ArticlePDF Available

Abstract and Figures

Background The Best Possible Self is a Positive Psychology Intervention which asks participants to write down about themselves in their best possible future. Previous studies have shown its efficacy to enhance wellbeing, but the mechanisms that underlie its efficacy are still unknown. Objective The aim of this study was to analyze the content of the essays of the BPS intervention and to examine how this content was related to the efficacy of the intervention to increase positive affect. Method Participants (N = 78) were randomized to either the Best Possible Self condition, or one of two variants of the intervention: one’s best self in the present, and one’s best self in the past. Qualitative analyses of the texts were carried out to explore the main themes and features of the essays. Then, a mixed-methods approach with quantitative and qualitative data was followed, in order to analyze the relationship between the content of the texts and the change in positive affect produced by the interventions. Results Significant differences between conditions were found in the content of the compositions. Regression analyses showed that different variables predicted the change in positive affect depending on the condition. Mediation analyses also found differences among conditions. Conclusions These findings suggest that these interventions respond to different underlying mechanisms which influence their efficacy. This study contributed to a better understanding of how Positive Psychology Interventions work, and how to increment their efficacy.
Content may be subject to copyright.
RESEARCH ARTICLE
Qualitative analysis of the Best Possible Self
intervention: Underlying mechanisms that
influence its efficacy
Alba CarrilloID
1
, Marian Martı
´nez-Sanchis
1
, Ernestina Etchemendy
2
, Rosa M. Baños
1,3
*
1Department of Personality, Assessment and Psychological Treatments, University of Valencia, Valencia,
Spain, 2Department of Psychology and Sociology, University of Zaragoza, Teruel, Spain, 3CIBER
Fisiopatologı
´a Obesidad y Nutricio
´n (CIBEROBN), Instituto Carlos III, Madrid, Spain
*banos@uv.es
Abstract
Background
The Best Possible Self is a Positive Psychology Intervention which asks participants to write
down about themselves in their best possible future. Previous studies have shown its effi-
cacy to enhance wellbeing, but the mechanisms that underlie its efficacy are still unknown.
Objective
The aim of this study was to analyze the content of the essays of the BPS intervention and
to examine how this content was related to the efficacy of the intervention to increase posi-
tive affect.
Method
Participants (N = 78) were randomized to either the Best Possible Self condition, or one of
two variants of the intervention: one’s best self in the present, and one’s best self in the past.
Qualitative analyses of the texts were carried out to explore the main themes and features
of the essays. Then, a mixed-methods approach with quantitative and qualitative data was
followed, in order to analyze the relationship between the content of the texts and the
change in positive affect produced by the interventions.
Results
Significant differences between conditions were found in the content of the compositions.
Regression analyses showed that different variables predicted the change in positive affect
depending on the condition. Mediation analyses also found differences among conditions.
Conclusions
These findings suggest that these interventions respond to different underlying mechanisms
which influence their efficacy. This study contributed to a better understanding of how Posi-
tive Psychology Interventions work, and how to increment their efficacy.
PLOS ONE | https://doi.org/10.1371/journal.pone.0216896 May 17, 2019 1 / 15
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Citation: Carrillo A, Martı
´nez-Sanchis M,
Etchemendy E, Baños RM (2019) Qualitative
analysis of the Best Possible Self intervention:
Underlying mechanisms that influence its efficacy.
PLoS ONE 14(5): e0216896. https://doi.org/
10.1371/journal.pone.0216896
Editor: Simone Rodda, University of Auckland,
NEW ZEALAND
Received: January 18, 2019
Accepted: April 29, 2019
Published: May 17, 2019
Copyright: ©2019 Carrillo et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited.
Data Availability Statement: Data files are
available at https://doi.org/10.5281/zenodo.
2662545.
Funding: This work was supported by the
Conselleria d’Educacio
´, Investigacio
´, Cultura i
Esport (Spain) under the grant “INTERSABIAS”
(PROMETEO/2018/110). The funders had no role
in study design, data collection and analysis,
decision to publish, or preparation of the
manuscript.
Introduction
Historically, individuals have made profuse efforts to achieve the road of happiness and wellbe-
ing. Lately, these efforts have crystallized in the Positive Psychology research movement,
whose aim is to provide an evidence-based framework for the study of what makes people
happy and how to bolster their wellbeing levels [1]. Although there is a lack of a unified defini-
tion of wellbeing, one of the main historical approaches proposes wellbeing as the balance
between positive and negative emotions and a high sense of satisfaction with life, also known
as subjective wellbeing (SWB) [2]. Positive Psychology Interventions (PPIs) emerged precisely
as a response to the societal need of increasing people’s overall wellbeing levels, including
SWB. This applied portion of Positive Psychology consists of activities aimed at increasing pos-
itive emotions, cognitions or behaviors [3,4]. Research on the efficacy of these interventions
has burgeoned since its beginning, and nowadays there are multiple published studies about
new and heterogeneous exercises that can help people flourish (e.g. gratitude letters, acts of
kindness, using signature strengths). Indeed, several meta-analyses have shown that PPIs are
effective approaches to increase wellbeing with small to moderate effect sizes [4,5].
Lately, as a consequence of the progression in the knowledge of these interventions,
research interest on the mechanisms that explain the efficacy of PPIs is growing. To date, some
authors have attempted to explain why and under which circumstances PPIs work, developing
some theoretical models that can be applied to all PPIs in general (e.g. [6,7]). However, this
field is considerably recent, and these models still need to be validated [8]. In addition, they
are applied to the complete range of PPIs despite their heterogeneity, hence there is a lack of
knowledge about the circumstances that make each intervention individually effective [5,6].
One of the most widely used PPIs is the Best Possible Self (BPS) intervention, in which par-
ticipants are asked to write down about their best possible self in a future where they have
achieved everything desired, after working hard towards it. This intervention was developed
initially by King [9], and it was based on the trauma writing paradigm, which had found that
writing sessions about upsetting and negative topics (as a traumatic event) produced both
physical and mental health improvements [10,11]. As a response to the emerging interest in
the positive side of life [1], the focus of attention in research changed from the trauma writing
paradigm to the positive writing paradigm (i.e., writing about positive topics) and its effects on
wellbeing, being the BPS intervention one of its main examples. Based on the writing paradigm
of Pennebaker, King [9] developed this intervention and compared it with a writing disclosive
exercise about a traumatic event. Results showed that BPS intervention produced the same
benefits as trauma-focused writing on health. This intervention, in addition, produced signifi-
cant increases in positive mood and wellbeing, and participants in this condition rated the
exercise as less upsetting than the trauma condition participants. These results are consistent
with the last meta-analysis about disclosive writing, which found no significant differences
between interventions focused on disclosing negative events and the ones focused on disclos-
ing positive events on psychological and health benefits [10]. As the author of this work stated,
given that trauma writing paradigms usually produce temporary increases in negative affect,
choosing the disclosure of positive events may be preferable, as it avoids this short-term nega-
tive side effects and it has shown the same positive results.
Since the first approach by King, many studies have been carried out in order to test the
efficacy of this PPI. A narrative review of this intervention concluded that it seems a viable
approach to produce positive outcomes on wellbeing, although little is known about how this
positive activity works [12]. In addition, a recent meta-analysis about 28 studies showed that
BPS is an efficacious intervention to improve wellbeing and found moderate effect sizes of BPS
over control groups on positive affect (d = .339 and d = .657) [13]. However, analyses of
Qualitative analysis of the Best Possible Self
PLOS ONE | https://doi.org/10.1371/journal.pone.0216896 May 17, 2019 2 / 15
Competing interests: The authors have declared
that no competing interests exist.
moderators (i.e., length, dosage, delivery method, etcetera) did not show significant results in
this review. Therefore, the characteristics of the BPS intervention that might influence its effi-
cacy are still unknown.
A methodological approach that has potential to unveil the possible mechanisms that
underlie the efficacy of a writing intervention is a qualitative analysis of its content. This
approach has the ability to uncover novel and deeper understandings of phenomena of interest
in Positive Psychology [14]. In addition, when combined with quantitative data in a mixed-
methods approach, qualitative data can help to identify significant predictors of wellbeing,
producing a more comprehensive outlook of relevant constructs and addressing questions as
why and how [15].
Recently, the benefits of positive writing have derived to an increasing interest on the quali-
tative variables of the writing tasks, although research is still scarce. In the case of the BPS
intervention, only a handful of studies explored the content that participants wrote about.
King [9] found that the BPS essays included a variety of topics, such as job success, self-
improvement, marriage and family, travel, or home ownership, although no further analyses
were carried out on the frequency of these topics. Hill and colleagues [16] analyzed the texts of
the BPS compositions in order to classify the goals included in the essays and found fourteen
categories. The most frequent goals were approach (those with references to approaching
something positive), intrapersonal (goals that mentioned only the self), and achievement (those
goals related to accomplishing a goal or achieving success). Correlation analyses were carried
out to explore the association between written goals and measures of life satisfaction and religi-
osity. Results showed that life satisfaction was negatively correlated with spirituality goals
(related to a higher power and/or to unity and justice). In addition, Loveday et al. [17] carried
out a thematic analysis of the BPS texts specifically focused on spare time using an explicit con-
ceptual framework on leisure [18]. Results showed that within the leisure area, affiliation (lei-
sure spent with other people), autonomy (leisure spent on oneself) and detachment-recovery
(leisure mentioned in relation to work) were the most frequent themes (33, 23, and 21 percent-
age of leisure sentences, respectively). However, this study only addressed the content of the
essays within the previously mentioned framework–focused on spare time, and only analyzed
the sentences coded as leisure, which represented 41% of the content, whereas the remaining
59% of the sentences categorized as non-leisure were not explored. As it can be seen, these first
approaches have explored the qualitative characteristics of the texts of the BPS essays, but they
were carried out within specific frames that might have constrained their results. Hence, there
is still a scarce knowledge about which content, in general, participants include in their essays
when they write about their BPS, and a broader approach could contribute to a better under-
standing about this subject. In addition, none of these studies have combined the content anal-
yses with quantitative data about the efficacy of the intervention, thus the role that the content
of the texts may play on its efficacy is still unknown.
This study is part of a larger project on the mechanisms that underlie the efficacy of the BPS
intervention [13], in which the role of temporality was explored. A randomized controlled trial
with three experimental conditions (the original BPS and two temporal variations: past best
self or BPAS and present best self or BPRES) showed that temporal focus did not affect the
ability of the intervention to increase positive affect as no statistically significant differences
emerged among conditions. Therefore, it is necessary to continue investigating on the under-
lying mechanisms that influence the efficacy of the BPS.
The main objective of the present work was to analyze the role of the content of the essays
of the BPS intervention on its efficacy. For this purpose, qualitative analyses of the BPS and the
BPRES and BPAS variants were carried out. More specifically, this work had two aims. On the
one hand, to analyze the content of the texts in order to identify the main themes and features
Qualitative analysis of the Best Possible Self
PLOS ONE | https://doi.org/10.1371/journal.pone.0216896 May 17, 2019 3 / 15
of the compositions of the three PPIs (BPAS, BPRES, BPS) and to explore the possible differ-
ences between conditions. On the other hand, to examine the influence that the identified
themes and features of the texts had on the efficacy of the interventions to increase positive
affect. As far as we know, this is the first study that systematically analyzes the content of the
texts of the BPS with a bottom-up approach (not forced by a predetermined model), and the
first one that includes the temporal variants of this intervention. In addition, this is the first
work that combines qualitative data of the interventions with quantitative data about their effi-
cacy. It was expected that the content of the texts would influence on the efficacy of the three
PPIs to increase positive affect. However, due to the exploratory nature of the analyses, no spe-
cific hypotheses were generated regarding the content of the texts on each intervention and
how it may affect the efficacy of the intervention.
Method
Participants
The initial sample consisted of 84 participants who were part of a larger study [13]. Their age
ranged from 18 to 40 years old (M = 20.23, SD = 4.10), and 77.2% of them were women. They
were randomized to one of three conditions: BPAS (N = 30), BPRES (N = 27), BPS (N = 27).
Two participants did not answer post-intervention assessment, and text analyses showed that
four participants did not follow the instructions of the assigned conditions. Consequently, six
participants were excluded from the study. The final sample consisted of 78 participants
(BPAS = 27, BPRES = 25, BPS = 26).
Interventions and procedure
This study included three PPIs, based on the original BPS exercise. The BPS intervention asks
participants to visualize themselves in the future after everything has gone as well as possible
[19,20]. Based on this intervention, two variants of the exercise were designed with the same
format and instructions, except for the time frame in which they were focused on. Concretely,
the Best Past Self condition (BPAS) required to recall a time in the past when participants con-
sidered they had displayed the best version of themselves, whereas the Best Present Self condi-
tion (BPRES) asked participants to think about the best version they offered to the world at the
present time.
The procedure was based in previous studies: participants were encouraged to write about
their best self and then to mentally visualize this content [19,21,22]. The complete intervention
lasted 7 days, in which participants came to the laboratory for the first session and then prac-
ticed the assigned exercise at home for one week. During the first session, participants signed
the informed consent, answered the pre-intervention assessment and listened to audiotaped
instructions of the assigned task. Regardless of the condition, they had to spend 15 minutes
writing their essay in a computer in the laboratory, and then 5 minutes visualizing its content
(their best self) [19,23]. Instructions in all conditions encouraged participants to include as
many sensorial details as possible, as the procedure included an explicit visualization compo-
nent in which they spent 5 minutes visualizing about their best self after writing about it
[19,23]. During the remaining 6 days, participants were encouraged to mentally visualize the
content of their essay once a day. After 7 days, participants received a link with the post-inter-
vention assessment.
This work was registered in the United States National Institute of Health Registration Sys-
tem (http://www.clinicaltrials.gov) with Clinical Trials Registration Number NCT03024853
and approved by the ethical committee of the University of Valencia (H1415802387094).
Qualitative analysis of the Best Possible Self
PLOS ONE | https://doi.org/10.1371/journal.pone.0216896 May 17, 2019 4 / 15
Scales
A mixed-methods approach using quantitative and qualitative methodologies was followed in
order to explore the relationship between the content and features of the texts and the change
in positive affect.
The quantitative outcome measure included was positive affect, as it has been widely used
in previous studies (e.g. [19,20,24]). The scale used to measure positive affect (PA) was the PA
subscale of the Positive and Negative Affect Scale, PANAS [25], which includes 10 positive
emotions (e.g., inspired) to measure positive mood. Respondents rate how they usually feel on
a 5-point Likert-type scale. In this study, a Spanish version was used [26].
Cronbach’s alpha for the original scale ranged from .86 to .90, and in this sample alpha
value was .90. Participants answered the scale the first day before practicing the assigned exer-
cise (pre-intervention assessment), and 7 days after the intervention started (post-intervention
assessment).
Coding of the essays
Essays were analyzed to explore two main areas. On the one hand, the content of the essays
(i.e., what did participants write about when they reflected on their best past, present or future
self). On the other hand, the features of the compositions or, in other words, how they
expressed these ideas (e.g., the number of words or its emotional valence).
The followed approach was based on the consensual qualitative research-modified
(CQR-M), a qualitative research method designed to be applied in large samples (i.e., more
than 15 participants) and relatively brief qualitative data, which can be used to describe little-
studied phenomena and establish a basis for further research. This method is defined as a bot-
tom-up approach, through which categories are derived from the data instead of forcing a pre-
determined structure on it [27]. With this method, as the authors state, a further
comprehension of the topic under research can be obtained by combining the newly described
phenomena with quantitative data.
In order to reach consensus, following the CQR-M guidelines, all team members discussed
disagreements at each step of the process. The coding team was composed by the first and sec-
ond authors (AC and MMS), who were experts on the interventions used in the study, knew
the instructions and procedure and had previously conducted studies with the included activi-
ties. The next procedure was followed: first, two independent coders (AC and MMS) read all
the essays independently and generated a list of themes and areas identified in the texts. Sec-
ondly, these themes were discussed by the researchers, and then the revised themes were
applied in the analyses of 30 randomized essays, in order to explore whether these were ade-
quate and captured all the relevant ideas. After a revision of the themes, all essays were ana-
lyzed independently by the two coders in order to categorize all the contained bins of
information with the designated themes and the subsequent areas. Interrater reliability and
frequency of themes were calculated (see Method and Results sections).
These themes were not mutually exclusive. In addition, since this analysis relied on bins of
information, they did not necessarily coincide with a complete sentence: it was possible that a
single sentence contained two ideas (for example, “the social area is very important in my life:
I like to communicate with people and I tend to be quite open and affectionate”, would be
coded as friendship and positive features), and it was also possible that the same idea expressed
in two or more sentences would be coded as one unit (for example, the two sentences “I want
to expose myself to what life brings to me. I want to feel inexperienced to able to improve”
would be categorized as positive features).
Qualitative analysis of the Best Possible Self
PLOS ONE | https://doi.org/10.1371/journal.pone.0216896 May 17, 2019 5 / 15
Themes of the texts. The final categories included could be grouped into four areas: per-
sonal, academic/professional, social, and leisure area. Regarding personal area, positive features
collected all phrases that expressed a personal improvement on one’s trait or psychological
ability, or an already present positive feature that remained constant (e.g., “In the future I
would like to have the same psychological abilities that I currently have”); skills referred to the
presence or the willingness to learn an ability or knowledge (e.g., “I would like to learn how to
play the piano or the harp”); and health was coded when participants talked about their
attempts to influence their physical health (e.g., “My best self figured out my intestinal problem
and now she’s thin and strong”). Concerning the academic or professional area, themes were
divided by the inner motives expressed in the texts, being intrinsic the content related with the
academic or professional area associated with intrinsic motives (e.g., “Now I have a job in
which I feel very happy, and I have realized that I love my job”), and extrinsic when extrinsic
motives were expressed (e.g., “I visualize myself wearing a suit and having quite a lot of
money”). With respect to the social area, friendship was coded on phrases containing social
relationships with friends or colleagues (e.g., “I felt very close to my childhood friends because
we were all going through the same phase”), family on mentions to relationships with mem-
bers of the family (e.g., “When I get home, I tell my family about my day and I hear about
theirs”), partner in the case of romantic relationships (e.g., “I had a partner with whom I
enjoyed our shared moments”), and help emerged when participants made an explicit reflec-
tion on their willingness to or their actions aimed at helping other people in different contexts
(e.g., “I decided I would watch over the happiness of others, trying to improve their lives”).
Lastly, the leisure area only included the leisure theme, which contained phrases related to how
their best selves spent their free time or practiced different hobbies (e.g., “I had time to watch
TV series and movies”).
Features of the texts. In addition, the collected features of the compositions were: length
of the essay (total number of words), quantity of sensorial details (e.g., “I was drinking tea, it
tasted stronger than usual. I added sugar and started to blow, it was so hot . . . I could see the
steam coming out of the cup”), emotional valence of the essay, and incongruousness. Emo-
tional valence was calculated as the subtraction of the total number of positive emotional states
(e.g., “It was some years ago, but the feeling still lingers: pride”, “I feel vigorous, energetic, tol-
erant and strong”) minus the number of negative emotional states (e.g., “In my future I keep
seeing a lot of stress and anxiety”, “I feel pretty demotivated in my academic life”) in each text.
Regarding incongruousness, it was coded on phrases in which participants talked about a posi-
tive feature explicitly expressed as no longer present (e.g., “I have the feeling that I enjoyed the
little things more than I do now”), or the willingness to reduce or eliminate the presence of a
personal feature (e.g., “My best self would learn not to overthink everything, because right
now I brood a lot about everything”).
Finally, all essays were coded independently by two researchers (AC and MMS). Disagree-
ments were resolved by consensus and by consultation with a third researcher expert in the
field (RMB). Intercoder reliability was assessed with Kappa coefficients and correlations
between coders for all categories. Kappa values ranged from .78 to 1, and correlations ranged
from .87 to 1 (see Table 1). These results indicate high levels of agreement [28].
Data analyses
Analyses of the texts were carried out with ATLAS.ti software for Windows (v. 7.5.4). Statisti-
cal analyses were conducted using the SPSS software for Windows (v. 24). In order to test the
differences between conditions on the content and features of the texts, two multivariate analy-
sis of variance (MANOVAs) were carried out, one for the content themes and another for the
Qualitative analysis of the Best Possible Self
PLOS ONE | https://doi.org/10.1371/journal.pone.0216896 May 17, 2019 6 / 15
text features. Pairwise comparisons using Bonferroni adjustment were conducted when signif-
icant differences were found among conditions. To examine the content themes and text fea-
tures that predicted the change in PA, a stepwise multiple regression analysis was conducted
entering the change in PA as dependent variable, and all themes and text features as indepen-
dent variables. Change in PA was calculated using pre-intervention PA scores and post-inter-
vention PA scores (i.e., change = post-intervention PA—pre-intervention PA), where positive
values for change in PA reflected an improvement. Finally, ten parallel multiple mediation
analyses (one for each theme) were performed in each condition to test whether the effect of
the content of the text on change in PA was mediated by the features of the text, using the pro-
cedure described by Hayes [29] from the PROCESS macro (version 2.16), choosing “model 4”.
In our proposed mediation models, we included the features of the texts as mediators in the
relationship between the themes of the essays and the change in PA. That is, we explored
whether the effects produced by the themes of the texts on the change in PA were mediated by
how these texts were written. These analyses were carried out for each condition. Bias-cor-
rected bootstrap 95% confidence intervals (CI) based on 5,000 samples were used to assess the
specific and total indirect effects. A CI that did not include the zero value indicated a signifi-
cant indirect effect, implying that the effect of the theme on the change in PA was mediated by
the features of the texts. Pairwise comparisons between specific indirect effects were carried
out to test whether one indirect effect was statistically different from another through the con-
fidence interval.
For both regression and mediation analyses, the frequency of participants who included
each theme and feature in their text was calculated for each condition. This was done as some
themes or features were especially uncommon in some conditions. Therefore, if a specific
Table 1. Kappa values and intercoder correlations.
Kappa values Correlation values
Themes of the texts
Personal area
Positive features .81 .97
Skills .78 .87
Health .88 .90
Academic/professional area
Intrinsic .78 .89
Extrinsic .91 .91
Social area
Friendship .78 .91
Family .91 .90
Partner .97 .92
Help .85 .92
Leisure area
Leisure .85 .92
Essay features
Positive emotional states .90 .98
Negative emotional states .80 .92
Incongruousness .90 .98
Sensorial details 1 1
Notes: For all correlations, p<.001. Positive and negative emotional states were subsequently used to calculate the
emotional valence of the texts.
https://doi.org/10.1371/journal.pone.0216896.t001
Qualitative analysis of the Best Possible Self
PLOS ONE | https://doi.org/10.1371/journal.pone.0216896 May 17, 2019 7 / 15
theme or feature appeared in less than 25% of the texts (that is, less than 7 participants of one
condition included it in their texts), it was considered that the theme/feature was no represen-
tative of the sample on that specific condition, and thus it was not included in the analyses of
that condition. For example, sensorial details were not included in the mediation analyses in
BPS condition as it appeared in less than 25% of the texts in this condition.
Results
Descriptive analyses of the themes
Means and standard deviations of each theme and feature of the texts on the different condi-
tions can be found in Table 2. Generally, the most frequent themes of the texts on the three
conditions taken together were positive features (M = 2.09, SD = 1.55), friendship (M = 1.18,
SD = 0.98), and intrinsic (M = 0.86, SD = 0.79), and the least frequent ones were skills
(M = 0.21, SD = 0.57), health (M = 0.28, SD = 0.45), partner (M = 0.37, SD = 0.58) and help
(M = 0.32, SD = 0.59). The mean valence of the essays taking all conditions was 2.21
(SD = 2.10).
Differences between conditions on the content of the texts
Table 2 shows the mean, standard deviations, and the MANOVA results for the effect of condi-
tion on the themes of the essays. The MANOVA revealed that, using Pillai’s trace, there was a
significant effect of condition on the presence of the different themes, V= 0.72, F(20, 134) =
3.79, p<.001, η
2p
= .36. According to Cohen’s indications [28], the effect size was large (η
2p
>
.14). Separate univariate ANOVAs revealed significant effects of condition on positive features,
skills,friendship,family and partner. No significant effects of condition were found on health,
help,leisure or on the academic/professional area, neither on intrinsic or extrinsic themes.
Post-hoc comparisons using Bonferroni adjustment revealed that, regarding personal area,
positive features were more frequent in BPRES than in BPAS and BPS, and skills appeared
more frequently in BPRES than in BPAS. Regarding social area, friendship was more frequent
Table 2. Means (M) and standard deviations (SD) for the themes of the essays per condition.
Themes BPAS
M (SD)
BPRES
M (SD)
BPS
M (SD)
TOTAL
M (SD)
ANOVA results Post-hoc comparisons
Personal area
P. features 1.52 (1.42) 3.04 (1.74) 1.77 (0.99) 2.09 (1.55) F(2, 75) = 8.50, p<.001, η
2p
= .185 BPRES >BPAS, p= .001; BPRES >BPS, p= .006
Skills 0.00 (0.00) 0.40 (0.71) 0.23 (0.65) 0.21 (0.57) F(2, 75) = 3.49, p= .036, η
2p
= .085 BPRES >BPAS, p= .027
Health 0.19 (0.40) 0.24 (0.44) 0.42 (0.50) 0.28 (0.45) F(2, 75) = 2.04, p= .137, η
2p
= .052 n.s.
Academic / professional area
Intrinsic 0.85 (0.72) 0.80 (0.87) 0.92 (0.80) 0.86 (0.79) F(2, 75) = .16, p= .857, η
2p
= .004 n.s.
Extrinsic 0.52 (0.58) 0.40 (0.65) 0.85 (0.78) 0.59 (0.69) F(2, 75) = 3.02, p= .055, η
2p
= .074 n.s.
Social area
Friendship 1.67 (1.33) 0.88 (0.67) 0.96 (0.53) 1.18 (0.98) F(2, 75) = 5.83, p= .004, η
2p
= .135 BPAS >BPRES, p= .009; BPAS >BPS, p= .020
Family 0.37 (0.56) 0.64 (0.64) 0.88 (0.65) 0.63 (0.65) F(2, 75) = 4.58, p= .013, η
2p
= .109 BPS >BPAS, p= .010
Partner 0.37 (0.69) 0.16 (0.37) 0.58 (0.58) 0.37 (0.58) F(2, 75) = 3.46, p= .036, η
2p
= .085 BPS >BPRES, p= .031
Help 0.22 (0.58) 0.20 (0.41) 0.54 (0.71) 0.32 (0.59) F(2, 75) = 2.77, p= .069 η
2p
= .069 n.s.
Leisure area
Leisure 0.56 (1.01) 0.31 (0.74) 0.65 (0.75) 0.49 (0.83) F(2, 75) = 1.74, p= .182, η
2p
= .044 n.s.
Notes: P. features = Positive features, n.s. = not significant.
https://doi.org/10.1371/journal.pone.0216896.t002
Qualitative analysis of the Best Possible Self
PLOS ONE | https://doi.org/10.1371/journal.pone.0216896 May 17, 2019 8 / 15
in BPAS than in BPRES and BPS, family was more frequent in BPS than BPAS, and partner
appeared more frequently in the texts in BPS than in BPRES.
Differences between conditions on the features of the texts
Table 3 shows the means, standard deviations, and the MANOVA results for the effect of
condition on the features of the essays. The MANOVA showed that, using Pillai’s trace, there
was a significant effect of condition on the presence of the features of the texts, V= 0.22,
F(8, 146) = 2.31, p= .023, η
2p
= .11. According to Cohen’s indications [28], the effect size was
moderate (η
2p
>.06).
Separate univariate ANOVAs and post-hoc analyses using Bonferroni adjustment revealed
that the number of sensorial details was higher in BPAS than in BPRES texts, and incongru-
ousness appeared significantly more often in BPS than in BPRES. A tendency to reach signifi-
cance on the effect of condition on the valence of the essays was found, being more positive in
BPAS than in BPS. No significant differences between conditions were found on length.
Analyses of the predictors of the change in PA: do the themes and features
of the texts predict the change in PA?
Three stepwise multiple regression analyses, one for each condition, were used to examine
which themes and features predicted change in PA. Variance Inflation Factor ranged from
1.00 to 1.01, indicating no problems with multicollinearity [30,31]. All the themes and features
were entered simultaneously. For BPAS, only emotional valence remained as a significant pre-
dictor of change in PA (β= 0.84, t= 2.84, p= .009). The model was statistically significant,
F(1,25) = 8.05, p= .009, R
2
= .24, R
2Adjusted
= .21, explaining 21% of the variance. By contrast,
for BPS, length of the essay (β= 0.02, t= 2.07, p= .050) and extrinsic theme (β= 3.71, t= 3.02,
p= .006) remained as significant predictors of change in PA. The model was statistically signif-
icant, F(1,23) = 4.29, p= .050, R
2
= .39, R
2Adjusted
= .34, explaining 34% of the variance in PA.
In the case of BPRES, none of the variables remained as significant predictors.
Parallel multiple mediation analyses: do the features of the texts mediate
the relationship between the themes of the texts and the change in PA?
Coefficients, Standard Errors (SE) and Confidence Intervals (CI) of the parallel multiple medi-
ations for the significant models can be found in Table 4.
In BPAS condition, there were significant indirect effects of friendship and partner on
change in PA through emotional valence, b= 0.76, 95% CI [0.22, 1.95] and b= 0.98, 95% CI
[0.06, 3.29], respectively (see Fig 1), as bias-corrected bootstrap 95% confidence intervals (CI)
for the indirect effects, based on 5.000 bootstrap samples, did not included zero. Neither the
total effect, b= -0.97, t = -2.02, p= .056, nor the direct effect, b= -0.55, t= -1.02, p= 0.319
Table 3. Means (M) and standard deviations (SD) for the features of the essays per condition.
Essay features BPAS
M (SD)
BPRES
M (SD)
BPS
M (SD)
TOTAL
M (SD)
ANOVA results Post-hoc comparisons
Valence 2.96 (2.14) 1.92 (1.98) 1.69 (2.00) 2.21 (2.10) F(2,75) = 2.93, p= .060, η
2p
= .072 BPAS >BPS, p= .079
1
Incongruousness 0.19 (0.62) 0.04 (0.20) 0.77 (1.61) 0.33 (1.04) F(2,75) = 3.81, p= .027, η
2p
= .092 BPS >BPRES, p= .034
Sensorial details 0.63 (1.21) 0.04 (0.20) 0.15 (0.78) 0.28 (0.88) F(2,75) = 3.54, p= .034, η
2p
= .086 BPAS >BPRES, p= .045
Length 278.78 (92.97) 252.04 (83.05 249.15 (89.85) 260.33 (88.73) F(2,75) = 0.87, p= .412, η
2p
= .023 n.s.
Notes:
1
= marginally significant, n.s. = not significant.
https://doi.org/10.1371/journal.pone.0216896.t003
Qualitative analysis of the Best Possible Self
PLOS ONE | https://doi.org/10.1371/journal.pone.0216896 May 17, 2019 9 / 15
were significant. No significant indirect effects were found for the rest of the themes and fea-
tures, as all CI included zero. Thus, results imply that, when participants in BPAS condition
wrote about the themes friendship and partner, they wrote more positive texts (i.e., with higher
emotional valence), and that produced higher changes in PA.
Regarding BPRES, no significant indirect effects were found, as all CI included zero and all
p>.05.
For BPS, there were significant indirect effects of positive features and family on change in
PA through length (i.e., number of words), b= 1.20, 95% CI [0.13, 3.95], and b= 2.83, 95% CI
[0.50, 7.64] respectively, given that bias-corrected bootstrap 95% confidence intervals (CI) for
Table 4. Coefficients, Standard Errors (SE) and Confidence Intervals (CI) of the parallel multiple mediations for the significant models.
BPAS condition BPS condition
Friendship as predictor Partner as predictor P.features as predictor Family as predictor
Coefficients (SE) 95% CI Coefficients (SE) 95% CI Coefficients (SE) 95% CI Coefficients (SE) 95% CI
Indirect effects
Total indirect effect 1.25 (0.59) [0.28, 2.78] 0.86 (0.94) [-0.96, 2.75] 0.73 (0.97) [-0.80, 3.29] 3,30 (1,68) [0.78, 7.89]
S.P. in Valence 0.76 (0.38) [0.22, 1.94] 0.98 (0.67) [0.06, 3.29] -0.14 (0.39) [-1.39, 0.36] 0.07 (0.38) [-0.37, 1.39]
S.P. in Length 0.09 (0.39) [-0.26, 1.40] -0.01 (0.19) [-0.61, 0.14] 1.20 (0.80) [0.13, 3.95] 2.83 (1.62) [0.50, 7.64]
S.P. in Sensorial details 0.40 (0.38) [-0.15, 1.23] -0.10 (0.52) [-2.13, 0.48] - - - -
S.P. in Incongruousness - - - - -0.33 (0.44) [-1.82, 0.16] 0.41 (0.62) [-0.10, 2.62]
Contrasts
Valence–Length 0.67 (0.59) [-0.18, 1.70] 0.99 (0.72) [0.06, 3.43] -1.33 (0.97) [-4.31, -0.02] -2.75 (1.73) [-7.76, -0.29]
Valence–Sensorial details 0.35 (0.59) [-0.48, 1.80] 1.07 (0.75) [0.02, 3.40] - -
Length–Sensorial details -0.31 (0.53) [-1.28, 0.76] 0.08 (0.56) [-0.80, 1.64] - -
Valence–Incongruousness - - - - 0.19 (0.62) [-0.91, 1.61] -0.33 (0.71) [-2.26, 0.57]
Length–Incongruousness - - - - 1,53 (0.84) [0.30, 3.96] 2.42 (1.76) [-0.28, 7.45]
Notes: “Incongruousness” was not included in the analyses in BPAS condition as it appeared in less than 25% of the texts in this condition, and the same procedure was
followed for “Sensorial details” in BPS condition. P. features = Positive features. S.P. = Specific change
https://doi.org/10.1371/journal.pone.0216896.t004
Fig 1. Parallel multiple mediations between content themes and change in PA throughfeatures of the texts in BPAS condition. Notes: All coefficients
represent unstandardized regression coefficients (and standard error in parenthesis). p<.05; �� p<.01; �� p<.001. PA = Positive affect.
https://doi.org/10.1371/journal.pone.0216896.g001
Qualitative analysis of the Best Possible Self
PLOS ONE | https://doi.org/10.1371/journal.pone.0216896 May 17, 2019 10 / 15
the indirect effects, based on 5.000 bootstrap samples, did not include zero (see Fig 2). Again,
neither the total effect, b= 1.38, t = 1.52, p= 0.141, nor the direct effect, b= 0.63, t= 0.71,
p= 0.488 were significant. No significant indirect effects were found for the rest of the themes
and features, as all CI included zero. These results suggest that, when participants in BPS con-
dition wrote about their positive features or family, they wrote longer texts, and that produced
higher changes in PA.
Discussion
This study showed that, despite the similar effects produced by writing about one’s best self in
the past, present or future on positive mood [13], these interventions respond to different
underlying mechanisms. The procedure of the included PPIs was identical (to write about the
best version of oneself), and the only difference between them was the time frame in which
participants had to focus on: their past, present or future. Notably, significant differences were
found in the content on the compositions depending on the condition. When writing about
their best past self (BPAS), participants more frequently included their social relationships
with their friends than the other conditions. In addition, they added more sensorial details
than the ones who wrote about their present self, which goes in line with previous studies that
suggest that recalling past events exhibit more sensorial details than imagining future events,
as the latter needs more mental work to supply these [32,33]. In the case of participants who
wrote about their best present self (BPRES), they talked more frequently about their personal
area, including their skills more often than in the past condition, and their positive features
more often than the rest of the conditions. Lastly, when participants wrote about their best
possible self in the future (BPS), their texts focused more on their familial relationships, being
their family more frequently included than in the past condition, and their partner more fre-
quently than in the present condition. In addition, they included more incongruousness in
their essays comparing with the present condition. Some of these results are in consonance
with previous studies about self-descriptions, which showed that participants’ descriptions of
their current self are more focused on oneself, followed by their past self-descriptions and least
of all their future self-descriptions, which were more socially oriented [34].
Fig 2. Parallel multiple mediations between content themes and change in PA throughfeatures of the texts in BPS condition. Notes: All coefficients
represent unstandardized regression coefficients (and standard error in parenthesis). p<.05; �� p<.01; �� p<.001. PA = Positive affect.
https://doi.org/10.1371/journal.pone.0216896.g002
Qualitative analysis of the Best Possible Self
PLOS ONE | https://doi.org/10.1371/journal.pone.0216896 May 17, 2019 11 / 15
As regards to predictions of change in PA, emotional valence arose as a significant predictor
of change in PA in the BPAS texts, whereas the length of the essay and academic or profes-
sional theme extrinsically motivated remained as significant predictors of change in PA for
participants in the BPS condition. That is, when writing about their best past self, the more
positive the compositions participants wrote, the better results on their levels of positive mood
they obtained. Conversely, when writing about their best possible self in the future, the more
words participants wrote, or the more they included the extrinsic academic or professional
theme, the more benefits they achieved on their positive mood levels. In the case of present
condition, none of the variables remained as significant predictors.
With respect to mediation analyses, significant indirect effects of friendship and partner on
PA change through emotional valence in the case of BPAS were found. For the BPS condition,
significant indirect effects of text length on PA change through positive features and family. In
other words, when participants wrote about their past self and talked about their relationship
with their friends or their partners, this led to greater positivity in their texts, which produced
improvements in their levels of positive mood. In the case of participants who wrote about
their best possible self in the future, when they focused on their own positive features or their
relationships with their family, this produced longer texts, which led to better results in their
levels of positive emotions. In the case of participants who wrote about their best current self,
no indirect effects were found.
Based on these results, it is possible to conclude that there are differences in the content and
form of the compositions of the three PPIs as well as their underlying mechanisms: even that
all of them consisted of writing about their best selves, the themes and features of their essays
were different, and the factors that predicted and mediated the change in the level of positive
mood were also different. It seems that positive emotional valence in combination with social
themes as friendship or partner play an important role in the BPAS condition, whereas the
length of the essay combined with positive features or family have an impact on the efficacy of
the BPS condition. It is worth to note, however, that the analyses did not find significant results
on the BPRES condition.
These results can have important implications. Approaching to disentangle the working
mechanisms of psychological interventions can help practitioners to use the interventions to
their most potential. In this sense, the results obtained in this specific study may help to boost
the effects of the interventions by, for example, modifying or highlighting specific features in
the instructions. Since emotional valence seems to be a key component of the BPAS condition,
it could be beneficial to encourage participants to include as many positive emotional states as
possible when they write about their best past self. However, some participants can feel frus-
trated if they are not able to naturally include positive emotional states when they are asked to
do so. In this case, and based on the results on the mediation analyses, emphasizing the social
area (writing about their friends or partner) could indirectly boost the efficacy of this PPI. Fol-
lowing the same rationale, the length of the text is an important factor in the future condition.
It is possible to encourage participants to write down as much as possible. However, it is not
feasible to know how much they should write, and it is possible that some discomfort reactions
could arise in a participant who does not accomplish to write as much as asked. In the same
manner, after mediation analyses results, asking participants to focus on their positive features
and family relationships in their texts could indirectly amplify the efficacy of the intervention.
This study has some limitations that are necessary to address. First, the sample included
was considerably young (M = 20.23, SD = 4.10). Even though this is not a limitation per se,
including a sample with more heterogeneous age could have helped us to produce more gener-
alizable results. In addition, it would be highly interesting to conduct future studies in a more
heterogeneously aged sample, in order to explore whether older participants show the same
Qualitative analysis of the Best Possible Self
PLOS ONE | https://doi.org/10.1371/journal.pone.0216896 May 17, 2019 12 / 15
pattern as the younger ones. Second, we were not able to find which mechanisms underlie in
the efficacy of writing about one’s present best self (BRES). It is possible that the term “present
self” seemed too broad to participants, which led to an excessively heterogeneous time range to
find significant results. Previous research has found that there are significant differences
between recalling near and far past events, as well as between imagining near or far future
events [33,35]. Participants writing about their best self in the present could have focused on
their present moment, but it is also possible that some of them included near past or even
future time frames, as it was not predefined in the instructions, thus different processes may
have been affecting on this condition. Future studies should explore this condition in more
detail, either encouraging participants to focus on a specific time frame or exploring which
time range they included in their texts.
This work has been the first attempt to study which are the underlying mechanisms of the
BPS intervention and the two variants derived from it, and the role that these mechanisms
have on their efficacy. There is evidence about PPIs being efficacious resources to improve
wellbeing over different populations, but the research about the mechanisms that produce
those benefits is still in its infancy [5,12,13]. Hence, there is still much more to investigate in
this regard. This study helped to shed light on the importance of the idiosyncratic features of
PPIs in order to better understand how they work. Results obtained provide a richer knowl-
edge about the process that takes part when participants practice the BPS intervention and its
temporal variants. This deeper understanding can be a powerful tool to increment their effi-
cacy by, for example, modifying the instructions of the exercises. We encourage researchers to
continue the investigations on this topic, as a better knowledge about why and how PPIs work
will help psychologists and other professionals to make the most of these valuable resources.
Acknowledgments
CIBERobn is an initiative of the ISCIII.
Author Contributions
Conceptualization: Alba Carrillo, Marian Martı
´nez-Sanchis, Rosa M. Baños.
Formal analysis: Alba Carrillo, Marian Martı
´nez-Sanchis.
Investigation: Alba Carrillo, Marian Martı
´nez-Sanchis, Ernestina Etchemendy.
Methodology: Alba Carrillo, Marian Martı
´nez-Sanchis, Rosa M. Baños.
Supervision: Ernestina Etchemendy, Rosa M. Baños.
Writing – original draft: Alba Carrillo, Marian Martı
´nez-Sanchis.
Writing – review & editing: Ernestina Etchemendy, Rosa M. Baños.
References
1. Seligman MEP, Csikszentmihalyi M. Positive Psychology: An Introduction. Am Psychol. 2000; 55: 5–
14. https://doi.org/10.1037/0003-066X.55.1.5 PMID: 11392865
2. Diener E. Subjective well-being. Psychol Bull. 1984; 95: 542–575. https://doi.org/10.1037/0033-2909.
95.3.542 PMID: 6399758
3. Mitchell J, Vella-Brodrick D, Klein B. Positive psychology and the internet: A mental health opportunity.
E-Journal Appl Psychol. Swinburne University; 2010; 6: 30–41. https://doi.org/10.7790/ejap.v6i2.230
4. Sin NL, Lyubomirsky S. Enhancing Well-Being and Alleviating Depressive Symptoms With Positive
Psychology Interventions: A Practice-Friendly Meta-Analysis. J Clin Psychol. 2009; 65: 467–487.
https://doi.org/10.1002/jclp.20593 PMID: 19301241
Qualitative analysis of the Best Possible Self
PLOS ONE | https://doi.org/10.1371/journal.pone.0216896 May 17, 2019 13 / 15
5. Bolier L, Haverman M, Westerhof GJ, Riper H, Smit F, Bohlmeijer E. Positive psychology interventions:
A meta-analysis of randomized controlled studies. BMC Public Health. BMC Public Health; 2013; 13: 1.
https://doi.org/10.1186/1471-2458-13-1
6. Lyubomirsky S, Layous K. How Do Simple Positive Activities Increase Well-Being? Curr Dir Psychol
Sci. 2013; 22: 57–62. https://doi.org/10.1177/0963721412469809
7. Quoidbach J, Mikolajczak M, Gross JJ. Positive Interventions: An Emotion Regulation Perspective.
Psychol Bull. 2015; 141: 25–27. https://doi.org/10.1037/a0038648 PMID: 25621978
8. Wang RAH, Nelson-Coffey SK, Layous K, Bao KJ, Davis OSP, Haworth CMA. Moderators of wellbeing
interventions: Why do some people respond more positively than others? PLoS One. 2017; 12: 1–19.
https://doi.org/10.1371/journal.pone.0187601 PMID: 29107994
9. King LA. The health benefits of writing about life goals. Personal Soc Psychol Bull. 2001; 27: 798–807.
10. Frattaroli J. Experimental disclosure and its moderators: A meta-analysis. Psychol Bull. 2006; 132:
823–865. https://doi.org/10.1037/0033-2909.132.6.823 PMID: 17073523
11. Pennebaker JW, Seagal JD. Forming a Story: The Health Benefits ol Narrative. J Clin Psychol. 1999;
55: 1243–1254. https://doi.org/10.1002/(SICI)1097-4679(199910)55:10<1243::AID-JCLP6>3.0.CO;2-
NPMID: 11045774
12. Loveday PM, Lovell GP, Jones CM. The Best Possible Selves Intervention: A Review of the Literature
to Evaluate Efficacy and Guide Future Research. J Happiness Stud. Springer Netherlands; 2016; 1–22.
https://doi.org/10.1007/s10902-016-9824-z
13. Carrillo A. “My best self”: Efficacy and underlying mechanisms of a Positive Psychology Intervention
[Internet]. University of Valencia. 2018. Available: http://roderic.uv.es/handle/10550/68238
14. Hefferon K, Ashfield A, Waters L, Synard J. Understanding optimal human functioning–The ‘call for
qual’ in exploring human flourishing and well-being. J Posit Psychol. Routledge; 2017; 12: 211–219.
https://doi.org/10.1080/17439760.2016.1225120
15. Plano Clark VL. Mixed methods research. J Posit Psychol. Routledge; 2017; 12: 305–306. https://doi.
org/10.1080/17439760.2016.1262619
16. Hill ED, Terrell HK, Arellano A, Schuetz B, Nagoshi CT. A Good Story: Using Future Life Narratives to
Predict Present Well-Being. J Happiness Stud. 2015; 16: 1615–1634.
17. Loveday PM, Lovell GP, Jones CM. The importance of leisure and the psychological mechanisms
involved in living a good life: A content analysis of best-possible-selves texts. J Posit Psychol. Rout-
ledge; 2017; 9760: 1–11. https://doi.org/10.1080/17439760.2017.1374441
18. Newman DB, Tay L, Diener E. Leisure and Subjective Well-Being: A Model of Psychological Mecha-
nisms as Mediating Factors. J Happiness Stud. 2014; 15: 555–578. https://doi.org/10.1007/s10902-
013-9435-x
19. Meevissen YMC, Peters ML, Alberts HJEM. Become more optimistic by imagining a best possible self:
effects of a two week intervention. J Behav Ther Exp Psychiatry. 2011; 42: 371–378. https://doi.org/10.
1016/j.jbtep.2011.02.012 PMID: 21450262
20. Sheldon KM, Lyubomirsky S. How to increase and sustain positive emotion: The effects of expressing
gratitude and visualizing best possible selves. J Posit Psychol. 2006; 1: 73–82. https://doi.org/10.1080/
17439760500510676
21. Boselie JJLM, Vancleef LMG, Peters ML. The effects of experimental pain and induced optimism on
working memory task performance. Scand J Pain. Clinical Psychological Science, Maastricht Univer-
sity, Netherlands; 2016; 12: 25–32. https://doi.org/10.1016/j.sjpain.2016.03.001 PMID: 28850487
22. Peters ML, Meevissen YMC, Hanssen MM. Specificity of the Best Possible Self intervention for increas-
ing optimism: Comparison with a gratitude intervention. Ter Psicolo
´gica. 2013; 31: 93–100. Available:
http://www.redalyc.org/articulo.oa?id=78525710009
23. Holmes EA, Coughtrey AE, Connor A. Looking at or Through Rose-Tinted Glasses? Imagery Perspec-
tive and Positive Mood. Emotion. 2008; 8: 875–879. https://doi.org/10.1037/a0013617 PMID:
19102599
24. Odou N, Vella-Brodrick DA. The Efficacy of Positive Psychology Interventions to Increase Well-Being
and the Role of Mental Imagery Ability. Soc Indic Res. 2013; 110: 111–129. https://doi.org/10.1007/
s11205-011-9919-1
25. Watson D, Clark LA, Tellegen A. Development and validation of brief measures of positive and negative
affect: the PANAS scales. J Pers Soc Psychol. 1988; 54: 1063–1070. PMID: 3397865
26. Lo
´pez-Go
´mez I, Herva
´s G, Va
´zquez C. Adaptacio
´n de las “escalas de afecto positivo y negativo”
(PANAS) en una muestra general Española. Behav Psychol Psicol Conduct. 2015; 23: 529–548.
27. Spangler PT, Liu J, Hill CE. Consensual qualitative research for simple qualitative data: An introduction
to CQR-M. In: Hill CE, editor. Consensual Qualitative Research: A Practical Resource for Investigating
Qualitative analysis of the Best Possible Self
PLOS ONE | https://doi.org/10.1371/journal.pone.0216896 May 17, 2019 14 / 15
Social Science Phenomena. 1st ed. Washington, DC: American Psychological Association; 2012. pp.
269–284.
28. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale: Erlbaum; 1988.
29. Hayes A. Introduction to mediation, moderation, and conditional process analysis: A regression-based
approach. Guilford Press; 2013.
30. Bowerman BL, O’Connell RT. Linear statistical models: An applied approach. 2nd ed. Belmont, CA:
Duxbury; 1990.
31. Myers R. Classical and modern regression with applications. 2nd ed. Boston, MA: Duxbury; 2000.
32. Grysman A, Prabhakar J, Anglin SM, Hudson JA. The time travelling self: Comparing self and other in
narratives of past and future events. Conscious Cogn. Elsevier Inc.; 2013; 22: 742–755. https://doi.org/
10.1016/j.concog.2013.04.010 PMID: 23703026
33. Arnold KM, Mcdermott KB, Szpunar KK. Imagining the near and far future: The role of location familiar-
ity. Mem Cogn. 2011; 39: 954–967. https://doi.org/10.3758/s13421-011-0076-1 PMID: 21312016
34. Shao Y, Yao X, Ceci SJ, Wang Q. Does the self drive mental time travel? Memory. 2010; 18: 855–862.
https://doi.org/10.1080/09658211.2010.514272 PMID: 20924944
35. D’Argembeau A, Van Der Linden M. Phenomenal characteristics associated with projecting oneself
back into the past and forward into the future: Influence of valence and temporal distance. Conscious
Cogn. 2004; 13: 844–858. https://doi.org/10.1016/j.concog.2004.07.007 PMID: 15522635
Qualitative analysis of the Best Possible Self
PLOS ONE | https://doi.org/10.1371/journal.pone.0216896 May 17, 2019 15 / 15
... The coding strategy was adapted from Braun and Clarke's (2006) method for thematic analysis. Thematic analysis is a relatively flexible form of qualitative analysis, and has been used in prior research to analyze data generated through similar writing interventions to better understand the lived experiences of individuals who completed the intervention (e.g., Abreu, Riggle, & Rostosky, 2019;Carrico et al., 2015;Carrillo, Martínez-Sanchis, Etchemendy, & Baños, 2019;Rawlings, Brown, Stone, & Reuber, 2017). Because identity disruption was a novel construct, our first step was to develop themes that thoroughly characterized the subjective experience of identity disruption, based on participants' reports of their identity-relevant experiences. ...
... Quantitative analyses were conducted to examine the relationship between identity disruption themes and psychosocial outcomes, including social support, PTSD symptoms, satisfaction with life, and reintegration difficulty. Such mixed-methods approaches to quantifying the associations between qualitatively coded themes and quantitatively measured outcomes of interest are regularly used in narrative identity research (e.g., Adler, 2012;McLean & Pratt, 2006;Syed & Azmitia, 2008), and have also been previously applied in the context of writing interventions (e.g., Carrillo et al., 2019). ...
Article
Full-text available
Most research and theory on identity integration focuses on adolescents and young adults under age 30, and relatively little is known about how identity adjusts to major life events later in life. The purpose of the present study was to operationalize and investigate identity disruption, or a loss of temporal identity integration following a disruptive life event, within the developmental context of established adulthood and midlife. We used a mixed-methods approach to examine identity disruption among 244 Afghanistan and Iraq war veterans with reintegration difficulty who participated in an expressive writing intervention. Participants completed measures of social support, posttraumatic stress disorder (PTSD) symptom severity, satisfaction with life, and reintegration difficulty at baseline right before writing, and 3 and 6 months after the expressive writing intervention. The expressive writing samples were coded for identity disruption using thematic analysis. We hypothesized that identity disruption would be associated with lower social support, more severe PTSD symptoms, lower satisfaction with life, and greater reintegration difficulty at baseline. Forty-nine percent (n = 121) of the sample indicated identity disruption in their writing samples. Identity disruption was associated with more severe PTSD symptoms, lower satisfaction with life, and greater reintegration difficulty at baseline, and with less improvement in social support. The findings suggest that identity disruption is a meaningful construct for extending the study of identity development to established adult and midlife populations, and for understanding veterans' adjustment to civilian life. (PsycInfo Database Record (c) 2020 APA, all rights reserved).
... For instance, the Best Possible Self intervention involves a positive thinking technique that has been shown to enhance optimism levels (Meevissen et al., 2011;Peters et al., 2010). Stemming from positive psychology research, this relatively simple intervention asks participants to write down aspects about their future best possible self (Carillo et al., 2019;Meevissen et al., 2011;Sheldon & Lyubomirsky, 2006). People are asked to write about their goals and wishes in personal domains (e.g., psychological skill), professional domains (e.g., occupational accomplishments), and relational domains (e.g., relationships with loved ones), and are also asked to start their sentences with "in the future I will" to promote realistic and positive orientation (Meevissen et al., 2011). ...
Article
Although research has examined the relationship between stress and life satisfaction for individuals with multiple sclerosis (MS), less is known about the mechanism through which disability-related stress affects life satisfaction. The purpose of the study was to examine the intermediary role of optimism and mental health in the relationship between disability-related stress and life satisfaction in people with MS. In this cross-sectional study, the sample consisted of 373 adults with MS (mean age = 47.77 years; SD = 11.70). Descriptive statistics, correlation analyses, and a serial mediation analysis were conducted in this study. Our findings suggested that disability-related stress was inversely associated with optimism, mental health, and life satisfaction. Optimism was positively associated with mental health and life satisfaction. Mental health was positively associated with life satisfaction. The serial mediation analysis results suggested that disability-related stress was negatively associated with life satisfaction through optimism and mental health. Our findings provided implications for clinicians to facilitate optimism and mental health promotion for people with MS.
... In general, there is a lack of optimism-based intervention studies in populations with and without CVD, but existing CBT interventions have shown to improve short-term optimism (83). Another possible strategy to improve optimism is the "Best Possible Self " (BPS) intervention, which is based on the positive writing paradigm, i.e., patients/clients write about the positive topics, such as focusing on their best possible future, that improves their positive mood and psychological well-being (90). Two meta-analytic studies have shown that BPS interventions could effectively improve optimism compared to controls (91,92). ...
Article
Full-text available
Background: Optimism has been shown to be positively associated with better cardiovascular health (CVH). However, there is a dearth of prospective studies showing the benefits of optimism on CVH, especially in the presence of adversities, i.e., psychosocial risks. This study examines the prospective relationship between optimism and CVH outcomes based on the Life's Simple 7 (LS7) metrics and whether multilevel psychosocial risks modify the aforementioned relationship. Methods: We examined self-reported optimism and CVH using harmonized data from two U.S. cohorts: Multi-Ethnic Study of Atherosclerosis (MESA) and Jackson Heart Study (JHS). Modified Poisson regression models were used to estimate the relationship between optimism and CVH using LS7 among MESA participants ( N = 3,520) and to examine the relationship of interest based on four biological LS7 metrics (body mass index, blood pressure, cholesterol, and blood glucose) among JHS and MESA participants ( N = 5,541). For all CVH outcomes, we assessed for effect measure modification by psychosocial risk. Results: Among MESA participants, the adjusted risk ratio (aRR) for ideal or intermediate CVH using LS7 comparing participants who reported high or medium optimism to those with the lowest level of optimism was 1.10 [95% Confidence Interval (CI): 1.04–1.16] and 1.05 (95% CI: 0.99–1.11), respectively. Among MESA and JHS participants, the corresponding aRRs for having all ideal or intermediate (vs. no poor) metrics based on the four biological LS7 metrics were 1.05 (0.98–1.12) and 1.04 (0.97–1.11), respectively. The corresponding aRRs for having lower cardiovascular risk (0–1 poor metrics) based on the four biological LS7 metrics were 1.01 (0.98–1.03) and 1.01 (0.98–1.03), respectively. There was some evidence of effect modification by neighborhood deprivation for the LS7 outcome and by chronic stress for the ideal or intermediate (no poor) metrics outcome based on the four biological LS7 metrics. Conclusion: Our findings suggest that greater optimism is positively associated with better CVH based on certain LS7 outcomes among a racially/ethnically diverse study population. This relationship may be effect measure modified by specific psychosocial risks. Optimism shows further promise as a potential area for intervention on CVH. However, additional prospective and intervention studies are needed.
... Importantly, there is a precedence in the literature for using qualitative analysis to better understand the workings not only of positive psychology interventions 49,50 but also the BPS intervention specifically. 51,52 Reflexive Thematic Analysis (TA) 53 was applied to identify, analyse and report patterns within our data set. TA was chosen as it is best suited to questions related to people's experiences, views and perceptions and the construction of meaning, which we felt was important for understanding how people engaged with the BPS intervention. ...
Article
Full-text available
Background: Public health initiatives seek to modify lifestyle behaviours associated with risk (e.g., diet, exercise, and smoking), but underpinning psychological and affective processes must also be considered to maximize success. Objective: This study aimed to qualitatively assess how participants engaged with and utilized the best possible self (BPS)-intervention specifically as a type 2 diabetes (T2D) prevention tool. Design and methods: Fourteen participants engaged with a tailored BPS intervention. Reflexive thematic analysis analysed accounts of participant's experiences and feasibility of use. Results: All participants submitted evidence of engagement with the intervention. The analysis considered two main themes: Holistic Health and Control. The analysis highlighted several nuanced ways in which individuals conceptualized their health, set goals, and received affective benefits, offering insights into how people personalized a simple intervention to meet their health needs. Conclusions: To our knowledge, this is the first study to tailor the BPS intervention as a public health application for the prevention of T2D. The intervention enabled users to identify their best possible selves in a way that encouraged T2D preventive behaviours. We propose that our tailored BPS intervention could be a flexible and brief tool to assist public health efforts in encouraging change to aid T2D prevention. Public contribution: The format, language and application of the BPS intervention were adapted in response to a public consultation group that developed a version specifically for application in this study.
Article
Full-text available
Research indicates that brief 2-min positive psychology interventions (PPIs) increase well-being during COVID-19 lockdowns. The present study extended this to assess the effectiveness over two-weeks. Participants (n = 150) were randomly allocated to one of three PPIs; nostalgia, gratitude, best possible self (BPS), or control. The interventions were slightly adapted for the lockdown and were completed three times, every seven days over two-weeks. Well-being measures were completed immediately after the first intervention (T1), after the next two interventions (T2-T3) and at one-week follow-up (T4) (but no baseline measure of well-being was taken). At T1, participants in the nostalgia, gratitude, and BPS intervention had higher self-esteem than those in the control intervention. At T1 and T2, participants in the gratitude and BPS intervention reported higher social connectedness than participants in the nostalgia and control intervention. Then at follow-up (T4), participants in the nostalgia, gratitude, and BPS intervention had lower fear of COVID-19 than those in the control intervention. Overall, the results show the benefits of nostalgia, gratitude and optimism, compared to the control, during lockdown. Supplementary information: The online version contains supplementary material available at 10.1007/s10902-022-00513-6.
Article
Background Diabetes self-management (DSM) is crucial for glycemic control among type-2 diabetic (T2D) people. Method This was a sequential exploratory mixed-method approach to assess whether a health-based coaching program designed to fit the unmet needs of T2D was the best intervention for improving DSM practice. Twenty-eight participants from different backgrounds were involved in phase 1 (Qualitative study) to explore DSM knowledge and practice, any difficulties obstructing such knowledge and practice, and the feasibility of implementing an intervention program nationwide. Sixty patients were recruited for phase 2 (Quasi-experimental study). A health-based coaching program, constructed to fit the unmet needs from phase 1 was implemented among thirty patients in an experimental group. By comparison, 30 patients in the control group received their usual care. Diabetes and DSM knowledge, DSM practice, and health outcomes were measured and compared between the two groups at baseline and after the 12th week of the intervention. Results The following problems were found: (1) a low perception of susceptibility to and severity of illness, (2) inadequate DSM knowledge and skills, (3) a lack of motivation to perform DSM practice, and (4) social exclusion and feelings of embarrassment. After the implementation of the program among the experimental group, all the variables improved relative to baseline and to the control group. Conclusion A health-based coaching program can improve DSM knowledge and practice and health outcomes. A nationwide program is recommended to promote DSM practice among Indonesian communities.
Article
Background Diabetes self-management (DSM) is crucial for glycemic control among type-2 diabetic (T2D) people. Method This was a sequential exploratory mixed-method approach to assess whether a health-based coaching program designed to fit the unmet needs of T2D was the best intervention for improving DSM practice. Twenty-eight participants from different backgrounds were involved in phase 1 (Qualitative study) to explore DSM knowledge and practice, any difficulties obstructing such knowledge and practice, and the feasibility of implementing an intervention program nationwide. Sixty patients were recruited for phase 2 (Quasi-experimental study). A health-based coaching program, constructed to fit the unmet needs from phase 1 was implemented among thirty patients in an experimental group. By comparison, 30 patients in the control group received their usual care. Diabetes and DSM knowledge, DSM practice, and health outcomes were measured and compared between the two groups at baseline and after the 12th week of the intervention. Results The following problems were found: (1) a low perception of susceptibility to and severity of illness, (2) inadequate DSM knowledge and skills, (3) a lack of motivation to perform DSM practice, and (4) social exclusion and feelings of embarrassment. After the implementation of the program among the experimental group, all the variables improved relative to baseline and to the control group. Conclusion A health-based coaching program can improve DSM knowledge and practice and health outcomes. A nationwide program is recommended to promote DSM practice among Indonesian communities.
Thesis
Full-text available
Positive Psychology Interventions are valuable resources to promote wellbeing (Bolier et al., 2013). Within this framework, the Best Possible Self (BPS) intervention seems to be a promising approach (Loveday et al., 2016). This is a positive intervention that asks individuals to imagine themselves in the best possible future (King, 2001). Although there are many individual published studies about its efficacy, it is still unknown what the overall efficacy of this intervention is. In addition, research on the mechanisms which lie beneath the efficacy of this positive activity is scarce. Therefore, the aim of this dissertation is twofold: to explore the overall efficacy of the BPS, and to analyze the role of the mechanisms that can influence its efficacy. Concretely, the role of the temporal focus will be examined. Subsequently, this dissertation has the following specific objectives: 1) To review the overall efficacy of the BPS intervention based on the existing evidence, and to explore the role of the possible moderator variables related to the intervention implementation. 2) To contribute to a more accurate measurement of SWB considering the temporal frame. 3) To design and develop two temporal variants of the original BPS (Best Past Self and Best Present Self), applied through ICTs. 4) To analyze the efficacy of the three temporal versions of BPS, applied through ICTs, to increase wellbeing. 5) To analyze the possible underlying mechanisms that lie beneath their effectiveness, through qualitative analyses of the texts. This dissertation contains five studies (organized in four papers) and two additional chapters aimed at addressing the previously mentioned objectives. Chapter 1 described a general introduction of the main topics of this dissertation, including the main characteristics and effects of PPIs and the BPS intervention. In addition, the role of possible factors that can influence the efficacy of these interventions was briefly exposed, as well as the impact that ICTs can have in the field of PPIs. Chapter 2 consists of a systematic review and meta-analysis on the efficacy of the BPS compared with controls, which include the general efficacy levels of the intervention as well as the analyses of possible moderator variables. Chapter 3 is aimed at describing the Spanish validation of a scale that measures life satisfaction along the lifespan. In addition, exploring the temporal aspects of SWB and its relationship with sociodemographic variables and the affective components of SWB. Chapter 4 includes two randomized controlled trials (Study 1 and Study 2) in which the efficacy of the temporal variations of the BPS implemented through ICTs were compared with a control condition. Chapter 5 includes a mixed method design in which a qualitative analysis of the texts included in Study 1 was carried out and combined with quantitative data about the efficacy of the intervention on positive affect. Finally, Chapter 6 presents a general discussion that includes a summary of the main conclusions outlined by the results obtained in the previous publications, as well as the limitations and future directions of this dissertation.
Article
Full-text available
Interventions rarely have a universal effect on all individuals. Reasons ranging from participant characteristics, context and fidelity of intervention completion could cause some people to respond more positively than others. Understanding these individual differences in intervention response may provide clues to the mechanisms behind the intervention, as well as inform future designs to make interventions maximally beneficial for all. Here we focus on an intervention designed to improve adolescent wellbeing, and explore potential moderators using a representative and well-powered sample. 16-year old participants (N = 932) in the Twins Wellbeing Intervention Study logged online once a week to complete control and wellbeing-enhancing activities consecutively. Throughout the study participants also provided information about a range of potential moderators of intervention response including demographics, seasonality, personality, baseline characteristics, activity fit, and effort. As expected, some individuals gained more from the intervention than others; we used multi-level modelling to test for moderation effects that could explain these individual differences. Of the 15 moderators tested, none significantly explained individual differences in intervention response in the intervention and follow-up phases. Self-reported effort and baseline positive affect had a notable effect in moderating response in the control phase, during which there was no overall improvement in wellbeing and mental health. Our results did not replicate the moderation effects that have been suggested by previous literature and future work needs to reconcile these differences. They also show that factors that have previously been shown to influence baseline wellbeing do not also influence an individual’s ability to benefit from a wellbeing intervention. Although future research should continue to explore potential moderators of intervention efficacy, our results suggest that the beneficial effect of positive activities in adolescents were universal across such factors as sex and socioeconomic status, bolstering claims of the scalability of positive activities to increase adolescent wellbeing.
Article
Full-text available
This paper explored the psychological mechanisms by which leisure enhances well-being by using sentence-by-sentence coding of the best-possible-selves text produced by 112 participants. Of the 1097 sentences, 41% were coded as leisure indicating that leisure is an important component of optimal well-being. The data showed that Australians have significantly less leisure in their daily lives than our sample desired; older and wealthier individuals placed a greater emphasis on leisure but there were no significant differences based on gender. Application of the Detachment-Recovery, Autonomy, Mastery, Meaning, Affiliation (DRAMMA) framework showed the following allocation of sentences to psychological mechanisms: Detachment-Recovery-21%, Autonomy-23%, Mastery-12%, Meaning-11% and Affiliation-33%. In their ideal future, participants imagined that they have the time and money to do what they want, particularly, to travel. We showed leisure is not solely associated with ‘having fun’; 59% of participants wanted to use their leisure time to learn, improve, or contribute to society.
Article
Full-text available
Since its inception in 2001, the best possible selves (BPS) activity has been the focus of more than 30 studies which have shown it to be a viable intervention for increasing optimism, positive affect, health and well-being. It is timely to critically review the findings from the BPS literature and suggest directions for future research. The majority of BPS studies have used an experimental methodology and have administered the BPS activity to diverse groups including students, adults, depressive individuals and suicidal inpatients. The BPS intervention can be effective when administered in-person or on-line and repeating the activity appears to enhance efficacy. Suggestions for future research include: (a) investigation of mediator variables, (b) additional outcome variables such as hope and appreciation, (c) comparative studies regarding dosage to enhance effectiveness, (d) extension of the BPS into a best-possible-other activity, (e) diversity of delivery methods, and (f) thematic content analysis of BPS text.