The Grateful Disposition: A Conceptual and Empirical Topography
Michael E. McCullough
Southern Methodist University
Robert A. Emmons
University of California, Davis
Southern Methodist University
In four studies, the authors examined the correlates of the disposition toward gratitude. Study 1 revealed
that self-ratings and observer ratings of the grateful disposition are associated with positive affect and
well-being, prosocial behaviors and traits, and religiousness/spirituality. Study 2 replicated these findings
in a large nonstudent sample. Study 3 yielded similar results to Studies 1 and 2 and provided evidence
that gratitude is negatively associated with envy and materialistic attitudes. Study 4 yielded evidence that
these associations persist after controlling for Extraversion/positive affectivity, Neuroticism/negative
affectivity, and Agreeableness. The development of the Gratitude Questionnaire, a unidimensional
measure with good psychometric properties, is also described.
Gratitude, as it were, is the moral memory of mankind.
The gratitude of most men is but a secret desire of receiving more
Despite centuries of reflection, scholars in the humanities con-
tinue to debate whether the disposition to be grateful is a trait
worthy of admiration or of contempt. In one camp, philosophers
such as Seneca, Adam Smith, and Georg Simmel have praised the
value of a grateful disposition for individual and social well-being.
In another, figures such as Aristotle, Epicurus, and La Rochefou-
cauld have concluded that manifestations of gratitude are no more
than thin veils over human beings’ self-interest, or messy emo-
tional ties that make people unnecessarily beholden to their bene-
factors (Harpham, 2000; Roberts, 2000).
Psychologists have done little to contribute to this centuries-old
debate (McCullough, Kilpatrick, Emmons, & Larson, 2001). How-
ever, psychology’s inattention to gratitude belies the considerable
individual differences in gratitude of which laypersons are aware.
People easily call to mind individuals in their lives who seem to be
grateful almost to a fault and others who seem perfectly qualified
for the infamous label “ingrate.” Are these individual differences
real? Do people who consider themselves grateful also appear
grateful to others? What are the typical characteristics of people
who demonstrate a grateful disposition? What is the relationship of
gratitude to measures of affect, prosociality, and spirituality? Psy-
chology is perfectly positioned to help illuminate the nature of
gratitude and its place in human functioning.
Gratitude as an Affective Trait
Rosenberg (1998) proposed that the common forms of affective
experience could be structured hierarchically according to speci-
ficity, temporal stability, pervasiveness in consciousness, and ef-
fects on other psychological systems. Rosenberg placed affective
traits, defined as “stable predispositions toward certain types of
emotional responding” that “set the threshold for the occurrence of
particular emotional states” (p. 249) at the top of the hierarchy. She
considered moods, which “wax and wane, fluctuating throughout
or across days” (p. 250), as subordinate to affective traits, and
emotions, which are “acute, intense, and typically brief psycho-
physiological changes that result from a response to a meaningful
situation in one’s environment” (p. 250), as subordinate to both
affective traits and moods.
Gratitude, like other affects, conceivably could exist as an
affective trait, a mood, or an emotion. The present article is
concerned primarily with gratitude as an affective trait that we call
the grateful disposition or disposition toward gratitude. We define
the grateful disposition as a generalized tendency to recognize and
respond with grateful emotion to the roles of other people’s be-
nevolence in the positive experiences and outcomes that one
Facets of the Grateful Disposition
Affective traits lower one’s threshold for experiencing certain
emotional states (Rosenberg, 1998). For example, hostility lowers
Michael E. McCullough and Jo-Ann Tsang, Department of Psychology,
Southern Methodist University; Robert A. Emmons, Department of Psy-
chology, University of California, Davis.
This research was generously supported by a grant from the John
Templeton Foundation. We gratefully acknowledge the assistance of Rob-
ert Scott and Spirituality and Health magazine, and Lori Aveni and
Stephan Sain with the collection and management of data for Study 2.
Correspondence concerning this article should be addressed to Michael
E. McCullough, Department of Psychology, Southern Methodist Univer-
sity, P.O. Box 750442, Dallas, Texas 75275-0442, or to Robert A. Em-
mons, Department of Psychology, University of California, 1 Shields
Avenue, Davis, California 95616. E-mail: firstname.lastname@example.org or
Journal of Personality and Social Psychology Copyright 2002 by the American Psychological Association, Inc.
2002, Vol. 82, No. 1, 112–127 0022-3514/02/$5.00 DOI: 10.1037//0022-35220.127.116.11
the threshold for experiencing anger. Insofar as the grateful dis-
position creates a reduced threshold for recognizing and respond-
ing with gratitude to the role of other people’s benevolence in
one’s positive outcomes, this disposition might cause several dis-
crete emotional experiences. We use the term facets to refer to the
following elements of the grateful disposition rather than the term
dimensions because we suspect that these elements are not distinct
or independent but, rather, co-occur.
The first facet of the grateful disposition can be called intensity.
A dispositionally grateful person who experienced a positive event
is expected to feel more intensely grateful than would someone
less disposed toward gratitude. A second facet can be called
frequency. A dispositionally grateful person might report feeling
grateful many times each day, and gratitude might be elicited by
even the simplest favor or act of politeness. Conversely, for
someone less disposed toward gratitude, gratitude would be expe-
rienced less frequently. A third facet can be called span. Gratitude
span refers to the number of life circumstances for which a person
feels grateful at a given time. Dispositionally grateful people might
be expected to feel grateful for their families, their jobs, their
health, and life itself, along with a variety of other benefits. People
less disposed to gratitude might experience gratitude for fewer
aspects of their lives. A fourth facet can be called density, which
refers to the number of persons to whom one feels grateful for a
single positive outcome. When asked to whom one feels grateful
for a certain outcome (say, obtaining a good job), a dispositionally
grateful person might list many other people, including parents,
friends, family, and mentors. Someone less disposed toward grat-
itude might feel grateful to fewer people for the same outcome.
The Grateful Disposition and Attributional Breadth
The grateful disposition is not merely a tendency to experience
a particular affect (i.e., gratitude); it also emerges from particular
attributions regarding the causes of one’s positive outcomes.
Weiner (1986) proposed that gratitude became distinct from hap-
piness through a two-step process. First, people recognize that they
have obtained a positive outcome, which causes happiness. Sec-
ond, people attribute their happiness to an external source (viz.,
another person who acted intentionally), and consequently, happi-
ness is labeled as gratitude. Thus, attributions are central to grat-
itude, and attributional style may be central to the disposition
Attribution theory in its basic (e.g., Weiner, 1986, 1995) and
clinical (Abramson, Seligman, & Teasdale, 1978) formulations
sets the stage for hypotheses regarding the attributional style of
dispositionally grateful people. Weiner’s (1986) insights lead to
the hypothesis that dispositionally grateful people tend to attribute
their positive outcomes to the effort of other people. Correspond-
ingly, dispositionally grateful people might seem less likely to
attribute their successes and good fortune to their own efforts or
positive qualities, and thus, more prone to the psychological dif-
ficulties associated with the externalizing attributional style (e.g.,
Abramson et al., 1978; Gladstone & Kaslow, 1995).
However, the fact that grateful people tend to recognize the
benevolence of other people in their positive outcomes does not
necessarily mean that they discount their own causal effort. In-
stead, what might distinguish grateful people is an ability to stretch
their attributions to incorporate the wide range of people who
contribute to their well-being. The thirteenth century theologian
and philosopher Thomas Aquinas (cited in Harpham, 2000) ob-
served that recognizing the many people who contribute to one’s
positive outcomes was essential to gratitude:
The nature of a debt to be paid must needs vary according to various
causes giving rise to the debt, yet so that the greater always includes
the lesser. Now the cause of debt is found primarily and chiefly in
God, in that He is the first principle of all our goods: secondarily it is
found in our father, because he is the proximate principle of our
begetting and upbringing: thirdly it is found in the person that excels
in dignity, from which general favors proceed; fourthly it is found in
a benefactor, from whom we have received particular and private
favors, on account of which we are under particular obligation to him.
(Summa Theologica, Question 106, Article 1)
Consider two successful Olympic swimmers—one more and
one less dispositionally grateful. The grateful swimmer would, no
doubt, recognize her own effort in obtaining an Olympic gold
medal. However, because of her grateful disposition, she may also
recognize other people’s benevolent contributions to her success in
addition to her own effort. For example, she also might attribute
her success to the effort of her coaches, her parents, teachers who
tolerated her demanding workout schedule, and even teammates or
competitors who challenged her to swim at her best. Clearly, some
of these people had strong and direct influences on her success,
whereas others’ influences may have been more distal and subtle.
Conversely, a less grateful Olympic swimmer might not consider
the full range of people who contributed (albeit perhaps distally
and subtly) to her success, and might therefore attribute her suc-
cess solely to her own effort. Thus, the distinctive attributional
quality of grateful people may not be that they have a more
externalizing style of attributions for their positive outcomes, but
rather, that they recognize the many people who contribute to their
Personality Correlates of the Grateful Disposition:
A Superordinate Traits Analysis
Given the preceding analysis, we suspect that the grateful dis-
position is linked to other personality traits, most notably (a)
positive affective traits and well-being, (b) prosocial traits, and (c)
religion/spirituality. To some extent, the correlations of the grate-
ful disposition with these three clusters of individual differences
might be explained in terms of the Big Five (John & Srivastava,
1999) or Five-Factor (McCrae & Costa, 1999) personality
The Emotions and Well-Being of Grateful People
Grateful people may be prone to positive emotions and subjec-
tive well-being. Several theorists and researchers (e.g., Lazarus &
Lazarus, 1994; Mayer, Salovey, Gomberg-Kaufman, & Blainey,
1991; Ortony, Clore, & Collins, 1988; Weiner, 1986) have noted
that gratitude typically has a positive emotional valence. From this
fact, we surmise that the disposition toward gratitude is rooted in
basic tendencies to experience positive emotions and subjective
well-being. However, there are other reasons to expect that grate-
ful people experience positive emotions and heightened well-
being: Seeing oneself as the beneficiary of other people’s gener-
osity may lead one to feel affirmed, esteemed, and valued, which
may boost self-esteem and perceived social support. Moreover,
highly grateful people may possess a worldview in which every-
thing they have—and even life itself—is a gift. This level of
appreciation for the good things in one’s life may lead grateful
people to avoid taking benefits for granted. As a result, they may
be less prone to habituate to positive life circumstances, which
might also help sustain their happiness and subjective well-being
Invoking the Big Five taxonomy, we hypothesized that dispo-
sitionally grateful people have high levels of Extraversion/positive
affectivity and low levels of Neuroticism/negative affectivity be-
cause these two superordinate personality dimensions are highly
relevant to emotional experience. We also suspected that grateful
people experience higher levels of other specific positive emotions
such as happiness, vitality, optimism, and hope, as well as greater
satisfaction with life. Conversely, we suspected that they tend to
experience low levels of negative emotions such as anxiety, de-
pression, and envy.
The Prosocial Traits of Grateful People
Gratitude has been called an “empathic emotion” (Lazarus &
Lazarus, 1994) because it is predicated on the capacity for recog-
nizing the beneficial actions of other people in one’s life. Mc-
Cullough et al. (2001) proposed that gratitude is relevant to the
moral domain in the same way that affects such as guilt (Tangney,
1991), shame (Keltner & Buswell, 1996), empathy (Batson, 1991),
and even contempt, anger, and disgust (Rozin, Lowery, Imada, &
Haidt, 1999) are morally relevant. Specifically, gratitude might be
considered a prosocial affect because it is a response to behaviors
that other people enact to contribute to one’s welfare, and may
actually motivate such behaviors in turn (McCullough et al., 2001;
McCullough & Tsang, in press).
The prosocial nature of gratitude suggests the possibility that the
grateful disposition is rooted in the basic traits that orient people
toward sensitivity and concern for others. Again invoking the Big
Five taxonomy, we suggest that grateful people may tend to be
higher in Agreeableness, which appears to facilitate prosocial and
other-oriented behavior (Koole, Jager, van den Berg, Vlek, &
Hofstee, 2001). Indeed, Saucier and Goldberg (1998) reported that
a two-item personality measure consisting of the adjectives grate-
ful and thankful was correlated r ⫽ .31 with Agreeableness.
Moreover, one might expect the disposition toward gratitude to be
related to other traits that emerge from Agreeableness such as a
capacity for empathy, willingness to forgive, and a tendency to
offer help and support to others. Grateful people probably are also
disinclined to experience negative interpersonal emotions such as
envy, which is not simply a negative affect, but rather a type of
negative emotion (viz., resentment and disdain) that is directed
specifically at other people (Smith, Parrott, Diener, Hoyle, & Kim,
1999). Relatedly, we expect that dispositionally grateful people are
less concerned with materialistic pursuits, which would be some-
what incompatible with a focus on generosity and helpfulness.
The Spiritual Traits of Grateful People
Just as gratitude accompanies the recognition of the positive
contributions of other people to one’s well-being, dispositionally
grateful people may also be oriented toward recognition of non-
human forces that might contribute to their well-being in a broader,
more existential sense (viz., luck, chance, God, or some other
conception of the divine). Stated another way, grateful people may
tend to be spiritually inclined—a trait that is surprisingly distinct
from the Big Five (Piedmont, 1999; Saucier & Goldberg, 1998).
Justification for this hypothesis comes from several sources.
First, many world religions commend gratitude as a desirable
human trait (see Carman & Streng, 1989; Emmons & Crumpler,
2000), which may cause spiritual or religious people to adopt a
grateful outlook. Second, when confronted with a positive out-
come that cannot be attributed to intentional human effort, such as
a lovely sunset or the gift of sight, spiritually inclined people may
still be able attribute these positive outcomes to a human or
nonhuman agent (viz., God or a higher power) and thus, experi-
ence more gratitude. Third, spiritually inclined people also tend to
attribute positive outcomes to God’s intervention, but not negative
ones (Lupfer, De Paola, Brock, & Clement, 1994; Lupfer, Tolliver,
& Jackson, 1996). As a result, many positive life events that are
not due to the actions of another person (e.g., pleasant weather,
avoiding an automobile accident) may be perceived as occasions
for gratitude to God, although negative events (e.g., a long winter,
an automobile accident) would likely not be attributed to God.
It is possible that causal links between the grateful disposition
and spirituality run in the other direction as well. Allport,
Gillespie, and Young (1948) found that 37% of their college-
student sample cited gratitude as a reason why they were religious.
For these many reasons, we hypothesized that dispositionally
grateful people score higher on measures of spirituality and
Is the Grateful Disposition a Distinct Construct?
Thus far, we have posited that the grateful disposition is real,
unique, and worthy of attention in its own right, even if also
correlated with a host of better understood traits and characteris-
tics. In this situation, it is prudent to anticipate the objection that
the grateful disposition is no more than a repackaging of some
other construct. This objection has prima facie merit for theoretical
and practical reasons. Theoretically, if gratitude is simply a man-
ifestation of more fundamental constructs (e.g., Extraversion/pos-
itive affectivity, Neuroticism/negative affectivity, or Agreeable-
ness), then the grateful disposition might reveal little of enduring
interest about human personality and social functioning. Practi-
cally, if gratitude is simply the product of such better understood
traits, then examining gratitude on its own terms might only
contribute to the explosion of traits (and measures) that has char-
acterized psychology during the twentieth century.
To this objection, one might argue that even if gratitude were
simply a linear combination of more basic constructs, the concept
could still reveal important insights at a different level of analysis.
For example, if gratitude could be reduced to a linear combination
of the Big Five (which most likely it cannot; Saucier & Goldberg,
1998), then gratitude still could be of interest as a characteristic
adaptation that more extraverted (or less neurotic or more agree-
able) people use to navigate their worlds (see McCrae & Costa,
The objection could also be examined empirically. If the items
used as indicators of the grateful disposition correlate as highly
MCCULLOUGH, EMMONS, AND TSANG
with the items on a measure of some other lower order construct—
say, hope, happiness, optimism, vitality, or satisfaction with
life—as they do with each other, then the concept of gratitude
would seem unnecessary for understanding the class of human
behavior or experience in question. Also, if gratitude’s associa-
tions with such lower order constructs can be explained in terms of
their common associations with superordinate traits (e.g., Extra-
version/positive affectivity, Neuroticism/negative affectivity,
Agreeableness) one might safely conclude that the disposition
toward gratitude is a basic adaptation emerging from more basic
Overview of the Studies
In the studies described below, we addressed several objectives
regarding the disposition toward gratitude. First, we developed
self-report measures of the grateful disposition and examined their
convergence with informant ratings. Second, we confirmed that
the disposition toward gratitude is empirically distinct from con-
structs such as life satisfaction, vitality, happiness, hope, and
optimism. Third, we tested the hypotheses that the disposition
toward gratitude is correlated with lower order traits such as (a)
emotions and well-being; (b) prosocial traits such as empathy,
forgiveness, and willingness to help others; and (c) spirituality,
religiousness, and nonmaterialistic attitudes. Fourth, we examined
the relation of the disposition toward gratitude with the Big Five
personality traits. Fifth, we explored whether the associations of
gratitude with lower order traits occurred through their common
associations with the superordinate traits of Extraversion/positive
affectivity, Neuroticism/negative affectivity, and Agreeableness,
as well as with social desirability.
Study 1 was an initial investigation of the correlates of the
grateful disposition using two methods for assessing gratitude.
First, we constructed a self-report measure called the Gratitude
Questionnaire–6 (GQ-6). We examined the convergence of the
GQ-6 with informant ratings and confirmed that gratitude was
empirically distinct from a host of other conceptually related
constructs. Then, we examined the correlations of the self-ratings
and informant ratings of the disposition toward gratitude with
measures of positive affect/well-being, prosociality, spirituality/
religiousness, and the Big Five.
The participants in Study 1 were 238 undergraduate psychology students
(174 women, 57 men, 7 unrecorded). Participants’ mean age was 21 years
(range ⫽ 19–44). Participants received course credit for participating.
Self-Report Measures of Gratitude
We administered 39 positively and negatively worded items that assess
experiences and expressions of gratefulness and appreciation in daily life,
as well as feelings about receiving from others. Items reflected the grati-
tude intensity facet (e.g., “I feel thankful for what I have received in life.”),
the gratitude frequency facet (e.g., “Long amounts of time can go by before
I feel grateful to something or someone.”), the gratitude span facet (e.g. “I
sometimes feel grateful for the smallest things.”), and the gratitude density
facet (e.g., “I am grateful to a wide variety of people.”). Respondents
endorsed each item on a 7-point Likert-type scale ranging from 1 (strongly
Measures of Affectivity and Life Satisfaction
Life satisfaction. The five-item Satisfaction With Life Scale (SWLS;
Diener, Emmons, Larsen, & Griffin, 1985) assesses the cognitive compo-
nent of subjective well-being. Items are rated on a 7-point scale ranging
from 1 (strongly disagree)to7(strongly agree). The SWLS has a 2-month
test–retest correlation coefficient of .82 and coefficient alpha of .87 (Diener
et. al., 1985). The SWLS is widely used and well-validated (Pavot &
Vitality. Subjective vitality is a feeling of aliveness, energy, and en-
thusiasm, representing the “nexus of physical and psychological well-
being” (Ryan & Frederick, 1997). It was assessed with the seven-item
Vitality Scale (a sample item is “I have energy and spirit”). The Vitality
Scale is related to both physical and mental health outcomes and possesses
good psychometric properties.
Subjective happiness. The Subjective Happiness Scale (Lyubomirsky
& Lepper, 1999) is a four-item scale that measures global subjective
happiness. The scale has high test–retest reliability, internal consistency,
and self–peer correlations.
Optimism. The widely used Life Orientation Test (LOT; Scheier,
Carver, & Bridges, 1994) is an eight-item scale for assessing dispositional
optimism. Scheier et al. reported a coefficient alpha of .82 and test–retest
stabilities ranging from .56 to .79 across four time periods.
Hope. The Adult Trait Hope Scale (Snyder, et al., 1991) is an eight-
item measure that assesses two dimensions of hope: agency and pathways.
The hope scale possesses favorable psychometric properties and predicts
several measures of performance, health, and adjustment (Snyder, 2000).
Positive and negative affect. The Positive and Negative Affect Scales
(PANAS; Watson, Clark, & Tellegen, 1988) measure general tendencies to
experience positive (e.g., pride) and negative (e.g., guilt) affect. Partici-
pants used a 5-point Likert-type scale ranging from 1 (very slightly or not
at all)to5(extremely) to indicate how well each of 20 adjectives described
“how [they] generally feel.” Coefficient alphas of the positive and negative
scales range in the middle to upper .80s (Watson et al., 1988).
Psychological symptoms. We measured anxiety and depressive symp-
toms with the Anxiety and Depression scales of the Brief Symptom
Inventory (Derogatis & Spencer, 1982). Participants rate the items on these
subscales with a 5-point scale ranging from 1 (not at all)to5(extremely).
Items in the Anxiety scale describe symptoms characteristic of high levels
of manifest anxiety and the cognitive and somatic correlates of anxiety.
The Depression scale consists of items that characterize clinical depression
including negative moods, low motivation, and social isolation. Derogatis
and Spencer (1982) reported coefficient alphas of .81 for the Anxiety scale
and .85 for the Depression scale.
Measures of Prosociality
Dispositional empathy. The disposition toward empathy was measured
with the Empathic Concern and Perspective-Taking subscales of the Inter-
personal Reactivity Index (Davis & Oathout, 1987). These subscales have
adequate internal consistency (
⫽ .73 and .71, respectively, Davis &
Oathout, 1987), and are frequently used to measure empathic disposition.
Social desirability. Social desirability was assessed with the Balanced
Inventory of Desirable Responding (BIDR; Paulhus, 1998). This 40-item
inventory assesses two self-favoring defensive tendencies: self-deceptive
enhancement (SDE) and impression management (IM). Coefficient alphas
range from .67 to .77 for SDE and .77 to .85 for IM (Paulhus, 1998).
Measures of Spirituality and Religiousness
Spiritual transcendence. The Spiritual Transcendence Scale (STS;
Piedmont, 1999) is a 24-item scale consisting of subscales for assessing
three dimensions of spirituality: Prayer fulfillment (e.g., “The desires of my
body do not keep me from my prayers or meditations”), universality (“I
believe there is a larger meaning to life”), and connectedness (“Iam
concerned about those who will come after me in life”). Coefficient alphas
for these subscales range from .65 to .85. Subscale scores are moderately
correlated with conventional indexes of religiousness and are relatively
independent of the Big Five (Piedmont, 1999). In the present study, we
used the total scale score as a measure of spirituality.
Self-transcendence. Fifteen items from the self-transcendence subscale
of the Character and Temperament Inventory (Cloninger, Svrakic, &
Przybeck, 1993) were used to assess three aspects of spirituality: self-
forgetful versus self-conscious experience, transpersonal identification ver-
sus self-isolation, and spiritual acceptance versus rational materialism.
Kirk, Eaves, and Martin (1999) developed a 15-item measure (sample item
is “I have had personal experiences in which I felt in contact with a divine
and wonderful spiritual power”) from the full-length 33-item subscale.
Cloninger et al. reported alphas in the low .70s for the 33-item version.
Internal consistency of the 15-item version in the present study was
alpha ⫽ .86. Items were endorsed on a 4-point Likert-type scale ranging
from 1 (agree)to4(disagree).
Other religious variables. Participants also completed several other
items related to religiousness, including items assessing the importance of
religion (“How important is religion in your life?”), the frequency with
which they attend religious activities (“On average, how often do you
attend religious services?”), the number of religious friends they have
(“Are your friends involved in religious activities?” with none, a few, most,
and all as response options), the amount of time they spent reading
scripture (“How often do you read sacred scriptures?”) and other religious
literature (“How often do you read other religious literature?”), the fre-
quency with which they pray (“How often do you pray?”), the extent to
which they felt they had a personal relationship with God (“To what extent
do you have a personal, unique, close relationship with God?”), and the
extent to which they experience spiritual union with God (“Do you have
experiences where you feel a union with God and gain spiritual truth?”).
Measures of the Big Five
The Big Five Inventory (John, Donahue, & Kentle, 1991) consists of 44
brief descriptive phrases that are prototypical markers for five broad
personality dimensions: Agreeableness, Conscientiousness, Extraversion,
Neuroticism, and Openness. Alpha reliabilities and test–retest reliabilities
for the five subscales range from .80 to .90 (John & Srivastava, 1999).
The 238 participants were asked to identify four people (friends, rela-
tives, or romantic partners) who knew them well. Participants asked these
informants to complete an informant rating form that included several
measures. Questionnaires were worded specifically for male and female
targets. Two or more informant reports were returned for 168 individuals.
Of the 639 total informants, 444 (69.5%) were friends or roommates, 54
(8.5%) were boyfriends or girlfriends, 49 (7.7%) were parents, 3 (⬍ 1%)
were spouses, 60 (9.4%) were other family members, and 29 (4.5%) were
people who knew the participants in some other way.
Gratitude ratings. The informant form included 12 items from the
gratitude item pool described above, with wording changed from the first
person to the third. Informants rated the participants on the same 7-point
scale that participants used in the self-ratings. Within informants, these 12
items had internal consistency reliabilities (
) of approximately .85. Inter-
rater reliability was calculated using the formula for Case 1 in Shrout and
Fleiss (1979, p. 421), using an analysis of variance components (Hoyt &
Melby, 1999). Using this formula, interrater reliability was .65.
Prosocial behavior. Informants were also asked to rate the frequency
with which they were the recipients of prosocial actions performed by the
target person (e.g. “[she/he] loaned me money,”“[she/he] provided com-
passion or sympathy,”“[she/he] made me feel good about myself”) during
“the last month or so.” Informants rated these items on a 5-point Likert-
type scale ranging from 1 (not at all)to5(frequently). We combined scores
on these informant reports into a single five-item measure of perceived
prosocial behavior in the last month.
Informants also completed several measures of their impressions of
participants’ general prosocial tendencies (e.g., “[She/he] tends to go out of
[his/her] way to help others,”“[She/he] has volunteered [his/her] time to
help others,” etc.). These items were rated on a 5-point scale ranging
from 1 (not at all characteristic of the participant)to5(extremely char-
acteristic of the participant).
Informants also rated participants with the Big Five Inventory (John et
During a class period, participants received a packet containing the
self-report scales and four envelopes and questionnaire packets containing
the informant-report scales. Participants were instructed to give each en-
velope to an individual who knew them well with the instructions that the
forms should be mailed directly back to the researcher within 1 week. They
were instructed not to discuss the questionnaires or ratings with their
informants. A total of 656 peer reports were returned. Participants for
whom at least three reports were not returned were not included in
aggregate informant reports described below.
Scale Construction: Initial Analyses
First, we conducted correlational and exploratory factor analy-
ses with the 39 items in our initial pool of items for measuring the
grateful disposition. An exploratory factor analysis on the 39 items
revealed one large factor that explained 27% of the total item
variance. Although there were 10 other factors with eigenvalues
exceeding unity, none of these factors accounted for more than 7%
of the total item variance. A scree plot also suggested only one
meaningful factor. From the item pool, we retained six items that
(a) loaded strongly on the first factor and (b) assessed unique
aspects of the grateful disposition. These six items constituted the
initial version of the GQ-6.
Structural Equation Models
The validity of a one-factor solution for the six retained items
was assessed with structural equation models with maximum-
likelihood estimation in EQS Version 5.7b (Bentler, 1995). To
assess goodness of fit, we examined the chi-square, comparative fit
index (CFI) and standardized root-mean-square residual (SRMR;
Bentler, 1995) statistics. Hu and Bentler (1998) noted that SRMR
is less sensitive to distribution and sample size, and recommended
its use in combination with CFI when using maximum-likelihood
estimation. CFI values greater than .95 and SRMR values less than
.05 are typically considered to indicate that a structural equation
model is adequately parameterized (Hu & Bentler, 1998), although
values as low as .90 and as high as .10, respectively, are
MCCULLOUGH, EMMONS, AND TSANG
First, we estimated a one-factor congeneric measurement model.
This model yielded a large and significant chi-square,
(9, N ⫽
235) ⫽ 30.34, p ⬍ .001. The CFI, which is less sensitive to sample
size, was also large (.95), indicating that the one-factor model
provided an adequate fit to the data. For our six-item scale, SRMR
was .04. The internal consistency reliability of the six-item scale
was alpha ⫽ .82. The final version of the GQ-6 appears in the
We proceeded to distinguish our measure of the grateful dispo-
sition from several related but distinct constructs: satisfaction with
life, vitality, subjective happiness, optimism, and hope. If the
grateful disposition construct is distinct from the latter five con-
structs, then it should be necessary to specify two unique but
correlated factors to account for the covariances among the items
for the GQ-6 and any of the other five scales.
For each of these tests of discriminant validity, we used
EQS 5.7b to estimate a one-factor solution specifying a single
latent construct underlying the items on the GQ-6 and each of the
other respective scales. Second, we estimated a two-factor solution
specifying that the items on the GQ-6 and each of the other
respective scales loaded on distinct but correlated latent variables.
If the addition of a second latent variable to account for the
covariances among both sets of items led to an improvement in
model fit (i.e., if the two-factor model was superior to the one-
factor model), we could conclude that both constructs are neces-
sary for describing the two sets of items, and thus, that the two
constructs are reasonably distinctive.
We evaluated the fit of these one-factor and two-factor models
with the chi-square, CFI, and SRMR indices. We analyzed the
relative improvements in goodness of fit associated with moving
from a one-factor model to a two-factor model with a nested
chi-square test. The nested chi-square test measures the difference
in chi-square values between nested models that differ by one
parameter. Differences in chi-square values themselves are chi-
square distributed, and can be evaluated for statistical significance
by evaluating them against the chi-square distribution with 1
degree of freedom. Statistically significant reductions in chi-square
suggest that the additional parameter improved model specifica-
tion (Hoyle & Panter, 1995).
Gratitude and life satisfaction. The one-factor model combin-
ing the gratitude items with SWLS items fit the data poorly,
N ⫽ 231) ⫽ 352.95, p ⬍ .001, CFI ⫽ .71, SRMR ⫽ .13. In
contrast, a model with gratitude and life satisfaction as two sepa-
rate factors showed a better fit,
(43, N ⫽ 231) ⫽ 74.89, p ⬍
.001, CFI ⫽ .97, SRMR ⫽ .04, ⌬
⫽ 278.06, p ⬍ .05, demon
strating that the introduction of a second factor to account exclu-
sively for the covariances among the items on the GQ-6 led to a
significant improvement in model fit. The gratitude and life satis-
faction latent factors were correlated at r ⫽.53, p ⬍ .05.
Gratitude and vitality. The one-factor solution for gratitude
and vitality did not fit the data well,
(65, N ⫽ 234) ⫽ 481.26,
p ⬍ .001, CFI ⫽ .68, SRMR ⫽ .14, and the two-factor solution fit
the data better,
(64, N ⫽ 234) ⫽ 166.44, CFI ⫽ .92, SRMR ⫽
⫽ 314.82, p ⬍ .05. The gratitude and vitality latent
factors were correlated at r ⫽ .46, p ⬍ .05.
Gratitude and happiness. The one-factor model for describing
the relations among the items on the GQ-6 and subjective happi-
ness (Lyubomirsky & Lepper, 1999) scales did not fit the data
(35, N ⫽ 234) ⫽ 355.54, p ⬍ .001, CFI ⫽ .66, SRMR ⫽
.12., and the two-factor solution fit the data better,
(34, N ⫽
234) ⫽ 103.90, CFI ⫽ .93, SRMR ⫽ .05, ⌬
⫽ 251.64, p ⬍ .05.
The gratitude and happiness latent factors were correlated at r ⫽
.50, p ⬍ .05.
Gratitude and optimism. We also evaluated the relations
among the items on the GQ-6 and the items on the LOT measure
of optimism. A one-factor model yielded a poor fit on all three
(54, N ⫽ 233) ⫽ 275.42, p ⬍ .001, CFI ⫽ .74,
SRMR ⫽ .11. In comparison, a two-factor solution with gratitude
and optimism as distinct factors fit the data more acceptably,
(53, N ⫽ 233) ⫽ 103.16, p ⬍ .001, CFI ⫽ .94, SRMR ⫽ .05.
The difference in chi-square between the one-factor and two-factor
models was significant, ⌬
⫽ 171.86, p ⬍ .05. The two latent
factors were correlated at r ⫽ .51, p ⬍ .05.
Gratitude and hope. We conducted similar tests with Snyder
et al.’s (1991) hope scale. We analyzed the relations between the
GQ-6 items and the items on the agency and pathways subscales
separately. The one-factor model combining gratitude and agency
items was a poor fit to the data,
(35, N ⫽ 232) ⫽ 192.70, p ⬍
.001, CFI ⫽ .82, SRMR ⫽ .08. The two-factor model with sepa-
rate factors for gratitude and agency appeared more adequate,
(34, N ⫽ 232) ⫽ 81.79, p ⬍ .001, CFI ⫽ .94, SRMR ⫽ .05. The
change in chi-square was significant, ⌬
⫽ 110.91, p ⬍ .05,
revealing the two-factor solution to be an improvement in fit over
a one-factor model. The gratitude and agency latent factors were
correlated at r ⫽ .67, p ⬍ .05.
Similar analyses were conducted comparing one-factor and two-
factor models of the covariances among the GQ-6 items and the
items on the pathways subscale of the hope scale. The one-factor
model yielded a poor fit,
(35, N ⫽ 233) ⫽ 273.79, p ⬍ .001,
CFI ⫽ .69, SRMR ⫽ .13. The two-factor model displayed a better
(34, N ⫽ 233) ⫽ 77.11, p ⬍ .001, CFI ⫽ .95, SRMR ⫽ .05,
and was a significant improvement in fit, ⌬
⫽ 196.68, p ⬍ .05.
The gratitude and pathways latent factors were correlated at r
⫽.42, p ⬍ .05.
Correlations With Informant Ratings of Gratitude
Because the GQ-6 appeared to be both psychometrically ade-
quate and sufficiently distinct from several other positive con-
structs, we proceeded to examine the convergence of self-reports
with informant reports of the grateful disposition. We computed a
composite informant rating of the grateful disposition by averaging
informants’ ratings on the 12 gratitude items and then averaging
across four informants. Participants’ ratings on the GQ-6 (M
⫽ 5.92, SD ⫽ 0.88) were positively correlated with mean infor-
mant ratings on the 12-item gratitude scale (M ⫽ 5.52, SD ⫽ 0.53),
r ⫽ .33, p ⬍ .01. Thus, people’s perceptions of their own grateful
disposition show a modest but significant correspondence with
external observers’ perceptions.
Correlations of Gratitude With Other Constructs
We correlated the GQ-6 and the mean informant ratings of the
grateful disposition with the other measures of personality. These
correlations appear in Table 1.
Affect and well-being. As predicted, both the self-report and
informant-report measures of the grateful disposition were corre-
lated positively with all of the measures of positive affect and
well-being (e.g., life satisfaction, vitality, subjective happiness,
optimism, hope, and positive affectivity), and negatively with all
of the measures of negative affect (negative affectivity, anxiety,
Prosociality. As predicted, the self-report and informant-
report measures of the grateful disposition were also correlated
positively with the Empathic Concern and Perspective-Taking
subscales of the empathy measure. In addition, the measures of the
grateful disposition were correlated positively with informants’
reports of participants’ prosocial behaviors. Participants who were
rated by themselves (and by their informants) as being more
grateful were reported to perform more prosocial behaviors (e.g.,
providing favors, as well as emotional and tangible support) for
their informants than were less grateful people. Moreover, infor-
mants rated grateful people as having more prosocial traits gener-
ally than did the informants of less grateful people.
Religiousness and spirituality. As predicted, the measures of
the grateful disposition were positively correlated with nearly all
of the measures of spirituality and religiousness, including spiritual
transcendence, self-transcendence, and the single-item religious
Social desirability. Gratitude was modestly correlated with the
SDE (r ⫽ .19, p ⬍ .01) and IM (r ⫽ .21, p ⬍ .01) subscales of the
social desirability scale. The mean of the informant ratings was
also correlated significantly with SDE (r ⫽ .17, p ⬍ .01) but not
with IM (r ⫽ .07, ns.).
The Big Five. The grateful disposition was related to several of
the Big Five (see Table 1). Specifically, gratitude was correlated
with Agreeableness (positively), Extraversion (positively), and
Neuroticism (negatively) for both self-ratings (i.e., the GQ-6) and
informant ratings. Perhaps most importantly, these correlations
replicated not only within raters, but across raters as well. For
example, self-ratings of Agreeableness were significantly corre-
lated with both self-rated and peer-rated gratitude and vice versa.
Regressing gratitude on the Big Five in self-ratings. We re-
gressed the GQ-6 on the Big Five self-ratings (see Table 2). The
self-reports of the Big Five personality variables explained 21% of
Correlations of Two Measures of Gratitude With Measures of
Affectivity and Well-Being, Prosociality, and
Spirituality/Religiousness (Study 1)
Affectivity and well-being
Life satisfaction (SWLS) .53**
Subjective happiness .50**
Optimism (LOT) .51**
Positive and negative affect
Positive subscale .31** .32**
Negative subscale ⫺.31** ⫺.31**
Psychological symptoms (BSI)
Anxiety ⫺.20** ⫺.11
Depression ⫺.30** ⫺.26**
Prosocial traits and behaviors
Empathic concern .28** .17*
Perspective taking .32** .22**
M of 5 peer-reported prosocial behaviors in past
Peer-rated prosocial traits
Goes out of way to do favors for others .14 .47**
Has volunteered time to help others .19* .42**
Tends to be generous with time and resources .22** .50**
Expects other people to do him/her favors ⫺.20** ⫺.41**
Is helpful and unselfish with others .18* .54**
Spiritual and religious variables
Spiritual transcendence (STS) .28** .30**
Self-transcendence (CTI) .24** .19*
Importance of religion .22** .23**
Frequency of attendance .23** .31**
Religious friends .29** .20*
Reading scripture .16* .21**
Other religious reading .14* .24**
Prayer .19** .32**
Personal relationship with God .25** .22**
Union with God .23** .17*
Social desirability (BIDR)
Self-deception .19** .17*
Impression management .21** .07
Big Five personality traits (self-ratings)
Agreeableness .39** .28**
Conscientiousness .27** .12
Extraversion .20** .26**
Neuroticism ⫺.30** ⫺.28**
Openness .23** .13
Big Five personality traits (peer ratings)
Agreeableness .17* .59**
Conscientiousness .14† .37**
Extraversion .16* .41**
Neuroticism ⫺.16* ⫺.49**
Openness .18* .45**
Note. GQ–6 ⫽ Gratitude Questionnaire–6; SWLS ⫽ Satisfaction With Life
Scale; LOT ⫽ Life Orientation Test; BSI ⫽ Brief Symptom Inventory; STS ⫽
Spiritual Transcendence Scale; CTI ⫽ Character and Temperament Inventory;
BIDR ⫽ Balanced Inventory of Desirable Responding.
† p ⬍ .10. * p ⬍ .05. ** p ⬍ .01.
Correlations between latent variables as estimated with maximum-
likelihood structural equation models.
Multiple Regression of Two Measures of Gratitude
on the Big Five (Study 1)
Predictor variable BSE
Gratitude Questionnaire–6 .21**
Openness .13 .09 .10
Conscientiousness .11 .09 .08
Extraversion .08 .07 .08
Agreeableness .41 .10 .29**
Neuroticism ⫺.14 .07 ⫺.13†
M of 12-item peer ratings .52**
Openness .25 .08 .20**
Conscientiousness ⫺.05 .08 ⫺.04
Extraversion .26 .06 .26**
Agreeableness .50 .08 .43**
Neuroticism ⫺.19 .06 ⫺.20**
† p ⬍ .10. ** p ⬍ .01.
MCCULLOUGH, EMMONS, AND TSANG
the variance in GQ-6 scores, R
(5, 223) ⫽ .21, p ⬍ .01. Agree
ableness predicted unique variance in GQ-6 scores (
⫽ .29, p ⬍
.01), but the unique contribution of Neuroticism was only margin-
ally significant (
⫽⫺.13, p ⬍ .07). Self-reported Openness,
Conscientiousness, and Extraversion were not uniquely related to
GQ-6 scores, all ps ⬎ .10.
Regressing gratitude and the Big Five in informant ratings.
The aggregate informant rating of the disposition toward gratitude
was also regressed on the informant ratings of the Big Five (see
Table 2). The informant measures of the Big Five explained 52%
of the variance in informant-rated gratitude, R
(5, 156) ⫽ .52, p ⬍
.01. Informant ratings of Openness (
⫽ .20, p ⬍ .01), Agreeable-
⫽ .43, p ⬍ .01), Extraversion (
⫽ .26, p ⬍ .01), and
⫽⫺.20, p ⬍ .01) predicted unique variance in
informant-rated gratitude. Informant ratings of Conscientiousness
did not predict unique variance in informant-rated gratitude, p ⬎
In Study 1, we constructed a self-report measure of the grateful
disposition (the GQ-6) that converged moderately with informant
ratings. As we suspected, the items that we developed to assess the
various facets of the grateful disposition were not factorially
distinct, but rather seemed to reflect a single underlying factor.
Although the self-ratings and informant ratings only shared about
10% common variance (i.e., their correlation was .33), such mod-
est correlations are the rule rather than the exception in the con-
vergence of self-ratings and informant ratings (Paulhus & Reyn-
olds, 1995). The degree of convergence is actually impressive
given the private nature of grateful cognitions and the fact that
self⫺observer convergence is strongest for highly visible traits
(Funder & Colvin, 1997). Ratings of the grateful disposition were
correlated with measures of positive emotionality and well-being,
including vitality, happiness, satisfaction with life, lack of depres-
sive and anxious symptoms, hope, and optimism.
Also, people who rated themselves (or who were rated by
others) as having a grateful disposition perceived themselves (and
were perceived by others) as having prosocial characteristics. They
were more empathic and were perceived as providing more con-
crete and emotional help to their peers, both within the last month
and in general. This evidence is consistent with McCullough et
al.’s (2001) hypothesis that gratitude motivates prosocial behavior.
This conclusion must be tempered by the fact that the correlations
between informant ratings of gratitude and prosocial behavior may
have been due to halo error (Thorndike, 1920), although the
amount of halo error in correlations of rated dimensions is less
than was once believed to be the case (Murphy, Jako, & Anhalt,
The grateful disposition was also related to measures of spiritual
and religious tendencies. Although these correlations were not
large (i.e., few of them exceeded r ⫽ .30), they suggest that
spiritually or religiously inclined people have a stronger disposi-
tion to experience gratitude than do their less spiritual/religious
counterparts. Thus, spiritual and religious inclinations may facili-
tate gratitude, but it is also conceivable that gratitude facilitates the
development of religious and spiritual interests (Allport et al.,
1948) or that the association of gratitude and spirituality/religious-
ness is caused by extraneous variables yet to be identified. The fact
that the correlations of gratitude with these affective, prosocial,
and spiritual variables were obtained using both self-reports and
peer reports of the grateful disposition suggests that these associ-
ations are substantive and not simply the product of monomethod
biases in measurement.
Confirmatory factor analyses revealed that gratitude is related
to, but not equivalent to, happiness, vitality, satisfaction with life,
optimism, and hope. Moreover, the Big Five accounted for ap-
proximately 21% of the variance in the GQ-6, suggesting that the
grateful disposition is not reducible to a linear combination of the
Big Five (see also Saucier & Goldberg, 1998).
Study 2 was based on a survey that we conducted with nonstu-
dents through the Internet. We cross-validated the measurement
model for the GQ-6, examined the convergence of the GQ-6 with
an alternative self-report measure of the grateful disposition (an
adjective-based scale), and examined the correlations of both of
these measures of the grateful disposition with positive and neg-
ative affect, the disposition to forgive, spirituality, and the Big
Participants were 1,228 adult volunteers (M age ⫽ 44.6, SD ⫽ 12.0;
Range ⫽ 18–75). The sample comprised mostly women (80% women,
15% men, 5% did not provide gender) who were predominantly White/
Caucasian (91% White/Caucasian, 3% Hispanic/Latino, 2% Black/African
American, 1% Asian, 3% other). All participants were either visitors to the
web site for the magazine Spirituality and Health (www.spiritualityhealth.
com) or visitors to one of two other nationally known web sites dealing
with issues related to spirituality, religion, and health, who then linked to
the Spirituality and Health web site.
Gratitude. Participants completed the GQ-6. They also used a 9-point
Likert-type scale ranging from 1 (inaccurate)to9(accurate) to indicate
how accurately three gratitude-related adjectives (grateful, thankful, and
appreciative) described them. Internal consistency reliability for the grat-
itude adjectives scale was .87.
Measures from Study 1. As in Study 1, participants completed the
PANAS (Watson et al., 1988), the SWLS (Diener et al., 1985), and
Piedmont’s (1999) STS (with items endorsed on 7-point rather than 5-point
The Big Five. We measured the Big Five with Saucier’s (1994) Big
Five Mini-Markers scale. This scale consists of 40 adjectives that partici-
pants rate on a 9-point Likert type scale ranging from 1 (extremely
inaccurate)to9(extremely accurate) to indicate how well each of the
adjectives describes their personalities. The scale demonstrates adequate
reliability and validity as a measure of the Big Five (Saucier, 1994).
Disposition to forgive. To measure a novel aspect of prosociality, we
had participants complete 10 items developed for the present study that
assessed their disposition to forgive. Items were based on McCullough
Worthington, and Rachal’s (1997) theorizing regarding forgiveness (i.e.,
that forgiveness involves prosocial changes in avoidance, revenge, and
conciliatory motivations). Participants indicated the extent to which they
engaged in 10 different responses when people anger or hurt them, includ-
ing positively worded items (e.g., “I don’t hold it against him/her for long”)
and negatively worded items (e.g., “I will find a way to even the score”),
the latter of which we reverse scored. These 10 items had an internal
consistency of alpha ⫽ .81.
After participants read an informed consent form and indicated that they
understood and consented to the terms of the study, they completed the
survey. Participants indicated their age, gender, and ethnicity and then
completed the items on the scales mentioned above. Once participants had
completed the surveys, their anonymous responses were sent from the
Spirituality and Health web server to Michael E. McCullough by e-mail.
Confirmatory Factor Analysis
First, we sought to replicate the single-factor measurement
model for the GQ-6. Using the large sample size in this study to
our advantage, we split participants into two roughly equal sam-
ples on the basis of the day of the month on which they were born
(N ⫽ 609 for people born before the 16th of the month and N ⫽
619 for people born after the 15th of the month) and conducted two
independent tests of the one-factor measurement model. Although
the first sample showed a large chi-square,
(9, N ⫽
609) ⫽ 55.41, p ⬍ .001, other indices indicated good fit, CFI ⫽
.94, SRMR ⫽ .04. The second sample had similarly good fit,
N ⫽ 619) ⫽ 56.83, p ⬍ .001, CFI ⫽ .93, SRMR ⫽ .05. Therefore,
we concluded that a one-factor model provided good fit in this
Correlations of the Grateful Disposition With Other
The GQ-6 (M ⫽ 6.15, SD ⫽ .82) and the three-item gratitude
adjectives scale (M ⫽ 7.67, SD ⫽ 1.22) were correlated at
r(1182) ⫽ .65, p ⬍ .05. Using EQS 5.7b, we estimated that the
correlation of the two latent variables (i.e., corrected for measure-
ment error) was r ⫽ .75.
Table 3 shows the relationship of the GQ-6 and the gratitude
adjectives scale with other constructs. Both measures of the grate-
ful disposition were positively and moderately related to positive
affect, life satisfaction, spiritual transcendence, forgiveness,
Agreeableness, Conscientiousness, Extraversion, and Openness.
Both measures were negatively and moderately related to negative
affect and Neuroticism.
Regression of the Grateful Disposition on the Big Five
As in Study 1, we were interested in the extent to which each of
the Big Five predicted unique variance in the grateful disposition.
Table 4 shows the results of simultaneous multiple regression
analyses in which the two measures of the grateful disposition
were regressed on the Big Five. The Big Five variables predicted
28% of the variance in the GQ-6, R
(5, 1175) ⫽ .28, p ⬍ .01, with
all five predictors accounting for unique variance. Results were
similar for the gratitude adjectives scale, in which the Big Five
predicted 33% of the variance, R
(5, 1175) ⫽ .33, p ⬍ .01, with all
five predictors accounting for unique variance. The strongest
unique Big Five predictors for the GQ-6 were Neuroticism (
⫺.26) and Agreeableness (
⫽ .19), whereas the strongest unique
Big Five predictors of the adjective scale were Agreeableness (
.39) and Openness (
⫽ .17). Despite these differences, both
analyses revealed that the disposition toward gratitude is not
reducible to a linear combination of the Big Five, although, be-
cause of high statistical power, it was correlated significantly and
uniquely with all of the Big Five.
Study 2 yielded more evidence for the internal consistency and
one-factor structure of the GQ-6. As in Study 1, people who
indicated that they tended to experience gratitude—as measured
with both the GQ-6 and the gratitude adjectives scale—were
considerably higher in positive affectivity and well-being. More-
over, they reported being more forgiving when other people anger
or hurt them and more spiritually minded than did their less
grateful counterparts. Study 2 is particularly noteworthy because it
demonstrates that the correlates of the grateful disposition are
essentially the same in adults as in university students.
The consistent findings from Studies 1 and 2 led us to consider
two other potential correlates of the grateful disposition: materi-
alism and envy. We expected that the disposition toward gratitude
would be negatively related to materialism because grateful people
probably do not focus on acquiring and maintaining possessions
and wealth; rather, they would be expected to focus on savoring
the positive experiences and outcomes—both material and non-
material—that they have already experienced. Thus, we expect
that people could be both dispositionally grateful and strongly
materialistic only with great difficulty. Moreover, a focus on the
acquisition and consumption of material goods as a source of
happiness is to some extent incompatible with the strong orienta-
tion to other people that grateful people manifest through their
high levels of Agreeableness and other Agreeableness-related
In a related vein, gratitude is probably somewhat incompatible
with envy, which is a negative emotional state characterized by
resentment, inferiority, longing, and frustration about other peo-
ple’s material and nonmaterial successes (Parrott & Smith, 1993).
Grateful people, who tend to focus on the positive contributions of
others to their well-being, probably devote less attention to com-
paring their outcomes with those of other people and experience
Correlations Between Gratitude Scales
and Other Constructs (Study 2)
Positive affect .53** .57**
Negative affect ⫺.43** ⫺.23**
Life satisfaction .53** .38**
Forgiveness .36** .30**
Spiritual transcendence .53** .47**
Big Five personality traits
Agreeableness .41** .52**
Conscientiousness .26** .28**
Extraversion .32** .27**
Neuroticism ⫺.42** ⫺.31**
Openness .23** .33**
** p ⬍ .01.
MCCULLOUGH, EMMONS, AND TSANG
less envy as a result. In Study 3 we addressed these hypotheses and
adduced further evidence regarding the emotional, prosocial, and
spiritual correlates of the grateful disposition.
Participants were 156 undergraduate psychology students. They received
a small amount of course credit for their participation.
Measures from Studies 1 and 2. Participants completed several mea-
sures that were administered in Studies 1 and 2. These included the Big
Five Inventory (John et al., 1991), the BIDR (Paulhus, 1998), the Inter-
personal Reactivity Index (Davis & Oathout, 1987), Scheier et al.’s (1994)
LOT, Piedmont’s (1999) STS, and the set of single-item religious
Materialism. We measured materialism with two scales. The 18-item
Values-Oriented Materialism Scale (Richins & Dawson, 1992) comprises
three factorially distinct subscales. Acquisition centrality (i.e., the impor-
tance of material possessions) consists of seven items (e.g., “I enjoy
spending money on things that aren’t practical”). Happiness (i.e., viewing
material goods as essential for happiness and life satisfaction) consists of
five items (e.g., “I’d be happier if I could afford to buy more things”).
Possession-defined success (i.e., judging one’s own and others’ success by
number and quality of possessions accumulated) consists of seven items
(e.g., “I like to own things that impress people”). Richins and Dawson
reported test⫺retest reliabilities ranging from .82 to .87 and alpha reli-
abilities ranging from .71 to .88 for the subscales and the composite scale.
High scorers place more emphasis on financial security and less emphasis
on warm relationships with others compared with lower scorers. They also
are more likely to spend on themselves rather than on others and are less
likely to engage in voluntary simplicity behaviors (e.g., making items or
buying used goods rather than new ones, relying on bicycles for transpor-
The 23-item Belk Materialism Scale (Ger & Belk, 1990) measures four
factorially distinct dimensions of materialism. Possessiveness is measured
with seven items (i.e., “I get very upset if something is stolen from me,
even if it has little monetary value.”), nongenerosity is measured with six
items (e.g., “I don’t like to lend things, even to good friends.”), envy is
measured with five items (“I am bothered when I see people who buy
anything they want.”), and tangibilization (i.e., tendency to collect and
retain material possessions) is measured with five items (e.g., “I tend to
hang on to things I should probably throw out.”). Ger and Belk reported
internal consistency reliabilities ranging from alpha ⫽ .52–.67.
Envy. We measured envy with Smith et al.’s (1999) eight-item Dispo-
sitional Envy Scale. Items (e.g., “Frankly, the success of my neighbors
makes me resent them”) were rated on a 5-point Likert-type scale ranging
from 1 (strongly disagree)to5(strongly agree). Smith et al. reported
internal consistency reliability estimates in the range of alpha ⫽ .83 to .86
and 2-week test⫺retest stability of r ⫽ .80.
The questionnaires were distributed during a class session. Participants
were asked to complete the measures at home and to return them at the next
class session (2 days later).
Results and Discussion
Confirmatory Factor Analysis
We evaluated the one-factor solution for the GQ-6 with
EQS 5.7b to determine if the one-factor measurement model fit the
present data. Although chi-square was large,
(9, N ⫽
155) ⫽ 24.02, p ⬍ .01, the CFI and the SRMR suggested a good
fit (CFI ⫽ .95, SRMR ⫽ .05).
Correlates of the Disposition Toward Gratitude
Participants in Study 3 showed levels of dispositional gratitude
(GQ-6) similar to those found in Studies 1 and 2 (M ⫽ 5.82, SD ⫽
.86). Table 5 shows the correlations between the GQ-6 and the
other measures. As in Studies 1 and 2, gratitude was related to
measures of positive affect (e.g., optimism), prosociality (i.e.,
empathy), and spirituality/religiousness. As in Study 1, gratitude
was positively correlated with SDE (r ⫽ .34, p ⬍ .01) and IM (r ⫽
.29, p⬍ .01) subscales of the social desirability inventory.
Materialism and Envy
The correlations between the GQ-6 and the materialism sub-
scales, as shown in Table 6, were uniformly negative. Grateful
people report themselves as being less materialistic and less envi-
ous. In particular, grateful people report being more willing to part
with their possessions, more generous with them, less envious of
the material wealth of others, less committed to the idea that
material wealth is linked with success in life, and less convinced of
the idea that material wealth brings happiness. Apparently, mate-
rial success is not a very important factor in the happiness of
highly grateful people. Also, their lower scores on Smith et al.’s
(1999) measure of dispositional envy suggests that grateful people
experience less frustration and resentment over the achievements
and possessions of other people (r ⫽⫺.39). These results suggest
that it may indeed be difficult for one to be grateful and materi-
alistic or envious at the same time because these states require
directing one’s attention toward such divergent concerns.
The Big Five
Gratitude was significantly and positively correlated with
Agreeableness, Extraversion, and Conscientiousness and nega-
Multiple Regression of Two Measures of Gratitude
on the Big Five (Study 2)
Predictor variable BSE
Gratitude Questionnaire–6 .28**
Openness .08 .02 .11**
Conscientiousness .04 .02 .07*
Extraversion .09 .02 .16**
Agreeableness .16 .03 .19**
Neuroticism ⫺.14 .02 ⫺.26**
Gratitude adjectives scale .33**
Openness .19 .03 .17**
Conscientiousness .08 .02 .09**
Extraversion .09 .02 .10**
Agreeableness .50 .04 .39**
Neuroticism ⫺.05 .02 ⫺.06*
* p ⬍ .05. ** p ⬍ .01.
tively correlated with Neuroticism. Results of a simultaneous
multiple regression analysis using the Big Five to predict gratitude
are presented in Table 6. The Big Five predicted 23% of the
variance in the GQ-6, R
(5, 149) ⫽ .23, p ⬍ .01. As in Study 1,
⫽ .30, p ⬍ .01), Neuroticism (
⫽⫺.19, p ⬍
.05) and Extraversion (
⫽ .18, p ⬍ .05) predicted significant
amounts of unique variance in gratitude, whereas Openness and
Conscientiousness did not.
In Studies 1 through 3, we presented evidence that the disposi-
tion toward gratitude is associated with tendencies toward positive
emotions and well-being, prosocial traits, and spirituality. How-
ever, given the consistent correlations of the disposition toward
gratitude with Extraversion/positive affectivity, Neuroticism/neg-
ative affectivity, and Agreeableness, it would be useful to deter-
mine the extent to which the correlations of the GQ-6 with affec-
tive, prosocial, and spiritual constructs exist independently of their
mutual associations with these three clusters of higher order traits.
It would also be useful to know whether these associations exist
independently of the effects of social desirability.
To address these issues, we evaluated the extent of reduction in these
associations when Extraversion/positive affectivity, Neuroticism/negative
affectivity, Agreeableness, and social desirability are controlled. We meta-
analytically combined the relevant zero-order and partial correlations from
the data gathered for Studies 1 through 3. We combined coefficients with
the DSTAT statistical software (Johnson, 1989). Using DSTAT, we per-
formed Fisher’s (1928) r-to-z transformations on the coefficients and
weighted them by the inverse of their sampling error variance. Then, we
calculated weighted mean effect sizes and converted these mean effect
sizes back to the Pearson r metric.
Results and Discussion
Column 2 of Table 7 indicates the mean zero-order correlation
of the GQ-6 with each of the measures denoted in the row headings
using the data from Studies 1 through 3. For the sake of simplicity,
we used sum scores of the religious variables measured in Stud-
ies 1 and 3 instead of the single-item measures of religiousness as
individual criterion variables.
Mean Correlations With Extraversion/Positive Affectivity
The coefficients in Column 3 of Table 7 are mean partial
correlations of the GQ-6 with the measures denoted in the row
headings while controlling for Extraversion and (in the case of data
from Studies 1 and 2) positive affectivity. (Positive and negative
affectivity were not measured in Study 3. For partial correlations
from Study 3, we controlled Extraversion only.) Notably, the
correlations with several of the affective and spiritual variables
(e.g., life satisfaction, hope, self-transcendence, and spiritual trans-
cendence) were substantially reduced in magnitude (i.e., 10 cor-
relation points or more). With only one exception (the correlation
with self-transcendence), all correlations maintained their valence
and remained statistically significant. Thus, gratitude still accounts
for variance in nearly all of these measures of affect/well-being,
prosociality, and spirituality after controlling for Extraversion/
Mean Correlations With Neuroticism/Negative Affectivity
The coefficients in Column 4 of Table 7 represent the correla-
tions of the GQ-6 with the measures denoted in the row headings
after controlling for Neuroticism and (in the case of data from
Multiple Regression of the Gratitude Questionnaire–6
on the Big Five (Study 3)
Predictor variable BSE
Openness .02 .10 .01
Conscientiousness .14 .11 .10
Extraversion .20 .08 .18*
Agreeableness .39 .11 .30**
Neuroticism ⫺.23 .09 ⫺.19*
⫽ .23, p ⬍ .01.
* p ⬍ .05. ** p ⬍ .01.
Correlations Between the Gratitude Questionnaire–6 (GQ-6)
and Other Constructs (Study 3)
Scale Correlation with GQ-6
Optimism (LOT) .43**
Empathic concern .27**
Perspective taking .36**
Spiritual transcendence (STS) .30**
Importance of religion .22**
Frequency of attendance .15†
Religious friends .10
Reading scripture .18*
Other religious reading .09
Personal relationship with God .29**
Union with God .29**
Success (Richins) ⫺.25**
Centrality (Richins) ⫺.07
Happiness (Richins) ⫺.38**
Possessiveness (Belk) ⫺.17*
Nongenerosity (Belk) ⫺.35**
Envy (Belk) ⫺.34**
Tangibilization (Belk) ⫺.11
Envy (DES) ⫺.39**
Impression management .29**
Note. LOT ⫽ Life Orientation Scale; STS ⫽ Spiritual Transcendence
Scale; Richins ⫽ Richins Materialism Scale; Belk ⫽ Belk Materialism
Scale; DES ⫽ Dispositional Envy Scale.
† p ⬍ .10. * p ⬍ .05. ** p ⬍ .01.
MCCULLOUGH, EMMONS, AND TSANG
Studies 1 and 2) negative affectivity. Notably, the correlations with
several of the affective and spiritual variables (e.g., life satisfac-
tion, hope, optimism, anxiety, depression, envy, forgiveness, and
materialism) were substantially reduced (i.e., 10 correlation points
or more). The only correlations that did not maintain their valence
and statistical significance were the correlations with anxiety and
the possessiveness subscale of the Belk Materialism Scale. Thus,
gratitude still predicts significant variance in nearly all of these
measures of affect/well-being, prosociality, and spirituality after
controlling for Neuroticism/negative affectivity.
Mean Correlations With Agreeableness Held Constant
The coefficients in Column 5 of Table 7 represent the correla-
tions of the GQ-6 with the measures denoted in the row headings
after controlling for Agreeableness. Notably, correlations with
several prosocial variables (empathy, forgiveness, several materi-
alism subscales) were substantially reduced in magnitude (i.e., 10
correlation points or more). The only correlations that did not
maintain their valence and statistical significance were with the
possessiveness and nongenerosity subscales of the Belk Material-
ism Scale. Thus, gratitude accounts for unique variance in nearly
all of these criteria after controlling for Agreeableness.
Mean Correlations With Social Desirability Held
Finally, the coefficients in Column 6 of Table 7 represent the
correlations with the measures denoted in the row headings after
controlling for social desirability. The correlations with measures
of several affective, prosocial, and spiritual variables (life satis-
faction, envy, perspective-taking, spiritual transcendence, the envy
subscale of the Belk Materialism Scale, and the Success and
Happiness subscales of the Richins Materialism Scale) were re-
duced substantially (i.e., 10 correlation points or more). However,
only the correlation with the Success subscale failed to maintain
statistical significance. Thus, although socially desirable respond-
ing may explain some of our obtained associations, the control of
social desirability does not affect drastically our general conclu-
sions about the correlates of gratitude.
Correlations of the Gratitude Questionnaire–6 With Other Constructs Before and After Controlling for Extraversion/Positive
Affectivity (E/PA), Neuroticism/Negative Affectivity (N/NA), Agreeableness, and Social Desirability (Studies 1–3)
Scale Zero-order correlation Controlling E/PA Controlling N/NA Controlling Agreeableness Controlling social desirability
Measures of affect and well-being
Life satisfaction .51**
Measures of prosociality
Empathic concern .28**
Possessiveness (Belk) ⫺.17*
Nongenerosity (Belk) ⫺.35**
Envy (Belk) ⫺.34**
Tangibilization (Belk) ⫺.11
Success (Richins) ⫺.25**
Centrality (Richins) ⫺.07
Happiness (Richins) ⫺.38**
Measures of spirituality/religiousness
Spiritual transcendence .47**
Note. Belk ⫽ Belk Materialism Scale; Richins ⫽ Richins Materialism Scale.
Correlations combined from Studies 1 and 2, total N ⫽ approximately 1,372.
Correlation from Study 1 only, N ⫽ approximately 205.
combined from Studies 1 and 3, total N ⫽ approximately 340.
Correlation from Study 3 only, N ⫽ approximately 135.
Correlation from Study 2 only,
N ⫽ approximately 1,167.
Correlations from Studies 1, 2, and 3, total N ⫽ approximately 1,507.
Sum of the single-item measures of religiousness
used in Studies 1–3.
† p ⬍ .10. * p ⬍ .05. ** p ⬍ .01.
The results from the present studies help to paint a portrait of the
dispositionally grateful person. Consistent with our hypotheses,
grateful people appear to be different from their less grateful
counterparts in three interesting psychological domains: (a)
emotionality/well-being, (b) prosociality, and (c) spirituality/reli-
giousness. Compared with their less grateful counterparts, grateful
people are higher in positive emotions and life satisfaction and also
lower in negative emotions such as depression, anxiety, and envy.
They also appear to be more prosocially oriented in that they are
more empathic, forgiving, helpful, and supportive than are their
less grateful counterparts. Relatedly, grateful people are less fo-
cused on the pursuit of materialistic goals. Finally, people with
stronger dispositions toward gratitude tend to be more spiritually
and religiously minded. Not only do they score higher on measures
of traditional religiousness, but they also score higher on nonsec-
tarian measures of spirituality that assess spiritual experiences
(e.g., sense of contact with a divine power) and sentiments (e.g.,
belief that all living things are interconnected) independent of
specific theological orientation.
We attempted to account for the emotional, prosocial, and
spiritual traits of grateful people by appealing to the Big Five
personality taxonomy as an explanatory framework. Disposition-
ally grateful people were consistently more extraverted, more
agreeable, and less neurotic than their less grateful counterparts.
When these three superordinate traits were controlled, many of the
correlations between the disposition toward gratitude and mea-
sures of emotions/well-being, prosociality, and spirituality became
smaller in magnitude.
However, none of these three superordinate traits could com-
pletely account for the correlations of the Big Five with lower
order personality variables. Moreover, although the correlations
with the Big Five were robust, the Big Five only accounted for
approximately 30% of the variance in the disposition toward
gratitude. Even if one were to correct the obtained associations for
measurement error, the Big Five still would account for no more
than 40% to 45% of the variance in the disposition toward grati-
tude, so the disposition toward gratitude is by no means reducible
to a linear combination of them (see also Saucier & Goldberg,
1998). Nevertheless, the present findings regarding the correla-
tions with the Big Five could be useful in formulating a broader
theory of gratitude that accounts for its roots in basic personality
traits. Specifically, it might be fruitful to conceptualize the dispo-
sition toward gratitude in part as a characteristic adaptation that is
preferentially deployed by highly extraverted, minimally neurotic,
and highly agreeable people for navigating their worlds (see Mc-
Crae & Costa, 1999).
Measuring the Grateful Disposition
The grateful disposition can be measured dependably through
self-report and informant report. We developed a six-item self-
report measure that assesses individual differences in gratitude. It
converges well with observer reports and an adjective rating scale
of gratitude like the one used by Saucier and Goldberg (1998). The
GQ-6 has excellent psychometric properties, including a robust
one-factor structure and high internal consistency, especially in
light of its brevity. Moreover, it correlates in theoretically expected
ways with a variety of affective, prosocial, and spiritual constructs.
Therefore, the GQ-6 and its observer form (which we intend to
continue developing) may be of use in future research on gratitude.
Directions for Future Research
McCullough et al. (2001) recently proposed several directions
for future research on gratitude, including research on psychomet-
rics, the role of gratitude in motivating reciprocity, the relations
between gratitude and well-being, and the relations between grat-
itude and spirituality. The present studies addressed several of
these issues, but also suggested ways that these questions could be
refined. The present findings raised several new issues as well.
Is Gratitude a Unique Emotion?
The present results indicate that gratitude as an affective trait is
related but distinct from other traitlike measures of specific emo-
tions (e.g., dispositional happiness, vitality, optimism, hope, de-
pression, anxiety, and envy), but is gratitude distinct at the level of
discrete emotional experience? Some research suggests that it is
(Ellsworth & Smith, 1988; Schimmack & Reisenzein, 1997). For
example, Ellsworth and Smith found that the adjectives loving,
grateful, friendly, and admiring formed a cluster orthogonal to a
happiness/elation/contentment cluster. Gratitude also possesses a
unique pattern of attributions that distinguish it from positive
emotions such as happiness and contentment (e.g., Weiner, Rus-
sell, & Lerman, 1979).
However, questions remain regarding the distinctiveness of grat-
itude. Although theorists have suggested that gratitude may have
unique functions—particularly in the realm of reciprocal altruism
and prosocial behavior (de Waal, 2000; Fredrickson, 2000; Mc-
Cullough et al., 2001; Nesse, 1990; Trivers, 1971)—researchers
have not investigated thoroughly whether gratitude motivates
prosocial behavior in such contexts over and above the effects of
positive emotions generally (see Carlson, Charlin, & Miller, 1988).
In addition, no evidence indicates that discrete episodes of grati-
tude are accompanied by particular patterns of physiological
arousal or a unique facial display. Examining whether gratitude
possesses a unique pattern of action tendencies, physiological
arousal, or facial displays would help in assessing the uniqueness
The Grateful Emotions and Attributions of Dispositionally
Following Rosenberg’s (1998) hierarchical model of affective
phenomena, we speculated that people with a strong disposition
toward gratitude possess a low threshold for the experience of
grateful emotions in daily life. This lowered threshold for gratitude
might consist of several facets: (a) more intense experiences of
gratitude, (b) more frequent experiences of gratitude, (c) wider
gratitude spans (i.e., feeling grateful for a greater number of
specific circumstances in one’s life), and (d) denser gratitude
episodes (i.e., attributing each positive outcome to a greater num-
ber of people). Relatedly, we have speculated about the general
attributional styles of highly grateful people: Their attributional
styles may be characterized not by a tendency to attribute their
positive outcomes to external sources, but rather by a tendency to
MCCULLOUGH, EMMONS, AND TSANG
incorporate a wide variety of people who contribute to their
positive outcomes. Exploring these notions about the emotional
experience and attributions of dispositionally grateful people could
help to elucidate the affective and cognitive mechanisms that
constitute the grateful personality.
Health, Well-Being, and Coping
Finally, given the consistent connections of gratitude to affec-
tive traits and well-being, it might be useful to explore whether
gratitude actually promotes well-being. Pondering the circum-
stances in one’s life for which one is grateful appears to be a
common way of coping with both acute and chronic stressful life
events (e.g., Barusch, 1997; Coffman, 1996), perhaps in a manner
akin to the “benefit-finding” described by Affleck and Tennen
(1996). Moreover, experimental research suggests that discrete
experiences of gratitude and appreciation may cause increases in
parasympathetic myocardial control (McCraty, Atkinson, Tiller,
Rein, & Watkins, 1995), as well as improvements in more molar
aspects of health (Emmons & McCullough, 2000). These prelim-
inary findings may provide the impetus for more detailed investi-
gations of the possibility that gratitude plays a role in health and
well-being, perhaps even as a mediator of the robust association
between religiousness and physical health (for review see McCul-
lough, Hoyt, Larson, Koenig, & Thoresen, 2000).
From the age of Seneca to the present day, scholars in the
humanities have grappled with the nature of gratitude as a human
disposition and affective state. However, gratitude has escaped
systematic attention by psychologists almost completely (Emmons
& Crumpler, 2000; McCullough et al., 2001). The goal of the
present article was to present and test some basic propositions
about the nature of the disposition toward gratitude and its rela-
tionships to affect and well-being, prosociality, spirituality, and the
Big Five. Our findings, though intriguing, are but an initial look at
the disposition toward gratitude. Hopefully, future researchers can
make use of the insights and measurement tools described in the
present article to extend psychology’s understanding of gratitude
and its roles in individual and interpersonal functioning.
Abramson, L. Y., Seligman, M. E. P., & Teasdale, J. (1978). Learned
helplessness in humans: Critique and reformulation. Journal of Abnor-
mal Psychology, 87,49–74.
Affleck, G., & Tennen, H. (1996). Construing benefits from adversity:
Adaptational significance and dispositional underpinnings. Journal of
Personality, 64, 899–922.
Allport, G. W., Gillespie, J. M., & Young, J. (1948). The religion of the
post-war college student. Journal of Psychology, 25,3–33.
Barusch, A. S. (1997). Self-concepts of low-income older women: Not old
or poor, but fortunate and blessed. International Journal of Aging and
Human Development, 44, 269–282.
Batson, C. D. (1991). The altruism question. Hillsdale, NJ: Erlbaum.
Bentler, P. M. (1995). EQS structural equations program manual. Encino,
CA: Multivariate Software.
Carlson, M., Charlin, V., & Miller, N. (1988). Positive mood and helping
behavior: A test of six hypotheses. Journal of Personality and Social
Psychology, 55, 211–229.
Carman, J. B., & Streng, F. J. (Eds.). (1989). Spoken and unspoken thanks:
Some comparative soundings. Dallas, TX: Center for World Thanksgiv-
Cloninger, C. R., Svrakic, D. M., & Przybeck, T. R. (1993). A psychobi-
ological model of temperament and character. Archives of General
Psychiatry, 50, 975–990.
Coffman, S. (1996). Parents’ struggles to rebuild family life after Hurricane
Andrew. Issues in Mental Health Nursing, 17, 353–367.
Davis, M. H., & Oathout, H. A. (1987). Maintenance of satisfaction in
romantic relationships: Empathy and relational competence. Journal of
Personality and Social Psychology, 53, 397–410.
Derogatis, L. R., & Spencer, P. M. (1982). The Brief Symptom Inventory:
Administration, scoring, and procedures manual. Baltimore: Clinical
Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The
Satisfaction With Life Scale. Journal of Personality Assessment, 49,
de Waal, F. (2000). Payment for labour in monkeys. Nature, 404, 563.
Ellsworth, P. C., & Smith, C. A. (1988). Shades of joy: Patterns of
appraisal differentiating pleasant emotions. Cognition and Emotion, 2,
Emmons, R. A., & Crumpler, C. A. (2000). Gratitude as human strength:
Appraising the evidence. Journal of Social and Clinical Psychology, 19,
Emmons, R. A., & McCullough, M. E. (2000). Counting blessings versus
burdens: An experimental investigation of gratitude and subjective well-
being in daily life. Manuscript submitted for publication.
Fisher, R. A. (1928). Statistical methods for research workers (2nd ed.).
London: Oliver & Boyd.
Fredrickson, B. L. (2000, March 7). Cultivating research on positive
emotions. Prevention and Treatment, 3, Article 0001a. Retrieved
October 4, 2001, from http://journals.apa.org/prevention/volume3/
Funder, D. C., & Colvin, C. R. (1997). Congruence of others’ and self-
judgments of personality. In R. Hogan, J. Johnson, & S. Briggs (Eds.),
Handbook of personality psychology (pp. 617–647). San Diego, CA:
Ger, G., & Belk, R. W. (1990). Measuring and comparing materialism
cross-culturally. Advances in Consumer Research, 17, 186–192.
Gladstone, T. R. G., & Kaslow, N. J. (1995). Depression and attributions
in children and adolescents: A meta-analytic review. Journal of Abnor-
mal Child Psychology, 23, 597–606.
Harpham, E. J. (2000, October). The problem of gratitude in Adam Smith.
Paper presented at the Kindling the Science of Gratitude Conference,
Hoyle, R. H., & Panter, A. T. (1995). Writing about structural equation
models. In R. H. Hoyle (Ed.), Structural equation modeling: Concepts,
issues, and applications (pp. 158–176). Thousand Oaks, CA: Sage.
Hoyt, W. T., & Melby, J. N. (1999). Dependability of measurement in
counseling: An introduction to generalizability theory. The Counseling
Psychologist, 43, 430–447.
Hu, L., & Bentler, P. M., (1998). Fit indices in covariance structure
modeling: Sensitivity to underparameterized model misspecification.
Psychological Methods, 3, 424–453.
John, O. P., Donahue, E. M., & Kentle, R. L. (1991). The Big Five
Inventory—Versions 4a and 54. Berkeley, CA: University of California,
Berkeley, Institute of Personality and Social Research.
John, O. P., & Srivastava, S. (1999). The Big Five trait taxonomy: History,
measurement, and theoretical perspectives. In L. A. Pervin and O. P.
John (Eds.), Handbook of personality: Theory and research (2nd ed., pp.
102–138). New York: Guilford.
Johnson, B. T. (1989). DSTAT: Software for the meta-analytic review of
research literatures. Hillsdale, NJ: Erlbaum.
Keltner, D., & Buswell, B. N. (1996). Evidence for the distinctiveness of
embarrassment, shame, and guilt: A study of recalled antecedents and
facial expressions of emotion. Cognition and Emotion, 10, 155–171.
Kirk, K. M., Eaves, L. J., & Martin, N. G. (1999). Self-transcendence as a
measure of spirituality in a sample of older Australian twins. Twin
Koole, S. L., Jager, W., van den Berg, A. E., Vlek, C. A. J., & Hofstee,
W. K. B. (2001). On the social nature of personality: Effects of extra-
version, agreeableness, and feedback about collective resource use on
cooperation in a resource dilemma. Personality and Social Psychology
Bulletin, 27, 289–301.
Lazarus, R. S., & Lazarus, B. N. (1994). Passion and reason: Making sense
of our emotions. New York: Oxford University Press.
Lupfer, M. B., De Paola, S. J., Brock, K. F., & Clement, L. (1994). Making
secular and religious attributions: The availability hypothesis revisited.
Journal for the Scientific Study of Religion, 33, 162–171.
Lupfer, M. B., Tolliver, D., & Jackson, M. (1996). Explaining life-altering
occurrences: A test of the “God-of-the-gaps” hypothesis. Journal for the
Scientific Study of Religion, 35, 379–391.
Lyubomirsky, S., & Lepper, H. S. (1999). A measure of subjective hap-
piness: Preliminary reliability and construct validation. Social Indicators
Research, 46, 137–155.
Mayer, J. D., Salovey, P., Gomberg-Kaufman, S., & Blainey, K. (1991). A
broader conceptualization of mood experience. Journal of Personality
and Social Psychology, 60, 100–111.
McCrae, R. R., & Costa, P. T., Jr. (1999). A five-factor theory of person-
ality. In L. A. Pervin & O. P. John (Eds.), Handbook of personality:
Theory and research (2nd ed.; pp. 139–153). New York: Guilford.
McCraty, R., Atkinson, M., Tiller, W. A., Rein, G., & Watkins, A. D.
(1995). The effects of emotions on short-term power spectrum analysis
of heart rate variability. American Journal of Cardiology, 76, 1089–
McCullough, M. E., Hoyt, W. T., Larson, D. B., Koenig, H. G., &
Thoresen, C. E. (2000). Religious involvement and mortality: A meta-
analytic review. Health Psychology, 19, 211–222.
McCullough, M. E., Kilpatrick, S. D., Emmons, R. A., & Larson, D. B.
(2001). Is gratitude a moral affect? Psychological Bulletin, 127, 249–
McCullough, M. E., & Tsang, J. (in press). Parent of the virtues? The
prosocial contours of gratitude. In R. A. Emmons & M. E. McCullough
(Eds.), The psychology of gratitude. New York: Oxford University
McCullough, M. E., Worthington, E. L., Jr., & Rachal, K. C. (1997).
Interpersonal forgiving in close relationships. Journal of Personality and
Social Psychology, 73, 321–336.
Murphy, K. R., Jako, R. A., & Anhalt, R. L. (1993). Nature and conse-
quences of halo error: A critical analysis. Journal of Applied Psychol-
ogy, 78, 218–255.
Nesse, R. M. (1990). Evolutionary explanations of emotions. Human
Nature, 1, 261–289.
Ortony, A., Clore, G. L., & Collins, A. (1988). The cognitive structure of
emotions. New York: Cambridge University Press.
Parrott, W. G., & Smith, R. H. (1993). Distinguishing the experiences of
envy and jealousy. Journal of Personality and Social Psychology, 64,
Paulhus, D. L. (1998). Interpersonal and intrapsychic adaptiveness of trait
self-enhancement: A mixed blessing? Journal of Personality and Social
Psychology, 74, 1197–1208.
Paulhus, D. L., & Reynolds, S. (1995). Enhancing target variance in
personality perceptions: Highlighting the person in person perception.
Journal of Personality and Social Psychology, 69, 1233–1242.
Pavot, W., & Diener, E. (1993). Review of the Satisfaction With Life
Scale. Psychological Assessment, 5, 164–172.
Piedmont, R. L. (1999). Does spirituality represent the sixth factor of
personality? Spiritual transcendence and the five-factor model. Journal
of Personality, 67, 985–1013.
Richins, M. L., & Dawson, S. (1992). A consumer values orientation for
materialism and its measurement: Scale development and validation.
Journal of Consumer Research, 19, 303–315.
Roberts, R. C. (2000, October). The blessings of gratitude: Some concep-
tual remarks. Paper presented at the Kindling the Science of Gratitude
Conference, Dallas, TX.
Rosenberg, E. L. (1998). Levels of analysis and the organization of affect.
Review of General Psychology, 2, 247–270.
Rozin, P., Lowery, L., Imada, S., & Haidt, J. (1999). The CAD triad
hypothesis: A mapping between three moral emotions (contempt, anger,
disgust) and three moral codes (community, autonomy, divinity). Jour-
nal of Personality and Social Psychology, 76, 574–586.
Ryan, R. M., & Frederick, C. (1997). On energy, personality, and health:
Subjective vitality as a dynamic reflection of well-being. Journal of
Personality, 65, 529–565.
Saucier, G. (1994). Mini-markers: A brief version of Goldberg’s unipolar
big-five markers. Journal of Personality Assessment, 63, 506–516.
Saucier, G., & Goldberg, L. R. (1998). What is beyond the Big Five?
Journal of Personality, 66, 495–524.
Scheier, M. F., Carver, C. S., & Bridges, M. W. (1994). Distinguishing
optimism from neuroticism (and trait anxiety, self-mastery, and self-
esteem): A reevaluation of the Life Orientation Test. Journal of Person-
ality and Social Psychology, 67, 1063–1078.
Schimmack, U., & Reisenzein, R. (1997). Cognitive processes involved in
similarity judgments of emotions. Journal of Personality and Social
Psychology, 73, 645–661.
Shrout, P. E., & Fleiss, J. L. (1979). Intraclass correlations: Uses in
assessing rater reliability. Psychological Bulletin, 86, 420–428.
Smith, R. H., Parrott, W. G., Diener, E., Hoyle, R. H., & Kim, S. H. (1999).
Dispositional envy. Personality and Social Psychology Bulletin, 25,
Snyder, C. R. (2000). The past and possible futures of hope. Journal of
Social and Clinical Psychology, 19,11–28.
Snyder, C. R., Harris, C., Anderson, J. R., Holleran, S. A., Irving, L. M.,
Sigmon, S. T., Yoshinobu, L., Gibb, J., Langelle, C., & Harney, P.
(1991). The will and the ways: Development and validation of an
individual-differences measure of hope. Journal of Personality and
Social Psychology, 60, 570–585.
Tangney, J. P. (1991). Moral affect: The good, the bad, and the ugly.
Journal of Personality and Social Psychology, 61, 598–607.
Thorndike, E. L. (1920). A constant error in psychological ratings. Journal
of Applied Psychology, 4,25–29.
Trivers, R. L. (1971). The evolution of reciprocal altruism. Quarterly
Review of Biology, 46,35–57.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and vali-
dation of brief measures of positive and negative affect: The PANAS
scales. Journal of Personality and Social Psychology, 54, 1063–1070.
Weiner, B. (1986). An attributional theory of motivation and emotion. New
Weiner, B. (1995). Judgments of responsibility. New York: Guilford.
Weiner, B., Russell, D., & Lerman, D. (1979). The cognition–emotion
process in achievement-related contexts. Journal of Personality and
Social Psychology, 37, 1211–1220.
MCCULLOUGH, EMMONS, AND TSANG
The Gratitude Questionnaire–6 (GQ-6)
Using the scale below as a guide, write a number beside each statement
to indicate how much you agree with it.
1 ⴝ strongly disagree
2 ⴝ disagree
3 ⴝ slightly disagree
4 ⴝ neutral
5 ⴝ slightly agree
6 ⴝ agree
7 ⴝ strongly agree
____1. I have so much in life to be thankful for.
____2. If I had to list everything that I felt grateful for, it would be a very
____3. When I look at the world, I don’t see much to be grateful for.
____4. I am grateful to a wide variety of people.
____5. As I get older I find myself more able to appreciate the people,
events, and situations that have been part of my life history.
____6. Long amounts of time can go by before I feel grateful to
something or someone.
Items 3 and 6 are reverse scored.
Received November 19, 2000
Revision received July 28, 2001
Accepted July 30, 2001 䡲