisms of change (12%)
Addressing defenses and its relation to outcome Impact of negative therapeutic process
The potential dangers of transference interpretations Various types of task analysis research
Research on accuracy of interpretations Findings on client deference
Therapist or treatment intervention by patient type (7%) Research demonstrating therapeutic equivalence (8%)
CBT has lower relapse rates for depression than antidepressants Neurobiological impact of psychotherapy
Severe depression responds best to therapy plus medication Brain research on amygdala and trauma
CBT performs as well as medication in various studies Brain research on how memory functions
a
Total is 96% because 4% of examples were not provided in sufficient number to be categorized.
The next three most common examples given were actively involved in the production and critical
categorized as research on psychotherapy process or consumption of psychotherapy research and the practice
mechanisms of change (12%), research demonstrating of psychotherapy would reveal less of a clinical-research
therapeutic equivalence (8%), and research on treat- gap than is commonly found in surveys of samples of
ment or therapist variables by patient type (7%). The clinicians. To this end, we surveyed members of the
next four types of examples reported were classified as Society for Psychotherapy Research regarding their
attachment related and other developmental research perceptions of the clinical utility of psychotherapy
(4%), research on emotional processes (3%), research and then asked them to contrast the impact of
research on psychotherapy and medication (3%), and psychotherapy research versus other sources of informa-
basic neuroscience research (3%). None of the other tion on their clinical practices. We asked our respon-
findings or types of research provided (4%) could be dents to (a) rate the overall impact of psychotherapy
classified into categories with five or more exemplars research on their clinical work, (b) compare the
and therefore were not classified. Examples included impacts of quantitative versus qualitative research,
research on relapse prevention, Scandinavian (c) evaluate the relative clinical utility of a variety of
psychosis research, doseeffect relationship research, different sources of information, including research,
meta-cognitive functioning in personality disorder, and (d) rate their overall degree of optimism about the
engagement and motivation procedures, and herme- future of psychotherapy research. We then asked them
neutic single case design research. to compare the impact of psychotherapy research on
their clinical work to a variety of other sources of
DISCUSSION information.
The major objective of this study was to find out As a reference point, we compared our findings
whether a survey of clinicianresearchers who are both with those obtained by Morrow-Bradley and Elliott
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V18 N4, DECEMBER 2011 366
If we were to focus exclusively on the various find- sense of humility of clinicians, and our openness to
ings described above, we might simply conclude that modifying our theories in an ongoing fashion (Elliott
even a sample of researcherclinicians shows consider- & Morrow-Bradley, 1994; Lampropoulos et al., 2002;
able skepticism about the clinical relevance of psycho- Safran & Muran, 1994). Further exploration of the
therapy research. At the same time, however, it should potentially valuable by-products of conducting psycho-
be noted that when we examine the distribution of rat- therapy research may have implications for the way in
ings for the question regarding the overall impact of which research training is incorporated into clinical
research on clinical practice (i.e., research has had an training programs. For example, placing greater empha-
important impact on my clinical practice), we see that sis on teaching students in clinical psychology programs
these ratings tend to cluster toward the top end of the to reflect more actively on what can be learned clini-
scale. Moreover, when we combine ratings for the top cally through the process of conducting psychotherapy
two points of the comparative helpfulness scale (i.e., research may facilitate a greater degree of mutual influ-
please rate the extent to which you agree about the ence between research training and clinical practice.
helpfulness of the following sources of information to This could potentially have a meaningful impact on
your clinical practice), over 90% of all respondents clinicians attitudes about research subsequent to
rated psychotherapy research (as well as other sources graduation.
of information) as either helpful or somewhat help- The findings that emerged from our qualitative
ful to their clinical practice. In addition, 84% of the analysis add additional texture to our understanding of
current sample was either somewhat optimistic or this samples attitudes toward psychotherapy research.
strongly optimistic about the potential impact of psy- The most commonly provided examples of influential
chotherapy research on clinical practice in the future. research findings fell into the category of therapeutic
Finally, when we compare SPR members responses to alliance research and research on therapeutic alliance
Morrow-Bradley and Elliotts (1986) sample of Divi- ruptures. This finding is consistent with the growing
sion 29 members responses to the single question emphasis on the importance of the therapeutic
regarding the impact of research on their clinical prac- relationship across a variety of psychotherapy traditions
tice, we find that substantially more SPR members (Goldfried & Safran, 1986; Wolfe & Goldfried, 1988)
than Division 29 members sampled by Morrow- and with initiatives such as the APAs Division of Psy-
Bradley and Elliott (1986) endorsed the top two points chotherapy Task Force on the identification of empiri-
(more impactful) of the rating scale (84% vs. 27%), cally supported elements of the therapeutic relationship
and substantially fewer endorsed the bottom two scale that contribute to good therapeutic outcome (Norcross,
points (9% vs. 31%). 2002, 2011). It is also consistent with Morrow-Bradley
Another interesting finding was that examining the and Elliotts (1986) findings that clinicians ranked
highest point on the rating scale, we see that respon- research that focuses on the development and impact
dents rated the activity of conducting psychotherapy of the therapeutic or helping alliance near the top in
research in and of itself considerably more relevant to terms of perceived usefulness of psychotherapy
their clinical practice than reading psychotherapy research topics.
research (46% compared to 29%) and as relevant to their Another finding emerging from our qualitative anal-
clinical practice as other sources of information, such as ysis is that when we group together examples classified
the experience of being a client or reading psycho- within the categories of research on the therapeutic
therapy theory. This finding suggests that the process alliance and ruptures in the alliance, change processes
of conducting psychotherapy research may actually be or mechanisms, and psychodynamically oriented
of greater clinical value than the product of psycho- research (which tended for the most part to focus on
therapy research. As a number of authors have indi- therapeutic processes or change mechanisms), we find
cated, the process of conducting psychotherapy that over 50% of the examples provided fall within the
research may contribute to our development as clini- general category of research on psychotherapy process
cians in a number of ways, including increasing our and change mechanisms rather than psychotherapy
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V18 N4, DECEMBER 2011 368
somewhat different format for the rating scale and somewhat optimistic or strongly optimistic about
anchors for one of the key questions (i.e., evaluating the potential impact of psychotherapy research in the future.
the overall impact of research on clinical practice) A second possibility to consider is that it may actu-
makes it critical that any inferences about similarities ally be the case that optimal use of information gener-
and differences in the attitudes of the two samples on ated by empirical research is as one source of
this question be drawn cautiously. Moreover, it would information to be combined by the clinician with oth-
have been helpful to be able to compare our findings ers (e.g., clinical experience, experience in clinical
with a more current survey of Division 29 members supervision, existing theory). Since the publication in
attitudes toward psychotherapy research. A final con- the early 1980s of Donald Schons classic work on the
cern is that respondents answers to the open-ended nature of professional expertise (Schon, 1983), an
question about examples of clinically meaningful important field of interdisciplinary investigation has
research findings were categorized and classified emerged that studies the process through which experts
through qualitative analysis procedures, and no inter- make decisions that guide their actions in real-world
rater rating of category classification was calculated. situations that are complex and ambiguous in nature
The category percentages provided in the results sec- (e.g., Collins, 2010; Dreyfus & Dreyfus, 1986;
tion are thus only rough estimates inevitably influenced Ericsson, 2009; Klein, 1998, 2009). Research in this
by our subjectively based categorization of the items. field (commonly referred to as the field of naturalistic
The above limitations notwithstanding, the current decision-making) has consistently found that skilled
survey does shed some light on the attitudes of a sam- practitioners across a wide range of professions respond
ple of researcherclinicians regarding the clinical rele- to relevant situations in a flexible, creative, and contex-
vance of psychotherapy research. Depending upon how tually sensitive fashion. Unlike novices, who tend to
one looks at the findings, one can see the glass as either apply procedures (based on theory, research findings,
half empty or half full. or both) in a standardized fashion, experts engage in
Given the active involvement and investment of this what Schon (1983) refers to as reflection-in-actionthat
sample in the enterprise of psychotherapy research, and is, principles of procedure are continuously elaborated
the complexity of their attitudes (described earlier), it in response to an ongoing appraisal of the emergent
would be difficult to argue that they display what situation. Knowledge acquired from multiple sources,
Baker et al. (2009, p. 83) have termed an insouci- including training, theory, research, and professional
ance in the face of research evidence. It seems rather experience, is stored at a tacit level and establishes the
that they value research, but only as one of many context for a type of holistic pattern analysis that permits
sources of information relevant to their clinical work them to see emerging possibilities for intervention
(and in some respects less than a number of other (Safran & Muran, 2000). Whether or not this is true
sources of information, including personal clinical for skilled psychotherapists is an empirical question.
experience). Given the fact that the sample is clearly
conversant with the latest research, it is difficult to NOTES
attribute the findings to a dissemination problem.
How then can we account for our findings? 1. Direct comparison of the two samples with respect to
One possibility is that unwillingness of this sample theoretical orientation is complicated by the fact that
to privilege research findings as a source of influence Morrow-Bradley and Elliott (1986) did not provide respon-
on their clinical practice in part reflects limitations in dents with the option of designating themselves as humanis-
tic.
the current state of the art of psychotherapy research.
2. A social science researcher with considerable expertise
Despite consistent efforts to improve the clinical utility
in qualitative research.
of psychotherapy research over the last few decades,
3. The senior author is psychoanalytically oriented and
we cannot assume that future developments in research conducts research on the therapeutic alliance and other
methodology will not further increase its clinical aspects of psychotherapy process and outcome. The three
utility. In fact, a total of 84% of our sample was either
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V18 N4, DECEMBER 2011 370
Lampropoulos, G. K., Goldfried, M. R., Castonguay, L. G., Rice, L. N., & Greenberg, L. S. (1984). Patterns of change:
Lambert, M. J., Stiles, W. B., & Nestoros, J. (2002). What Intensive analysis of psychotherapy process. New York:
kind of research can we realistically expect from practitio- Guilford Press.
ners? Journal of Clinical Psychology, 58, 12411264. Safran, J. D. (2001). When worlds collide: Psychoanalysis and
Lipsey, M. W., & Wilson, D. B. (1993). The efficacy of the empirically supported treatment movement. Psychoana-
psychological, educational, and behavioral treatment. lytic Dialogues, 11, 659681.
American Psychologist, 48, 11811209. Safran, J. D., & Aron, L. (2001). The implications of the
Lucock, M. P., Hall, P., & Noble, R. (2006). A survey of empirically supported treatment controversy for psycho-
influences on the practice of psychotherapists and clinical analysis. Psychoanalytic Dialogues, 11, 571582.
psychologists in the UK. Clinical Psychology and Psychother- Safran, J. D., Greenberg, L. S., & Rice, L. N. (1988). Inte-
apy, 13, 123130. grating psychotherapy research and practice: Modeling the
Luyten, P., Blatt, S. J., & Corveleyn, J. (2006). Minding the change process. Psychotherapy, 25, 117.
gap between positivism and hermeneutics in psychoana- Safran, J. D., & Muran, J. C. (1994). Toward a working
lytic research. Journal of the American Psychoanalytic Associa- alliance between research and practice. In P. F. Talley,
tion, 54, 571610. H. H. Strupp, & J. F. Butler (Eds.), Psychotherapy research
Morrow-Bradley, C., & Elliott, R. (1986). The utilization of and practice (pp. 206226). New York: Basic Books.
psychotherapy research by practicing psychotherapists. Safran, J. D., & Muran, J. C. (2000). Negotiating the therapeutic
American Psychologist, 41, 188197. alliance: A relational treatment guide. New York: Guilford Press.
Mussell, M. P., Crosby, R. D., Crow, S. J., Knopke, A. J., Schon, D. (1983). The reflective practitioner. New York: Basic
Peterson, C. B., Wonderlich, S. A., et al. (2000). Utiliza- Books.
tion of empirically supported psychotherapy treatments Seligman, M. E. (1995). The effectiveness of psychotherapy:
for individuals with eating disorders: A survey of psy- The Consumer Reports study. American Psychologist, 50,
chologists. International Journal of Eating Disorders, 27, 230 965974.
237. Smith, M. L., Glass, G. V., & Miller, T. I. (1980). The benefits
Norcross, J. (Ed.). (2002). Psychotherapy relationships that work: of psychotherapy. Baltimore: Johns Hopkins University Press.
Therapist contributions and responsiveness to patients. New Stewart, R. E., & Chambless, D. L. (2007). Does psychother-
York: Oxford University Press. apy research inform treatment decisions in private prac-
Norcross, J. (Ed.). (2011). Psychotherapy relationships that work tice? Journal of Clinical Psychology, 63, 267281.
(2nd ed.). New York: Oxford University Press. Stiles, W. B. (1993). Quality control in qualitative research.
Nunez, N., Poole, D. A., & Memon, A. (2003). Psychologys Clinical Psychology Review, 13, 593618.
two cultures revisited: Implications for the integration of Strauss, A., & Corbin, J. (1998). Basics of qualitative research:
science with practice. Scientific Review of Mental Health Grounded theory procedures and techniques (2nd ed.). Thou-
Practice, 2, 819. sand Oaks, CA: Sage.
Orlinsky, D. (1995). The graying and greening of SPR: A Strupp, H. H. (2001). Implications of the empirically sup-
personal memoir on forming the Society for Psychother- ported treatment movement for psychoanalysis. Psychoana-
apy Research. Psychotherapy Research, 5, 343350. lytic Dialogues, 11, 605619.
Persons, J. B., & Silberschatz, G. (1998). Are results of ran- Westen, D., Novotny, C. M., & Thompson-Brenner, H.
domized controlled trials useful to psychotherapists? Journal (2004). The empirical status of empirically supported
of Consulting and Clinical Psychology, 66, 126135. psychotherapies: Assumptions, findings, and reporting in
Rennie, D. L., Phillips, J. R., & Quartaro, G. K. (1988). controlled clinical trials. Psychological Bulletin, 130, 631663.
Grounded theory: A promising approach to conceptualiza- Wolfe, B. E., & Goldfried, M. R. (1988). Toward a more
tion in psychology. Canadian Psychology, 29, 139150. clinically valid approach to research. Journal of Consulting
Rennie, D. L., Watson, K. D., & Montiero, A. M. (2002). and Clinical Psychology, 66, 143150.
The rise of qualitative research in psychology. Canadian
Psychology, 43, 179189. Received September 19, 2011; accepted September 21, 2011.