Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

Copyright © The British Psychological Society

Reproduction in any form (including the internet) is prohibited without prior permission from the Society

The
British
Psychological
British Journal of Health Psychology (2008), 13, 1–7
q 2008 The British Psychological Society
Society

www.bpsjournals.co.uk

Exploring the boundary conditions of expressive


writing: In search of the right recipe

Joshua M. Smyth1* and James W. Pennebaker2


1
Department of Psychology, Syracuse University, New York, USA
2
Department of Psychology, University of Texas at Austin, Texas, USA

Imagine that you are experimenting with new recipes for a tasty sauce to complement
broccoli. One evening, you wander into the kitchen wearing new shoes. You pull a host
of ingredients and spices from your cupboard and refrigerator and mix them together in
a haphazard way. You cook the concoction at a low temperature in the oven in an open
saucepan for about 25 minutes. Voila! You have created the perfect sauce.
How can you and your friends recreate this amazing sauce? Which ingredients,
cooking methods, and implements were most critical to the success of the sauce? There
are a dizzying number of variables and theories that may partially explain the flavour and
texture. Some sound reasonable, such as cooking at a low temperature, and are assumed
to be necessary and may never be subjected to culinary scrutiny. Others, which may
appear to be superfluous, such as the new shoes, may be discarded by most subsequent
cooks regardless of their potential contribution. In the grand scheme of things, the
essence of the sauce may simply be a unique chemical reaction of two of the ingredients.
However, these reactions could come from a quirk in one of the ingredients or may
interact with the smell of the leather from your shoes. Only dozens of cooking
experiments, each of which varies an ingredient, a procedure, or even shoes, can tell.
In many ways, finding the true recipe for the sauce is the same problem we face in
understanding the effectiveness of the writing paradigm. Expressive writing methods
often do, yet sometimes do not, work. As researchers have explored some of the
boundary conditions of the paradigm, it is clear that many of our implicit assumptions
about when and why it works may not be accurate. As such, it seems time to revisit some
of the ideas surrounding the original recipe.
Over 20 years ago, the first expressive writing study found that individuals who were
asked to write for 15 minutes a day for four consecutive days showed improved health
over the next few months when compared with control participants who wrote about
superficial topics (Pennebaker & Beall, 1986). The prospect of a relatively simple
experimental procedure that could improve people’s physical health soon caught the

* Correspondence should be addressed to Joshua M. Smyth PhD, Department of Psychology, Syracuse University, Syracuse,
NY 13244-2340, USA (e-mail: jmsmyth@syr.edu) or James W. Pennebaker, Department of Psychology, University of Texas at
Austin, Austin, Texas 78712, USA (Pennebaker@mail.utexas.edu).

DOI:10.1348/135910707X260117
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society

2 Joshua M. Smyth and James W. Pennebaker

imagination of researchers around the world. Dozens upon dozens of studies were
published using the expressive writing paradigm on a wide range of participants, and
using varied instructions, settings, outcome measures, and theoretical frameworks.
By now, well over 200 expressive writing studies have been published with hardly
any two experiments alike. There have been nearly enough meta-analyses on the
expressive writing studies to conduct a meta-meta-analysis. The most recent and
comprehensive meta-analysis by Frattaroli (2006) indicates that the overall effect size of
expressive writing is quite modest – a Cohen’s d around .08. For a traditional laboratory
researcher, such an effect size is viewed as almost insignificant. However, for any
intervention, particularly such a brief one, to have an effect on a meaningful outcome
several weeks later is actually impressive. Nevertheless, it is clear that expressive writing
works some times and not others.
Because the first writing studies yielded clear changes in objective health markers,
many of us assumed that successful studies would require a reasonably faithful
replication of the original procedures to reproduce the findings. Further, as there was
not an agreed-upon theory that predicted or explained the results, the experimental
procedure itself served as the architecture to build later explanatory mechanisms – for
both better and worse. In other words, the sauce was tasty but none of us really could
explain why this was true.
Given the large number of studies that have found positive effects of writing, as well
as another group of studies that have failed to find anything, it is an appropriate time to
stand back and assess some of the boundary conditions of writing. At the same time, it is
important to explore new questions that have arisen as a result of the first writing
studies. We also need to acknowledge that many of the implicit assumptions of the
original writing paradigm, what we assumed to be the key ingredients to the sauce, are
simply wrong.

Re-examining some assumptions of expressive writing


Much like strolling into the kitchen with new shoes to cook up a new recipe, the
original writing study was an inductive exercise. The depth of the hypothesis was
‘Hmmm, I wonder if asking people to write about traumas could improve their health?’
Because the answer turned out to be yes, there was not any serious inspection of the
assumptions of the original question. Yet, we believe the time has come to more
carefully examine issues surrounding expressive writing. Some of those central
questions about the paradigm that need revisiting include:
Do people need to write about traumatic and/or negative experiences? Probably
not. A number of studies have now had people write about positive as well as negative
experiences with comparable effects (e.g. Burton & King, 2004; King & Miner, 2000;
Low, Stanton, & Danoff-Burg, 2006; Marlo & Wagner, 1999). Others have had people
simply write about their thoughts and feelings about diagnoses of diseases, major issues
in their lives, or other topics rather than traumas per se (see Pennebaker & Chung, 2007
for review). Even having people write about other people’s traumas, rather than their
own, seems to have palliative effects (Greenberg, Stone, & Wortman, 1996).
Is this a process that requires several days to accomplish? Probably not. The original
writing studies assumed that writing once a day for several days would allow people
time to process or work through their writing topics. Along these lines, in the original
meta-analysis by Smyth (1998), there was preliminary evidence to suggest that if the
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society

Boundary conditions of expressive writing 3

days over which people wrote were spaced out, the effects were stronger. Studies
testing this in an experimental fashion, however, failed to support this idea (e.g. Sheese,
Brown, & Graziano, 2004). As is apparent for at least two of the studies in the special
issue (King; Chung & Pennebaker), positive effects accrue if people write on three
occasions over a single hour or even as briefly as for 5 minutes on different days.
Although there has not been systematic study of the number of writing sessions,
Greenberg et al. (1996) found positive effects for a single writing time. If all of these
findings are to be trusted, we must ask ourselves about the nature of “working through”
or, more generally, if repeated or lengthy writing sessions are even necessary at all.
Is there a single theoretical process that explains the writing findings? Probably not.
Much of scientific thinking is, often appropriately, guided by the law of parsimony. That
is, we should embrace a simple explanation over multiple or complex explanations.
With due respect to Occam, the problem is that we are studying a real world
phenomenon, in real people, as opposed to an esoteric principle or theory. Multiple
interacting factors are likely driving the effectiveness of expressive writing. There is
surely a cascade of dimensions to writing that must be appreciated. For example,
writing forces people to label and acknowledge an emotional experience, which results
in a more complex cognitive representation of the event and surrounding emotions,
which, in-turn, may affect working memory. There are undoubtedly emotional changes
going on as well; abundant evidence shows processes consistent with habituation.
Writing may also affect sleep and produce a host of brain changes (e.g. De Moor et al.,
2002; Harvey & Farrel, 2003).
The writing paradigm appears to affect people along multiple dimensions.
Depending on one’s level of analysis, there will be different ways of explaining the
effects of writing (e.g. explanations at cellular, physiological, emotional, cognitive,
behavioural, interpersonal, etc. levels of analysis). Drawing on the sauce recipe analogy,
we can come up with overlapping explanations that could, in theory, invoke molecular,
chemical, functional, temporal, stylistic, or cultural explanatory models. All might be
true but none would be complete – the secret of the sauce is likely the interaction of
multiple factors across multiple levels of a complex system.
Is expressive writing only effective on certain people? It is hard to tell. Early research
suggested that people more likely to have trouble feeling or expressing their emotions to
others were more likely to benefit from writing (e.g. males: Smyth, 1998; alexithymics:
Paez, Velasco, & Gonzalez, 1999; hostility: Christensen & Smith, 1993; openness: Lewis
et al., 2005). However, these patterns of effects have often not held up or, on occasion
have been found to reverse (cf. Frattaroli, 2006; Norman, Lumley, Dooley, & Diamond,
2004; Smyth, Anderson, Hockemeyer, & Stone, 2002). At this point, then, there is no
clear picture of which personality or individual difference dimensions may moderate
expressive writing. In general, then, writing remains effective across a wide range of
participant characteristics. Thus, the aspect which is least understood, its ability to be
effective without any ‘must haves’ or requirements, is also possibly the most valuable.
The very thing that we have not been able to confidently identify is what
makes this intervention so promising as an intervention. Whatever that underlying
mechanisms may be, the writing technique is extremely malleable to differing situations,
circumstances, participant conditions, and people.
Is expressive writing culture-bound? Possibly, but not in the way most researchers
might think. In reality, successful writing studies have been reported across multiple
countries, languages, and ethnicities. Our belief is that people dealing with strong
emotional and cognitive reactions to important events, likely to be a universal condition,
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society

4 Joshua M. Smyth and James W. Pennebaker

can benefit from translating the experience into narrative language. Such translation
may occur through writing, talking with others, prayer, or other narrative and disclosure
processes. Having said this, there is reason to suspect that the traditional writing
paradigm may be becoming less effective because the idea of writing (or otherwise
disclosing) about stress and worries has become so commonplace – at least in many
Western societies.
We are not so narcissistic as to claim that the early writing work has caused the
culture to change in its degree of expressiveness. Rather, the popularity of expressive
writing has coincided with a cultural wave where being more open and emotionally
expressive is normative. Indeed, among our undergraduate students, the majority now
report that they have been encouraged to write about their problems for years.
The informational transformation resultant from the Internet also contributes –
MySpace, Facebook, and other social networking tools provide tremendous
normative opportunities for written (and other) disclosure. Changes such as these
(and many more not delineated here) may be altering the nature and meaning of
expressive writing. In fact, the recipe to create the perfect sauce may also itself be
changing over time.

The special issue


This Special Issue presents a selection of recent studies that have attempted to push
the boundaries of what we know about writing. Each of the studies explores
various ‘ingredients’ to the expressive writing sauce, and collectively represents three
broad themes of exploration: Methodology, Mediators and Moderators, and New
Outcomes.

Methodology
Methodological features have been continually tested and modified to test the
contextual limits of expressive writing, but uncertainties remain concerning the
structure of administering the intervention as well as the specific writing instructions.
For example, several studies raise some important questions about the nature of timing.
It appears that writing is effective even at drastically reduced lengths (i.e. 2 minutes;
Burton & King), suggesting that even small sessions of writing may ‘kick-start’ processes
(emotional, cognitive, and social) that can lead to benefits. Similarly, a single session of
writing for 10–15 minutes has been shown to effectively reduce negative emotions
about a traumatic event (when using positive framing – positive emotion words;
Fernandez & Paez). The spacing between writing sessions also seems to be flexible.
An investigation of spacing between writing sessions found that 1-hour intensive writing
sessions produced similar physical health benefits to the traditional 3-day procedure
(Chung & Pennebaker); although participants themselves perceived the writing as less
helpful in this shortened time span. These studies raise issues about what occurs
between writing sessions that may be relevant for the intervention to work. That is,
what (if any) cognitive, emotional, and/or social processes occur between writing
sessions (and are these time dependent)? which, if any, are necessary for improvement?
and what is the ‘best’ length and number for a writing intervention? Other variations,
such as single 30-minute writing sessions, point to boundaries of certain individuals (i.e.
personality characteristics) that may influence the writing experience (Cohen et al.).
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society

Boundary conditions of expressive writing 5

Another such example is the study by Corter and Petrie, which highlights the effect of
context on participants’ language use and experience.
Another point of manipulation in the methodology is the writing itself. Different
writing instructions are often used to test substitutions for the original ingredients. One
study instructed participants to write in different point of view perspectives
(first-person pronoun, second-person pronoun, and then third person-pronoun) in a
consecutive sequential order (Seih). The results indicate that the writing produced
benefits, especially among people with high anxiety, and demonstrated a three-step
process to success in which emotional experiences are directly expressed first, then
supported, and ultimately transformed. Participants have also been instructed to write
in more than one language. Kim instructed Korean-English and Spanish-English bilingual
students to expressively write in their native language, English, or both languages on
four occasions. In the end, participants who switched between writing in English and
their native language described the experience as more personal and emotional, and
were more socially engaged at follow-up, than students who only wrote in one language.
The study by Austenfeld and Stanton suggests that patients’ preferred emotion
regulation strategies should be considered in implementing expressive writing, claiming
that its benefits are largely dependent on matching the imposed coping strategy
(i.e. writing instructions) to the individual. Another study suggests that emotional and
cognitive word use predicts psychological outcomes in expressive writing. Van
Middendorp and Geenen examined associations between alexithymia and emotional
and cognitive word use with subsequent health. They suggest that poor cognitive–
emotional processing may impede the outcome of emotional disclosure interventions.
As seen here, alterations in the methodology of expressive writing sometimes reveals
not only the intervention’s strengths, but also its limitations. In this way, methodology is
important because it can be uniquely adjusted to the specific needs of a study, and also
because this malleability allows expressive writing to be administered in a wide range of
contexts and across individuals that vary in potentially important ways.

Mediators and moderators


Some studies focus directly on the nature of mediators or moderators, such as person
characteristics or coping styles. Moderators can be identified within a specific
population to identify potential strategies of tailoring the intervention for maximal
benefit. For instance, expressive writing was used to examine outcomes in a stigmatized
group, homosexual men, demonstrating that avoidant coping and somatic symptoms
improved for this particular group of socially constrained individuals (Swanbon &
Boyce). The effectiveness of EW was found to vary according to coping style and
educational level of patients with fibromyalgia, suggesting that subgroups of patients
may be differentially benefited by EW and extending the clinical relevance of EW
interventions (Junghaenal & Broderick). Personality markers (i.e. alexithymia, splitting,
repressive coping), too, are shown to differentially affect expressive writing benefits;
although significant effects were only found for objective physical health quality
(e.g. physician-assessed illness, self-reported health visits), not other physical and
psychological health outcomes (Baikie).
Placing moderators and mediators at the focus of many studies has resulted in many
different flavours, or psychological and physical health outcomes (the latter of which is
one component of expressive writing that has consistently been a main focus of
research). One example is a study that looked at how emotional approach coping and
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society

6 Joshua M. Smyth and James W. Pennebaker

self-efficacy could moderate the effects of expressive writing in people with migraine
headaches (Kraft et al.). Another study by O’Cleirigh and colleagues assigned individuals
living with HIV/AIDS to write about past traumas, and found that higher levels of
emotional disclosure and trauma processing mediate health and immunological benefits
in expressive writing. A best possible future self (BPS) writing topic, along with more
traditional trauma writing, was included in one study to test if EW could improve
working memory capacity (WMC) (Yogo & Fujihara). Although BPS writing groups
showed no improvements in WMC, the traumatic experience writing group did improve
WMC. This demonstrates expressive writing’s extended benefits across cultures and in
collectivist societies (e.g. Japan), and further suggests that varying instructions may be
differentially related to outcome types (in this case, cognitive function).

New outcome measures


Yet, other studies focus on exploring the boundary conditions of expressive writing in
terms of the outcome measures themselves. Weinman and colleagues studied the
healing process of a small punch biopsy wound, comparing rates of recovery between
an expressive writing and a control group. Results showed that participants assigned to
the expressive writing group healed significantly faster by week 2 and 3 compared to the
control group. This study holds significant meaning for expressive writing, providing
additional evidence for its beneficial impact not only for the numerous physiological
systems within the body, but also for the clinical recovery from physical wounds.
Another extension to the range of outcomes is the examination of patients with post-
traumatic stress disorder (PTSD) and endocrine stress responses (cortisol; Smyth,
Hockemeyer, & Tulloch). Expressive writing, at least in this case, appeared safe to use in
this clinical sample of volunteers with PTSD. Additionally, those assigned to the EW
group demonstrated significant improvement in regulating their emotional responses,
and cortisol responses to re-exposure to trauma memories were significantly attenuated
following EW. This study is an example of how EW boundaries can be tested for
feasibility, safety, and efficacy, and raises some precautionary measures that should be
used in proceeding with EW interventions in previously unexamined and at-risk
populations.

Concluding remarks
After 20 years and hundreds of studies, most of us still acknowledge that expressive
writing works – at least some of the time and for some people. The original sauce is not
as startling as it used to be. There are clearly many ways to cook it – most of which
would not have been predicted at the time of the original experiments. There are also
now a large number of important and valid questions about the original procedures,
participants, outcome measures, and underlying explanations.
As this special issue suggests, one of the most exciting aspects of the original writing
method has been its ability to generate new ways of thinking about emotions, cognitive
processing, and health. The writing paradigm has both been influenced by and has
influenced a new generation of theories, methods, and real-world applications. This
Special Issue is hardly meant as the final coda of this research enterprise. Rather, it is
meant as an opportunity to both look back at what we have learned and also to look
forward at the many remaining frontiers in EW research. The original sauce has come a
Copyright © The British Psychological Society
Reproduction in any form (including the internet) is prohibited without prior permission from the Society

Boundary conditions of expressive writing 7

long way, it has helped us appreciate far more than just broccoli, and we hope that it
continues to be utilized and scrutinized by many an investigative chef in the future.

References
Burton, C. M., & King, L. A. (2004). The health benefits of writing about intensely positive
experiences. Journal of Research in Personality, 38, 150–163.
Christensen, A. J., & Smith, T. W. (1993). Cynical hostility and cardiovascular reactivity during self-
disclosure. Psychosomatic Medicine, 55, 193–202.
De Moor, C., Sterner, J., Hall, M., Warneke, C., Gilani, Z., Amato, R., et al. (2002). A pilot study of
the effects of expressive writing on psychological and behavioral adjustment in patients
enrolled in a phase II trial of vaccine therapy for metastatic renal cell carcinoma. Health
Psychology, 21, 615–619.
Frattaroli, J. (2006). Experimental disclosure and its moderators: A meta-analysis. Psychological
Bulletin, 132, 823–865.
Greenberg, M. A., Stone, A. A., & Wortman, C. B. (1996). Health and psychological effects of
emotional disclosure: A test of the inhibition-confrontation approach. Journal of Personality
and Social Psychology, 71, 588–602.
Harvey, A. G., & Farrel, C. (2003). The efficacy of the Pennebaker-like writing intervention for poor
sleepers. Behavioral Sleep Medicine, 1, 115–124.
King, L. A., & Miner, K. N. (2000). Writing about the perceived benefits of traumatic events:
Implications for physical health. Personality and Social Psychology Bulletin, 26, 220–230.
Lewis, R. J., Derlega, V. J., Clarke, E. G., Kuang, J. C., Jacobs, A. M., & McElligott, M. D. (2005). An
expressive writing intervention to cope with lesbian-related stress: The moderating effects of
openness about sexual orientation. Psychology of Women Quarterly, 29, 149–157.
Low, C. A., Stanton, A. L., & Danoff-Burg, S. (2006). Expressive disclosure and benefit finding
among breast cancer patients: Mechanisms for positive health effects. Health Psychology, 25,
181–189.
Marlo, H., & Wagner, M. K. (1999). Expression of negative and positive events through writing:
Implications for psychotherapy and health. Psychology and Health, 14, 193–215.
Norman, S. A., Lumley, M. A., Dooley, J. A., & Diamond, M. P. (2004). For whom does it work?
Moderators of the effects of written emotional disclosure in women with chronic pelvic pain.
Psychosomatic Medicine, 66, 174–183.
Paez, D., Velasco, C., & Gonzalez, J. L. (1999). Expressive writing and the role of alexithymia as a
dispositional deficit in self-disclosure and psychological health. Journal of Personality and
Social Psychology, 77, 630–641.
Pennebaker, J. W., & Beall, S. K. (1986). Confronting a traumatic event: Toward an understanding
of inhibition and disease. Journal of Abnormal Psychology, 95, 274–281.
Pennebaker, J. W., & Chung, C. K. (2007). Expressive writing, emotional upheavals, and health. In
H. Friedman & R. Silver (Eds.), Handbook of health psychology (pp. 263–284). New York:
Oxford University Press.
Sheese, B. E., Brown, E. L., & Graziano, W. G. (2004). Emotional expression in cyberspace:
Searching for moderators of the Pennebaker disclosure effect via email. Health Psychology, 23,
457–464.
Smyth, J. M. (1998). Written emotional expression: Effect sizes, outcome types, and moderating
variables. Journal of Consulting and Clinical Psychology, 66, 174–184.
Smyth, J., Anderson, C., Hockemeyer, J., & Stone, A. (2002). Does emotional non-expressiveness or
avoidance interfere with writing about stressful life events? An analysis in patients with
chronic illness. Psychology and Health, 17(5), 561–569.

You might also like