Emotionally extreme life experiences are more meaningful

Abstract

What makes an experience meaningful? Diverse lines of research have provided contrasting evidence; that either positive or negative events are found particularly meaningful. In this paper, we propose that the extremity of an event, rather than its valence per se, may drive meaning, and test multiple mechanisms that might explain this effect. Across three studies (including one that was pre-registered), we show for the first time a quadratic relationship between event valence and meaningfulness, such that both extremely painful and extremely pleasant events are more meaningful than milder events. Furthermore, we show that this effect is partly mediated by shared features of extreme events; their emotional intensity and tendency to induce contemplation. While extreme positive and extreme negative events differ in many important ways, this research shows that they share key characteristics (including their extremity) that lead people to find them more meaningful.

Full text

Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=rpos20
The Journal of Positive Psychology
Dedicated to furthering research and promoting good practice
ISSN: 1743-9760 (Print) 1743-9779 (Online) Journal homepage: https://www.tandfonline.com/loi/rpos20
Emotionally extreme life experiences are more
meaningful
Sean C. Murphy & Brock Bastian
To cite this article: Sean C. Murphy & Brock Bastian (2019): Emotionally extreme life experiences
are more meaningful, The Journal of Positive Psychology, DOI: 10.1080/17439760.2019.1639795
To link to this article: https://doi.org/10.1080/17439760.2019.1639795
View supplementary material
Published online: 17 Jul 2019.
Submit your article to this journal
View Crossmark data

Emotionally extreme life experiences are more meaningful
Sean C. Murphy and Brock Bastian
Melbourne School of Psychological Science, The University of Melbourne, Melbourne, Australia
ABSTRACT
What makes an experience meaningful? Diverse lines of research have provided contrasting
evidence; that either positive or negative events are found particularly meaningful. In this
paper, we propose that the extremity of an event, rather than its valence per se, may drive
meaning, and test multiple mechanisms that might explain this effect. Across three studies
(including one that was pre-registered), we show for the first time a quadratic relationship
between event valence and meaningfulness, such that both extremely painful and extremely
pleasant events are more meaningful than milder events. Furthermore, we show that this effect is
partly mediated by shared features of extreme events; their emotional intensity and tendency to
induce contemplation. While extreme positive and extreme negative events differ in many
important ways, this research shows that they share key characteristics (including their extremity)
that lead people to find them more meaningful.
ARTICLE HISTORY
Received 15 October 2018
Accepted 11 June 2019
KEYWORDS
Meaning; positive
psychology; negative affect;
positive affect; pain
What makes some experiences in life seem meaningful,
while others fade into the background? While there is
no single answer to this question, research from several
avenues of inquiry suggest valence is a key factor: The
extent to which events are positive or negative appears
to be one of the most important determinants of
whether they are found to be meaningful (Fivush,
Hazzard, McDermott Sales, Sarfati, & Brown, 2003;
King, Hicks, Krull, & Del Gaiso, 2006; Tov & Lee, 2016).
Different theories conflict, however, in predicting which
end of the valence spectrum is most strongly related to
meaningfulness, leading researchers to focus on either
positive or negative experiences alone, or to compare
one to the other. Critically, this approach has meant
that the role of experience extremity (independent of
valence) has been relatively under-explored.
Furthermore, it has led to a focus on differences rather
than similarities in how the positivity and negativity of
experiences are related to meaningfulness. In this
paper, we explore the possibility that both extremely
pleasant and extremely painful events might be espe-
cially likely to be found meaningful, relative to more
neutral events –for some of the same reasons.
Why are positive and negative events meaningful?
Narrative identity theory suggests that our life story
(and thus our sense of who we are), is not an objective
history, but rather is made up of a carefully curated
selection of key events from our lives (McAdams,
2008). In a very real sense then, the events we find
meaningful (and thus worth incorporating into this
story) determine how we see ourselves and our place
in the world, and subsequently how meaningful we find
our lives (Beike & Crone, 2012). A key question arising
from this understanding of the self is which kinds of
experiences are more likely to contribute to our life
story.
Empirical work within the positive psychology litera-
ture has largely indicated that positive events will be
found particularly meaningful. For instance, when indi-
viduals are asked to recount extremely meaningful
events from their lives, they are most likely to recount
positive events, often featuring themes of rewarding
social interactions and significant life transitions
(Baum, 1988).
In recent years, evidence in favour of this possibility
has also accumulated from studies that ask people
about their experience, activities, and sense that life is
meaningful, on a regular basis. Individuals appear to
feel more meaning in their lives on days when they
experience positive events or affect, and less meaning
when they experience negative events or affect (King
et al., 2006; Machell, Kashdan, Short, & Nezlek, 2015; Tov
& Lee, 2016).
Outside the positive psychology literature, however,
several rich theories instead predict that negatively
valenced events should be found particularly
CONTACT Sean C. Murphy seanchrismurphy@gmail.com
Supplemental data for this article can be accessed here.
THE JOURNAL OF POSITIVE PSYCHOLOGY
https://doi.org/10.1080/17439760.2019.1639795
© 2019 Informa UK Limited, trading as Taylor & Francis Group

meaningful. Theories of meaning-making, for instance,
suggest that traumatic events may shake individuals’
global understanding of the world –by challenging the
notion that life is fair, or that things happen for
a reason. Individuals must then make concerted efforts
to either alter their understanding of the event to fit
within their global meaning framework, or alter their
framework to incorporate the event (Park, 2010). While
this theory does not speak to whether meaning-making
leads these events to be found more meaningful than
others in an absolute sense (Park, 2010), expending
effort often increases the value individuals see in an
experience (Inzlicht, Shenhav, & Olivola, 2018). Evidence
suggests that individuals are highly motivated to search
for meaning in traumatic events, with many feeling that
they find it (Sommer, Baumeister, & Stillman, 1998).
The theoretical case for meaningful negative events
is further supported by narrative identity theory.
McLean, Pasupathi, and Pals (2007) propose that dis-
ruptive and challenging events make for better ele-
ments of a life story, as they allow an individual to
react in ways that are diagnostic of self-characteristics,
gain life lessons or insights, or overcome difficulties. For
instance, children from violent neighbourhoods recall
negative events more coherently and with more refer-
ence to their feelings and thoughts than they do posi-
tive events (Fivush et al., 2003).
We suggest that these two literatures may both be at
least partly right. Asking which side of the valence
spectrum is associated with greater meaningfulness
may be the wrong question; perhaps both positive
and negatively valenced events are found particularly
meaningful compared to more affectively neutral
events. A handful of studies provide evidence which
points in this direction. For instance, Baumeister, Vohs,
Aaker, and Garbinsky (2013) found that the reported
overall frequency of both positive and negative life
events was related to a sense of global meaning in
life. Other work by Bohanek, Fivush, and Walker (2005)
found that both positive and negative experiences of
high emotional intensity were especially vivid and
memorable compared to those of mild emotional
intensity.
Some authors have indeed worked to integrate the
two literatures, suggesting that both positive and nega-
tive events will be found more meaningful, but for
different reasons. For instance, positive events might
reinforce expectations of stability and order in the
world around us, while negative events force us to
confront our ideas about what is truly meaningful
(King & Hicks, 2009). Surprisingly, though, there has
been little investigation into the possibility that positive
and negative life events may be meaningful for some of
the same reasons. In this paper, we propose a more
parsimonious hypothesis; that highly positive and nega-
tive events may both be particularly meaningful in part
because of the very extremity of valence (i.e. the dis-
tance from the neutral mid-point of a valence scale)
that they share.
What makes extreme events meaningful?
Emotionally extreme events share a number of charac-
teristics that might lead them to be found more mean-
ingful, independent of their valence. Perhaps the most
obvious of these is their intensity. Emotional intensity
is second of two factors in the valence-intensity model
of emotion. It is theoretically distinct from extremity,
but more extreme emotions do tend to also be more
intense (Kuppens, Tuerlinckx, Russell, & Barrett, 2013).
For example, feelings of depression and feelings of
terror might have be equally emotionally extreme (in
terms of their distance from a neutral valence point),
but the latter is likely to be much more emotionally
intense. Experiences that are intense in this way repre-
sent emotional peaks (Fredrickson, 2000; see also
Maslow, 1964), that are particularly likely to shape the
meaning we attribute to an event, and by extension
may figure more prominently in our life story
(Fredrickson, 2000).
A related possibility is that the perceived uniqueness
of events may play a role. Bohanek et al., (2005) found
that emotionally extreme events of either valence were
felt to be more unique to the individual. Events that are
felt to be rare or not to have happened to others may
stand out in memory, and be more likely to be incor-
porated in the life story and found meaningful.
Social relatedness and strong connections to others
are also commonly thought to be key factors in driving
feelings of meaning (Heine, Proulx, & Vohs, 2006). While
positive events (such as weddings and births) are often
associated with social closeness, particularly negative
experiences may also drive such bonding. Jong,
Whitehouse, Kavanagh, and Lane (2015) found that
traumatic experiences can lead to a strong sense of
shared social identity, and Bastian, Jetten, and Ferris
(2014) found that sharing a painful experience in the
lab increased social bonding and cooperative beha-
viour. If emotionally extreme events on both sides of
the valence spectrum can lead to an increased sense of
social bonding, this may be a shared path through
which they are found meaningful.
Another such path may be through contemplation.
A substantial body of work suggests that individuals
tend to contemplate negative events for some time
after the fact (Morris & Shakespeare-Finch, 2011).
2S. C. MURPHY AND B. BASTIAN

While research on the contemplation of positive events
remains sparse, one series of studies suggests that any
strong emotions induce contemplation after the fact
(Rimé, Philippot, Boca, & Mesquita, 1992). By deepening
the web of connections between a specific occurrence
and broader elements of an individual’s life, contempla-
tion may link events to personal values, giving the
event significance within a person’s life narrative.
The current studies
Drawing on the preceding analysis, we set out to test
for the first time whether both positive and negative
experiences might be found more meaningful than
neutral experiences, for some of the same reasons.
Across three studies, we elicited reports of key life
experiences across a continuous spectrum of valence,
and measured the meaningfulness of these experi-
ences. This allowed us to simultaneously model, and
thus disentangle, the impacts of event valence and
extremity on perceived meaningfulness. We also mea-
sured event qualities that might mediate the predicted
impact of emotional extremity on meaningfulness.
Study 1a and 1b
Study 1 consisted of two iterations on the same design,
designated 1a and 1b. In this study, we prompted
participants to reflect on the most significant event in
their lives over the previous year (1a) or three months
(1b).
1
We asked this way to elicit events across the
valence spectrum that were memorable, and because
sampling events completely at random might result in
recollections for which some of our questions might
make little sense (such as driving to work).
As a sense of life having significance has been sug-
gested to be a key component in meaning (Martela &
Steger, 2016), this approach has the potential drawback
of oversampling highly meaningful events. However,
we believed that there would still be sufficient natural
variation to conduct our investigation.
2
For each experience, we measured event valence
and meaningfulness, both on continuous scales. We
also measured our proposed mediators of extremity;
emotional intensity, event contemplation, sociality,
and uniqueness. In addition, we measured the extent
to which each event was felt to reflect personal growth
and a sense that it was part of an individual’s fate (or
had a ‘higher purpose’). We did not have any specific
predictions about these two factors, but included them
as they had previously been theorised as key to gen-
erating meaning in life events (Sommer et al., 1998).
Methods
Across all studies, we report all manipulations and data
exclusions. In studies 1a, 1b, and 2, participants completed
a number of measures not reported here, as part of
a separate project investigating the relationship between
trait rumination and meaning. In Study 3, which was pre-
registered, we report all measures collected.
Participants
For Study 1a, we recruited 253 participants through
Amazon Mechanical Turk (MTurk) for a study described
as being about significant life experiences. Participants
were paid $1.50 USD. Three participants were removed
for failing two or more of our three attention check
questions, one because their experience was copied
from online sources, and three for writing less than 20
words. This left 246 participants (114 female, 132
male, M = 35.38) for the analyses. For Study 1b, the
same procedures were followed, except participants
were paid $1.80. Here, 226 participants were recruited,
five were removed for failing attention checks, and one
for copying their experience from the internet, leaving
220 participants (99 female, 121 male, M = 35.06) for
the analyses.
Materials
Experience
Participants were asked to think back on their most
significant experience over the past year (three months
in 1b), and to describe the experience in a paragraph or
so, including what made it significant to them.
Meaning
Participants were next asked to rate how meaningful
the experience was, from 0 (‘a meaningless experience’)
to 10 (‘the most meaningful experience you can think
of’, or, in Study 1b, ‘the most meaningful experience
you can imagine anyone having.’).
Valence
Participants were separately asked to rate both the
extent to which the event was pleasant, and painful,
each measured from 1 (‘Not at all’)to7(‘Very’). To
prompt answers about the experience of the event
rather than retrospective feelings, participants sepa-
rately made these ratings for: ‘the events leading up
to the experience’,‘the actual moment or moments of
experience’, and ‘the events that followed after the
experience.’We used ratings of pleasantness and pain-
fulness during the event as valence measures.
THE JOURNAL OF POSITIVE PSYCHOLOGY 3

Emotional intensity
This was a single item ‘This experience was emotionally
intense’, rated from 1 (‘Completely disagree’)to7
(‘Completely agree’).
Social connection
Two items measured social connection; ‘To what extent
was this experience shared with others?’and ‘To what
extent did this experience make you feel closer to
others?’from 1 (‘Not at all’)to7(‘Very much’;α= .82
and α= .75).
Contemplation
This scale consisted of three items; ‘Ifind myself going
over this experience in my mind repeatedly’,‘Ifind
myself analyzing this experience to try to make sense
of it’, and ‘I am still trying to understand this experi-
ence’from 1 (‘Completely disagree’)to7(‘Completely
agree’;α= .77 and α= .79).
Uniqueness
This was a single item ‘To what extent was this experi-
ence unique to you (that is, very few other people
would have had a similar experience)’from 1 (‘Not at
all’)to7(‘Very much’).
Fate
Two items measured the belief that the event hap-
pened for a reason. ‘This experience happened for
a reason’, and ‘This experience was part of my fate/
destiny’, from 1 (‘Completely disagree’)to7
(‘Completely agree’;α= .75 and α= .66).
Personal growth
Personal growth was measured with three items: ‘This
experience made me a better person.’,‘This experience
shaped me as a person’and ‘This experience made me
the person I am today’from 1 (‘Completely disagree’)to
7(‘Completely agree’;α= .87 and α= .87).
Results
Wherever relevant, results from Study 1a and 1b are
reported together and in order. We originally intended to
use pleasantness and painfulness as separate indicators of
valence because positive and negatively valenced ratings
are often relatively independent (Watson, Clark, & Tellegen,
1988). In this case, however, the two ratings were very
highly correlated (r(242) = −.79, p< .001 and r(218) = −.77,
p< .001), suggesting they were best modelled as opposite
ends of a bipolar scale (Russell & Carroll, 1999). We thus
reverse-coded pleasantness, then averaged it with painful-
ness to create a single score measuring event valence.
Means, standard deviations, and zero-order correlations
are reported in Tables S1 and S2 (see Supplementary
Material).
We first tested for a main effect of valence on mean-
ingfulness. Event valence was not associated with rat-
ings of meaning in either sample. We next tested our
hypothesis that valence extremity was related to mean-
ingfulness. To do so, we fit a regression model including
both the linear and quadratic effects of event valence in
predicting meaning, centering at the mid-point of the
valence scale (4). While the linear relationship between
valence and meaning remained non-significant, the
quadratic terms were significant (β= .23, p< .001,
and β= .16, p = .020). This indicated that while painful
and pleasant experiences did not differ in meaningful-
ness overall, experiences at the ends of the valence
spectrum were indeed found more meaningful than
milder events (see Figure 1).
Mediations
We next explored what factors might mediate the effect
of valence extremity on meaningfulness. To do so, we
tested for quadratic mediation; effectively, whether each
factor was higher at extreme levels of valence, and also
positively related to meaningfulness (or vice versa).
The most consistent and strongest mediator was
emotional intensity. While this was higher for more
painful events across both studies (β= .33, p< .001
&β= .45, p< .001), there was also a significant
quadratic effect such that it was higher at more
extreme levels of valence (β= .22, p= .002 &
β= .13, p= .046), see Figure 2. Bootstrapped media-
tion tests (all mediations through the paper use 1000
bootstraps) revealed a significant indirect effect in the
first sample and marginal indirect effect in the second
(IE = .09, [.04,.15], p= .002 & IE = .05, [.01,.10],
p = .061). This indicated that particularly painful and
particularly pleasant events may both be meaningful,
in part, because of their shared tendency to be experi-
enced as emotionally intense.
Event sociality and contemplation both showed
similar (though weaker) patterns of mediation.
Sociality was lower for more painful events in both
studies (β=−.27, p< .001 & β=−.32, p< .001), but
was also higher at the extreme levels of valence
(though only significantly in Study 1b; β= .10,
p=.124 & β= .21, p= .002). This led to positive
indirect effects (again only significant in Study 1b;
IE = .03, [−.01,.06], p= .147, IE = .05, [.01,.11],
p= .015). Event contemplation was also higher after
more painful events overall (β= .35, p< .001 &
β= .45, p<.001), but showed a quadratic effect
4S. C. MURPHY AND B. BASTIAN

such that it was highest for events of extreme valence
(though only significant in Study 1b; β= .06, p= .370
&β= .15, p= .021, IE = .01, [−.02,.04], p= .389,
IE = .06, [.00,.12], p= .032). Thus, there was initial (if
mixed) evidence that extremely valenced events at
both ends of the spectrum may have been found
more social and contemplated more than neutral
events, which in part may have explained their shared
tendency to be found more meaningful. Event unique-
ness, personal growth, and fate were not related to
valence extremity, and so did not emerge as potential
mediators.
Post-hoc examination of the contemplation scale items
led us to believe that two of the three items (‘Iamstill
trying to understand this experience’and ‘Ifind myself
analysing this experience to try to make sense of it’)
indexed a lack of understanding of the event, rather than
contemplation itself. When examining the single remaining
item (‘Ifind myself going over this experience in my mind
repeatedly’), both samples exhibited a quadratic
Figure 1. Fitted quadratic curves showing predicted levels of meaningfulness as a function of valence in Studies 1a and 1b.
Figure 2. Fitted quadratic curves showing predicted levels of intensity, contemplation (single item), and sociality as a function of
valence in Studies 1a and 1b.
THE JOURNAL OF POSITIVE PSYCHOLOGY 5

relationship, (β= .15, p = .023 & .21, p = .002), which led to
significant indirect effects (IE = .05, [.01,.09] p = .037,
IE = .08, [.02,.13], p = .007). We thus decided to measure
this aspect of contemplation specifically in later studies.
Discussion
This Study provided an important first step in devel-
oping a more nuanced understanding of the role of
valence in the meaning of events. While several stu-
dies have contrasted the manner and extent to which
positive and negative events become meaningful, this
is the first evidence that strongly valenced events at
either end of the spectrum may share common fea-
tures that make them increasingly meaningful. This is
also the first time that the curvilinear relationship
between valence and event meaningfulness has
been clearly revealed, showing that the data appears
to fit a clear U-shaped curve, with more neutral
events occupying a ‘valley’of meaning.
We also found initial evidence that three of our
theoretically-derived factors were plausible media-
tors of the effect of valence extremity on meaning.
While these findings should be interpreted with cau-
tion, as some relationships were only significant in
one of the two studies, they are exciting. Aspects of
these relationships have previously been reported,
but this is the firsttimethatapotentiallinkhas
been drawn showing that sociality, contemplation
and the emotional intensity of events may all be
factors shared by extreme positive and negative
events, and key reasons why both are particularly
meaningful.Inaddition,themaineffects in our stu-
dies provided the firstdirectevidencethatevents
felt to be unique, emotionally intense, and that
received substantial contemplation were found to
be more meaningful.
This collection of initial findings provides a rich pic-
ture of what it takes to create a meaningful event.
However, given that the strength of evidence for our
mediated pathways varied across Studies 1a and 1b, we
conducted a replication in Study 2 to establish whether
these patterns of results were reliable.
Study 2
To conduct this replication, we added a second item
to measure contemplation, excluding the two items
indexing a lack of understanding. To make room for
new trait measures introduced in service of our sepa-
rate rumination project, social connectedness and
uniqueness were dropped from this study; we return
to social connectedness in Study 3.
Methods
The methods for this Study were identical to Study
1b, with one exception. To tap contemplation more
directly, we added the item ‘I thought about this event
alot’to the contemplation scale. Factor analysis of
the resulting four items indicated that they formed
two separate scales; a two-factor model fitsubstan-
tially better than a single-factor model (χ
2
(1) = 117.27, df = 1, p < .001). We use this new two-
item scale as ‘contemplation’for the remainder of the
paper (α= .71).
We recruited 406 MTurk participants, each paid $1.80
USD. Six were removed for failing two attention checks,
and five for writing under 20 words, leaving 395 (185
female, 210 male, M = 36.08). We once again measured
painfulness, pleasantness, meaning, emotional inten-
sity, fate (α= .63), and growth (α= .84). As in the
previous study, we collapse pleasantness and painful-
ness (r(390) = .75, p < .001) into a single measure of
valence. Means, standard deviations, and zero-order
correlations are reported in Table S3.
Results
We replicated the association between event valence
and meaning found in both Study 1 samples, with no
linear relationship with valence (β=−.02, p= .574), but
a significant quadratic relationship (β= .19, p < .001)
indicating that more extremely valenced events were
found more meaningful. We next moved on to testing
whether other event factors could mediate these asso-
ciations with valence extremity.
Once again, while fate, growth, contemplation and
intensity were all positively associated with meaning,
only contemplation and emotional intensity emerged as
plausible mediators of valence. The more painful an event
was, the more it was contemplated and found emotionally
intense overall (β=.17,p< .001 and β= .36, p< .001,
respectively), but quadratic relationships indicated that
both were also highest for events at extremes of the
valence spectrum (β= .18, p< .001 and β= .18, p< .001,
respectively), while these relationships were null for fate
and growth. This led to significant indirect effects of
valence extremity on meaningfulness through only con-
templation (IE = .06, [.03,.10], p= .002) and intensity
(IE = .07, [.03,.10], p= .001) see Figure 3.
6S. C. MURPHY AND B. BASTIAN

Discussion
The results from Study 2 confirmedtheroleplayedbyboth
emotional intensity and contemplation in driving the curvi-
linear association between event valence and meaning.
Events on either end of the valence spectrum were experi-
enced as more emotionally intense and contemplated more,
and this in part explained their increased meaningfulness.
The clearer pattern of results with the new contemplation
measure suggests this pattern is characteristic of contempla-
tion itself, and not rumination caused by a lack of under-
standing of the event. This makes sense; work on post-
traumatic growth suggests that a search for understanding
may lead painful events to be found more meaningful
(Michael & Snyder, 2005),butthisisunlikelytobeashared
cause of meaning across pleasant and painful events. We
next extended our work into a within-subjects design that
allowed us to replicate these findings with more control.
Study 3
In the first two studies, we asked individuals to report
significant events in their lives. This allowed us to study
life events that varied naturally across both valence and
our other factors, and were representative of the events
people found significant. However, this design came with
potential drawbacks; primarily, the potential for selective
reporting of life events.
People who report a painful event as their most sig-
nificant may differ systematically from those who choose
a pleasant one. If these differences included trait levels of
meaning, for instance, these person-level associations
might have masqueraded as event-level associations.
An additional concern was that, in the first two studies,
approximately two-thirds of reported events fell on the
‘pleasant’half of the valence scale. Thus, while the
reported painful events were just as meaningful as the
pleasant ones, they may not have been representative of
the average painful event.
To address these issues, we moved to a repeated-
measures design in which each participant reported both
a highly positive and highly negative event. This allowed us
to compare experiences within the same people, removing
person-level confounding variables. It also made sure we
were characterizing differences between events of similar
frequencies, rather than comparing rarer significant painful
events to more frequent pleasant events. Our goal was to
repeat our analyses from Study 2 under these improved
conditions. We also re-incorporated an improved measure
of event-related social bonding.
Pre-registration
We originally ran a slightly different version of Study 3
(which is now Study 3S in supplementary material) using
largely the same outcome measures as in Study 2. After
receiving editorial feedback concerned with the single-item
measures of meaningfulness and emotional intensity, we
re-ran the study with new and improved measures of these
constructs. This Study was pre-registered to solidify the
confirmatory nature of the hypotheses. The pre-
registration, including all measures and full analysis plan,
can be found at https://osf.io/94ex6/.
Methods
Participants
We recruited 521 participants through MTurk for a study
described as involving writing about their most pleasant
Figure 3. Fitted quadratic curves showing predicted levels of intensity and contemplation as a function of valence in Study 2.
THE JOURNAL OF POSITIVE PSYCHOLOGY 7

and painful life experiences in the last year. Participants
were paid $2.50 USD. We then removed responses in accor-
dance with our pre-registered plan. Thirteen responses
were removed as they did not contain codes to indicate
they were generated from our Mturk link, and contained
experiences clearly incoherent or copied from the internet.
Six participants were removed for failing both attention
checks, and thirty because at least one experience was
shorter than 20 words. This left 472 participants (228
female, 240 male, 4 undisclosed, M = 35.53).
Materials and procedure
Experience
Participants responded to two experience prompts in
counter-balanced order. In one, they were asked to report
on their ‘most rewarding, enjoyable, or pleasant’experi-
ence in the past year.
3
In the other, their ‘most adverse,
challenging, or painful’experience. After describing each
experience, participants completed our experience-
related questions from Study 2.
Meaning
In order to increase reliability, we turned our outcome mea-
sure of meaning into a three-item scale. This now consisted
of asking participants ‘How meaningful was the experience’,
‘To what extent was this experience an important moment
in your life’,and‘How significant was the experience’,on
ascalefrom0‘Notatall’to 10 ‘Extremely’;α= .87.
Valence
While we now manipulated the valence of reported events,
we still collected subjective ratings, as we intended to use
the naturally occurring variation to examine the curvilinear
associations between valence and other variables. Analysis
of our earlier studies indicated that ratings of valence
before, during, and after events were highly similar. Thus,
to reduce burden on participants, they were now asked
only to answer to what extent the experience was Pleasant
andtowhatextentitwasPainfulonscalesfrom1(‘Not at
all’)to7(‘Very’). These two measures were highly related
(β=−.88, p < .001) and were collapsed into a single valence
measure, coded so higher scores were more painful.
Social
To measure how much events facilitated social closeness,
we used an altered version of the scale from Study 1. We
wanted to tap participants’feelings of social connection,
rather than event characteristics. Thus, we removed the
question about how much the experience was shared
with others and instead added ‘This experience made me
feel more connected to others’and ‘This experience made
me feel bonded with others’, measured on a scale from 1
(‘Completely Disagree’)to7(‘Completely Agree’;α= .95).
Emotional intensity
To increase the reliability of our emotional intensity
measure, we expanded it to three items. We added
‘This experience was very emotional’and ‘this experi-
ence made me feel strong emotions’(α= .87).
We also once again measured fate (α= .69) and growth
(α= .84), as in Study 2. Means, standard deviations, and
zero-order correlations are reported in Table S5.
Results
We first conducted a manipulation check, with paired
t-tests confirming that experiences in the negative valence
condition were rated as significantly more painful (M = 6.02,
SD = 1.28) than in the positive valence condition (M = 1.8,
SD = 1.21, t(471) = 52.02, β= 1.72, p< .001).
Curvilinear effects
While our manipulation ensured that each participant
recorded both a positive and negatively valenced event,
there was substantial variation within conditions, with
responses given across the entire valence spectrum.
4
Thus, we used our continuous measure of reported valence
to replicate previous curvilinear associations and to once
again examine mediating pathways.
As we needed to model the data accounting for the
dependencies induced by our repeated-measures design,
we ran a series of mixed-effects regression models using
the lme4 package in the R statistical software (R Core Team,
2017). Each of these models included a random intercept for
participant.
First, we examined the direct relationship between
valence and meaningfulness. Contrary to previous studies,
reported painfulness was associated with lower levels of
event meaning on average (β=−.14, p< .001). However, we
once again found a strong curvilinear effect such that
extreme levels of valence were associated with higher levels
of meaning (β= .20, p< .001).
As in previous studies, we found that reported valence
curvilinearly predicted both emotional intensity and con-
templation, but not fate and growth. Emotional intensity
was associated with increased painfulness, while contempla-
tion was not, (β= .27, p<.001andβ=.03,p= .249,
respectively), however they each increased as events
approached the extremes of the valence spectrum (β=.21,
p<.001andβ=.12,p< .001, respectively). Clarifying the
mixed pattern of results from Study 1, reported valence was
not curvilinearly associated with social closeness; there was
a strong tendency for the reported pleasantness of events to
8S. C. MURPHY AND B. BASTIAN

be associated with social closeness (β=−.50, p < .001), but
social closeness neither increased nor decreased as events
becamemoreextremeinvalence(β=−.04, p = .175; See
Figure 4).
We used the mediation package in R along with the
lme4 package to test mediation while accounting for
the repeated measurement of participants. Results indi-
cated that contemplation (IE = .06, [.02,.09], p < .001),
and emotional intensity (IE = .12, [.08,.16], p < .001)
were significant mediators of the relationship between
valence and meaning, while social closeness was not
(IE = −.01, [−.03,.01], p = .200).
Linear effects
In previous studies, we focused on curvilinear relationships
between valence and meaning, as these were key to
testing our theoretical questions. However, the design of
this study also allows us to clearly characterise the linear
effects of our valence manipulation. We include these in
Figure 5 for completeness, but also to illustrate what can be
missed with this approach. When treating valence as bin-
ary, we see only linear effects, such as painful events being
characterised as more emotionally intense, or contempla-
tion apparently not varying as a function of valence.
5
The
curvilinear results reported above, by contrast, show far
more nuance in the effects of valence on these variables.
Discussion
Study 3 advanced upon our earlier efforts, as the use of
a within-subjects design eliminated several possible
alternative explanations for our results. We also made
several key measures more robust using highly reliable
Figure 4. Fitted quadratic curves showing outcomes as a function of valence in Study 3.
Figure 5. Mean differences in event characteristics across valence conditions in Study 3, with 95% confidence intervals.
THE JOURNAL OF POSITIVE PSYCHOLOGY 9

scales in place of single items, and pre-registered all
analyses to establish these results as confirmatory. We
thus confirmed, in this experimentally controlled and
balanced replication of earlier studies, that events char-
acterised by more extreme valence were found more
meaningful. We also confirmed that this was, in part,
because these events were more likely to be charac-
terised by high levels of emotional intensity and inspire
greater contemplation. Clarifying mixed findings from
earlier studies, this study indicated that there was no
detectable curvilinear effect of valence on social
bonding.
Events in the positive condition were found more
meaningful overall in this Study, where there had been
no difference previously. It seems plausible that this
deviation was due to the requirement that each parti-
cipant recall both types of experience. In previous stu-
dies, two-thirds of reported experiences were more
pleasant than painful. This suggests that significant
(and thus meaningful) painful events may be experi-
enced more rarely than significant pleasant events, and
thus in this balanced sample, a higher proportion of
non-significant painful events were reported. However,
it is encouraging to note that even in this balanced
sample, all other results replicated as expected.
General discussion
Across three studies, we used a novel design to model
curvilinear relationships between valence and meaning,
thus disentangling the directional effects of valence
from those of emotional extremity. In doing so, we
uncovered relationships that had gone undetected pre-
viously. Our findings provide strong support for the
hypothesis that emotional extremity of an event plays
a key role in whether events are found meaningful, as
the most meaningful events in our studies were those
that were either extremely pleasant or extremely pain-
ful. Importantly, our results also suggest that research
focusing only on the effects of valence may have con-
flated these with effects of extremity.
This work thus ties together separate literatures
which have previously focused largely on either positive
or negative events exclusively. There are undoubtedly
processes at work when individuals develop an under-
standing of meaning after traumatic events that are not
present when meaning is garnered from joyful and life-
affirming events, and vice versa. While we agree that
insight into these unique processes is important, if
studied in isolation, researchers risk creating divergent
literatures that miss important commonalities.
Indeed, we found that these commonalities may
reveal a more complete and nuanced picture about
what determines the events we find meaningful and
memorable. Extreme events were found more mean-
ingful, in part, because of their emotionally intensity, as
well as the contemplation they inspire. Thus, while
Figure 5 shows that highly positive and highly negative
events differ in a great many ways on average, our
models revealed commonalities that help explain why
both types of events are found especially meaningful.
Some of these commonalities are quite surprising.
We consistently found that positive and negative
events inspired contemplation to roughly the same
degree. This stands in stark contrast to the strong
focus on negative events in most previous research on
contemplation, and adds to a recent trend in seeking to
also understand how people contemplate positive
events (Feldman, Joormann, & Johnson, 2008).
Our findings also point to the importance of inten-
sity (or arousal) in building meaning in life. Alongside
valence, intensity is a key dimension relied on in map-
ping the emotional space (Kuppens et al., 2013) and
extremely positive or negative events also tend to be
experienced as more intense –our findings show that
this intensity is an important mediator of the effect of
valence extremity on meaning. This serves as
a counterpoint to the current trend of viewing low-
arousal positive experiences as most conducive to well-
being (e.g. Van Dam et al., 2018). Though mindfulness,
meditation, and feelings like calmness may be benefi-
cial across several domains, they may also be less likely
to create experiences that contribute to a person’s life
story. Rather, intense or peak experiences may be more
likely to define who we are, providing some insight into
why people at times behave in counter-hedonic ways,
seeking out objectively unpleasant experiences (e.g.
Bastian, 2018; Rozin, Guillot, Fincher, Rozin, &
Tsukayama, 2013).
Contrasts with past research
To our knowledge, only one other study has tested for
nonlinear relationships between valence and meaning.
Tov and Lee (2016) studied the relationship between
daily positive and negative affect and a sense that life is
meaningful, and found little evidence of a quadratic
relationship. Instead, their results suggested simply
that individuals felt more meaning on days when they
felt more positive.
The key difference between the two investigations
appears to be time-scale. Tov and Lee (2016) modelled
the relationship between valence and a global sense of
meaning on a given day, while we investigated discrete
events in the past. This suggests that positive events
may be found meaningful immediately, while negative
10 S. C. MURPHY AND B. BASTIAN

events ‘catch up’over time, eventually creating the
u-shaped curve revealed in our studies. This gels with
theoretical suggestions that meaning is more likely to
be constructed effortfully over time in the case of
negative events (King & Hicks, 2009), but does not
explain why contemplation did not then differ by
valence in our studies. Future research might examine
whether the valence of an event determines what func-
tion contemplating that event serves. It may be that
positive events elicit something closer to savouring,
while in negative events contemplation may be driven
by efforts to accommodate the event into one’s life
narrative.
This disparity leads into a broader question of how
the experience of meaningfulness differs across tem-
poral levels. We know very little about how the experi-
ence of meaningfulness in discrete, memorable events
resembles that of daily events, or how this meaning,
once generated, propagates to the global level. While
we have posited that a life filled with meaningful events
will lead to a richer life story and thus a greater sense of
meaning in life, empirical tests of this idea have yet to
be conducted. The study of meaning is growing at all
levels, but how these different ‘layers’relate remains
largely unexplored.
Another area of interest is the ‘live’process by which
individuals come to find events meaningful. One limita-
tion of our study was that all reports were retrospective;
individuals had already established the events we asked
about as meaningful, and we could not observe the
mental processes which drove this. As this design can-
not establish causality, it leaves open the possibility
that reciprocal processes influence our results; emo-
tional intensity may drive meaning in events, but mean-
ingful events may also come to be seen as more intense
over time, for instance. While we do not feel this would
diminish the importance of the link between the two,
establishing directionality more firmly would further
improve our understanding of how valence and extre-
mity shape, and are perhaps shaped by, meaning. To
answer this question would be costly (requiring long-
itudinal investigations that probe individuals shortly
after significant events, and at subsequent intervals),
but highly valuable.
A further limitation of our work is that it relied
only on American MTurk workers. While our sample
was diverse in age and gender, it is unclear whether
our results would generalise to other cultures that
may place different value on the intensity of experi-
ences. To the extent that what seems meaningful
can be strongly influenced by culture (Baumeister
et al., 2013), this is an important area for future
research.
Conclusions
In this paper, we found that both extremely pleasant and
extremely painful events are more meaningful than neu-
tral events, for some of the same reasons. These findings
suggest that pleasant and painful events may have more
in common when it comes to meaning than previously
thought, and that we may need to rethink how we study
the link between valence and meaning in the future.
When it comes to the search for meaning, the most
extreme experiences may be the best places to look.
Notes
1. To counteract possible ceiling effects, we changed from
a year to three months in Study 1b and adjusted the
meaningfulness question.
2. This was the case; see Figure S3 for histograms of
meaning variation across all Studies.
3. We asked about the previous year because of previous
participants wished to talk about older events; pilot test-
ing indicated that asking about valenced events led to
less skewed meaning ratings than significant events.
4. At least 20 experiences were reported at every value of
valence; see Figure S4 for histograms. For wordclouds
of what was written, see Figures S5 and S6.
5. See Table S6 for effect sizes.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
This work was supported by an Australian Research Council
Discovery Project grant awarded to Brock Bastian, grant num-
ber DP140103716.
ORCID
Sean C. Murphy http://orcid.org/0000-0002-6783-9906
References
Bastian, B. (2018). The other side of happiness: Embracing
a more fearless approach to living. London: Penguin UK.
Bastian, B., Jetten, J., & Ferris, L. J. (2014). Pain as social glue:
Shared pain increases cooperation. Psychological Science,25
(11), 2079–2085.
Baum, S. K. (1988). Meaningful life experiences for elderly
persons. Psychological Reports,63(2), 427–433.
Baumeister, R. F., Vohs, K. D., Aaker, J. L., & Garbinsky, E. N.
(2013). Some key differences between a happy life and
a meaningful life. The Journal of Positive Psychology,8(6),
505–516.
Beike, D. R., & Crone, T. S. (2012). Autobiographical memory
and personal meaning: Stable versus flexible meanings of
THE JOURNAL OF POSITIVE PSYCHOLOGY 11

remembered life experiences. In Paul T. P. Wong (Ed.), The
human quest for meaning: Theories, research, and applica-
tions (pp. 315–334). New York: Routledge/Taylor and
Francis Group.
Bohanek, J. G., Fivush, R., & Walker, E. (2005). Memories of
positive and negative emotional events. Applied Cognitive
Psychology,19(1), 51–66.
Feldman, G. C., Joormann, J., & Johnson, S. L. (2008).
Responses to positive affect: A self-report measure of rumi-
nation and dampening. Cognitive Therapy and Research,32
(4), 507–525.
Fivush, R., Hazzard, A., McDermott Sales, J., Sarfati, D., &
Brown, T. (2003). Creating coherence out of chaos?
Children’s narratives of emotionally positive and negative
events. Applied Cognitive Psychology,17(1), 1–19.
Fredrickson, B. L. (2000). Extracting meaning from past affec-
tive experiences: The importance of peaks, ends, and spe-
cific emotions. Cognition and Emotion,14(4), 577–606.
Heine, S. J., Proulx, T., & Vohs, K. D. (2006). The meaning
maintenance model: On the coherence of social
motivations. Personality and Social Psychology Review,10
(2), 88–110.
Inzlicht, M., Shenhav, A., & Olivola, C. Y. (2018). The effort
paradox: Effort is both costly and valued. Trends in
Cognitive Sciences,22(4), 337–349.
Jong, J., Whitehouse, H., Kavanagh, C., & Lane, J. (2015).
Shared negative experiences lead to identity fusion via
personal reflection. PloS One,10(12), e0145611.
King, L. A., & Hicks, J. A. (2009). Detecting and constructing
meaning in life events. The Journal of Positive Psychology,4
(5), 317–330.
King, L. A., Hicks, J. A., Krull, J. L., & Del Gaiso, A. K. (2006).
Positive affect and the experience of meaning in life.
Journal of Personality and Social Psychology,90(1), 179–196.
Kuppens, P., Tuerlinckx, F., Russell, J. A., & Barrett, L. F. (2013).
The relation between valence and arousal in subjective
experience. Psychological Bulletin,139(4), 917–940.
Machell, K. A., Kashdan, T. B., Short, J. L., & Nezlek, J. B. (2015).
Relationships between meaning in life, social and achieve-
ment events, and positive and negative affect in daily life.
Journal of Personality,83(3), 287–298.
Martela, F., & Steger, M. F. (2016). The three meanings of
meaning in life: Distinguishing coherence, purpose, and
significance. The Journal of Positive Psychology,11(5),
531–545.
Maslow, A. H. (1964). Religions, values, and peak-experiences
(Vol. 35). Columbus, OH: Ohio State University Press.
McAdams, D. P. (2008). Personal narratives and the life story.
In O. John, R. Robins, and L. Pervin (Eds.), Handbook of
personality: Theory and research (3rd ed., pp. 241–261).
New York: Guilford Press.
McLean, K. C., Pasupathi, M., & Pals, J. L. (2007). Selves creating
stories creating selves: A process model of
self-development. Personality and Social Psychology
Review,11(3), 262–278.
Michael,S.T.,&Snyder,C.R.(2005). Getting unstuck: The roles of
hope, finding meaning, and rumination in the adjustment to
bereavement among college students. Death Studies,29(5),
435–458.
Morris, B. A., & Shakespeare-Finch, J. (2011). Rumination,
post-traumatic growth, and distress: Structural equation mod-
elling with cancer survivors. Psycho-Oncology,20(11),
1176–1183.
P
ark,C.L.(2010). Making sense of the meaning literature: An
integrative review of meaning making and its effects on adjust-
ment to stressful life events. Psychological Bulletin,136(2),
257–301.
R Core Team. (2017). R: A language and environment for sta-
tistical computing. Vienna, Austria: R Foundation for
Statistical Computing. Retrieved from https://www.
R-project.org/
Rimé,B.,Philippot,P.,Boca,S.,&Mesquita,B.(1992). Long-lasting
cognitive and social consequences of emotion: Social sharing
and rumination. European Review of Social Psychology,3(1),
225–258.
Rozin,P.,Guillot,L.,Fincher,K.,Rozin,A.,&Tsukayama,E.(2013).
Glad to be sad, and other examples of benign masochism.
Judgment and Decision Making,8(4), 439.
Russell, J. A., & Carroll, J. M. (1999). On the bipolarity of positive and
negative affect. Psychological Bulletin,125(1), 3–30.
Sommer, K. L., & Baumeister, R. F. (1998). The construction of
meaning from life events: Empirical studies of personal
narratives. In P. T. P. Wong & P. S. Fry (Eds.), The human
quest for meaning: A handbook of psychological research
and clinical applications (pp. 143–161). Mahwah, NJ:
Lawrence Erlbaum Associates Publishers.
Tov, W., & Lee, H. W. (2016). A closer look at the hedonics of
everyday meaning and satisfaction. Journal of Personality
and Social Psychology,111(4), 585–609.
VanDam,N.T.,vanVugt,M.K.,Vago,D.R.,Schmalzl,L.,Saron,C.D.,
Olendzki,A.,...Fox,K.C.(2018). Mind the hype: A critical
evaluation and prescriptive agenda for research on mindfulness
and meditation. Perspectives on Psychological Science,13(1),
36–61.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and
validation of brief measures of positive and negative affect: The
PANAS scales. Journal of Personality and Social Psychology,54(6),
1063–1070.
12 S. C. MURPHY AND B. BASTIAN