Journal of Adolescence 32 (2009) 633e650

www.elsevier.com/locate/jado

Gratitude and subjective well-being in early adolescence:

Examining gender differences
Jeffrey J. Froh a,*, Charles Yurkewicz b, Todd B. Kashdan c
a
Hofstra University, Department of Psychology, 210 Hauser Hall, Hempstead, NY 11549, USA
b
St. John’s University, Department of Psychology, SB36 Marillac Hall, 8000 Utopia Parkway, Jamaica, NY 11439, USA
c
George Mason University, Department of Psychology, Mail Stop 3F5, Fairfax, VA 22030, USA

Abstract

Gratitude was examined among 154 students to identify benefits from its experience and expression.
Students completed measures of subjective well-being, social support, prosocial behavior, and physical
symptoms. Positive associations were found between gratitude and positive affect, global and domain
specific life satisfaction, optimism, social support, and prosocial behavior; most relations remained even
after controlling for positive affect. Gratitude demonstrated a negative relation with physical symptoms,
but not with negative affect. Relational fulfillment mediated the relation between gratitude and physical
symptoms. Gratitude demonstrated strong relations with the following positive affects: proud, hopeful,
inspired, forgiving, and excited. The relation between gratitude and family support was moderated by
gender, indicating that boys, compared with girls, appear to derive more social benefits from gratitude.
Strengths, limitations, and implications are discussed.
Ó 2008 The Association for Professionals in Services for Adolescents. Published by Elsevier Ltd. All rights
reserved.

Keywords: Gratitude; Emotion; Subjective well-being; Life satisfaction; Happiness; Prosocial behavior; Gender; Posi-
tive psychology; Adolescence

* Corresponding author. Tel.: þ1 516 463 4027; fax: þ1 516 463 6052.
E-mail addresses: jeffrey.froh@hofstra.edu, http://people.hofstra.edu/jeffrey_j_froh/ (J.J. Froh), charlesyurkewicz@
yahoo.com (C. Yurkewicz), tkashdan@gmu.edu, http://mason.gmu.edu/~tkashdan/kashdan.html (T.B. Kashdan).

0140-1971/$30.00 Ó 2008 The Association for Professionals in Services for Adolescents. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.adolescence.2008.06.006
634 J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650

The goodness of the fairy tale was not affected by the fact that there might be more dragons
than princesses; it was good to be in a fairy tale. The test of all happiness is gratitude.
G.K. Chesterton
If the only prayer you ever say in your entire life is ‘‘Thank you,’’ it will be enough.
Meister Eckhardt
Psychologists have recently emphasized the need for promoting adolescent well-being, beyond the
existing focus on symptom reduction. Mitigating pathology is important, but its absence is different
from mental health (Keyes, 2007). The traditional approachdidentifying and fixing weaknessesd
may be limited in fostering the ‘‘good life’’ (Sheldon & King, 2001). Therefore, psychologists should
consider complementing existing practices by identifying and augmenting strengths, like gratitude.
There are two useful theories in studying gratitude. The first is Fredrickson’s (1998, 2001)
broaden-and-build theory of positive emotions. Negative emotions narrow our focus and restrict
our behavioral range. Positive emotions, however, yield nonspecific action tendencies beyond
physical action. The theory asserts that positive emotions generate broad thought-action reper-
toires that ultimately build durable physical, intellectual, and social resources. A meta-analysis
by Lyubomirsky, King, and Diener (2005), aggregating over 300 studies, suggests that success
engenders positive emotionsdbut also that positive emotions engender success. Indeed, happy
people tend to live longer, make more money, and enjoy enduring loving relationships. One
reason positive emotions (e.g., happiness) might cause success could be because of the durable
resourcesdphysical, intellectual, and socialdbuilt over time. These resources can then be tapped
into during times of adversity, as well as in times of growth.
The second theory describes gratitude as a moral emotion with three essential functions
(McCullough, Kilpatrick, Emmons, & Larson, 2001). As a moral barometer, gratitude signals
the beneficiary that a benefactor has bestowed a gift upon him. As a moral motive, gratitude
encourages prosocial behavior in the beneficiary either directly toward the benefactor or others.
Finally, as a moral reinforcer, gratitude increases the probability that the benefactor will act pro-
socially toward the beneficiary in the future. According to this conceptualization, by experiencing
gratitude, a person is motivated to carry out prosocial behavior, energized to sustain moral
behaviors, and inhibited from committing destructive interpersonal behaviors.

The psychological benefits of gratitude

Gratitude stems from the perception that one has experienced a positive outcome intentionally
provided by another person or ‘‘moral agent,’’ often but not necessarily a person (Emmons
& McCullough, 2003). The object of gratitude is other-directed to persons, or to impersonal
(nature) or non-human sources (God, fate, the cosmos). Gratitude may be defined as ‘‘a sense
of thankfulness and joy in response to receiving a gift, whether the gift be a tangible benefit
from a specific other or a moment of peaceful bliss evoked by natural beauty’’ (Emmons, 2004,
p. 554). As an emotion, grateful states result from recognizing that (a) one has obtained a positive
outcome; and (b) there is an external source for this positive outcome. Recognizing that the bene-
factor has expended effort to give them a gift further amplifies grateful feelings; for this reason,
gratitude is considered an empathic emotion (Lazarus & Lazarus, 1994). Because children develop
 J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650 635

a theory of mind at 3e4 years of age, they are able to understand that human behavior is inten-
tional and feel empathy (Leslie, 1987). Therefore, children seem capable of experiencing gratitude
and its function as a moral emotion (e.g., prosocial behavior).
Recent experimental research has demonstrated that gratitude causes prosocial behavior,
demonstrating its function as a moral motive (McCullough et al., 2001). Gratitude often causes
direct reciprocity, leading individuals to respond prosocially to a benefactor (Bartlett & DeSteno,
2006; Tsang, 2006, 2007); and it can cause upstream reciprocity, leading them to act prosocially
toward others (Bartlett & DeSteno, 2006). Grateful individuals may act prosocially as a way of
expressing their gratitude; however, over time these actions can enhance social relationships
(Emmons & Shelton, 2002; Harpham, 2004; Komter, 2004). Indeed, gratitude helps build trust
in social relationships (Dunn & Schweitzer, 2005). Thus, gratitude may maintain and build
resources of social support (Fredrickson, 2004).
Fredrickson’s (1998, 2001) broaden-and-build theory of positive emotions suggests that grati-
tude may also help individuals build other durable resources for well-being. Specifically, it may
nurture creativity, intrinsic motivation, purposefulness (Froh & Bono, in press), and spark an
upward spiral of positive emotions and outcomes. This may explain why grateful people tend to
be higher in vitality, optimism, religiousness, spirituality, (McCullough, Emmons, & Tsang,
2002), well-being (Emmons & Shelton, 2002; Watkins, Van Gelder, & Frias, in press) and relation-
ship quality (Algoe, 2006) and lower in negative affect (McCullough et al., 2002; Watkins, Wood-
ward, Stone, & Kolts, 2003) and physical symptoms (Emmons & McCullough, 2003). The
broaden-and-build theory might also help explain why gratitude would be positively linked to
hope and forgiveness. When grateful, our mindset is broadened to include the role others play
in aiding our welfare, increasing hope and optimism for a benevolent world with other people
helping rather than hindering our personal strivings. Furthermore, beyond both gratitude and
forgiveness being interpersonal emotions (McCullough & Witvliet, 2002), the common focus on
goodwill might make these emotions strongly related; gratitude is experienced when you receive
others’ goodwill toward you whereas forgiveness is experienced when you give goodwill to others.1
The present study aims to investigate relations between gratitude and subjective well-being, social
support, prosocial behavior, and physical symptoms in early adolescence. Because positive
outcomes have been examined in one known study of gratitude in adolescence (Froh, Sefick,
& Emmons, 2008), we also sought to define gratitude within the context of other positive emotions.
Prior studies of the structure of positive affect in children failed to include moral emotions such as
gratitude (e.g., Laurent et al., 1999). We were interested in which discrete positive affects converge
with gratitude. In the absence of prior factor analytic studies of the structure of gratitude and other
positive affects in children, we consider this secondary focus to be exploratory.

Gratitude in children and adolescents

Several studies aside (e.g., Baumgarten-Tramer, 1938; Becker & Smenner, 1986; Gleason &
Weintraub, 1976; Harris, Olthof, Terwogt, & Hardman, 1987; Russell & Paris, 1994), youth

1
We are grateful to an anonymous reviewer for these insightful relationships between gratitude, hope, and
forgiveness.
636 J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650

gratitude research is scant (see Bono & Froh, in press; Froh & Bono, in press, for reviews). Froh
et al. (2008) made the first attempt at exploring the relation between gratitude and subjective well-
being in early adolescents. In a daily gratitude journal-keeping exercise (i.e., counting blessings),
students in the gratitude condition reported significantly more gratitude compared with those
focused on irritants (i.e., the hassles group) and significantly greater satisfaction with their school
experience compared with both the hassles and no-treatment control groups. Moreover, students
in the gratitude condition reported significantly greater optimism for their upcoming week in
relation to the hassles condition. Correlations computed across conditions indicated that
feeling grateful toward receiving aid was significantly related ( p < 0.01) with positive well-being.
Counting blessings seems to be a promising intervention for adolescent well-being.
Gordon, Musher-Eizenman, Holub, and Dalrymple (2004) examined the content of school-
aged children’s responses to a countrywide essay assignment in which they were instructed to
describe the objects of their gratitude. The most common themes were family, basic needs, friends,
and teachers/schools. Gender differences were found, with girls tending to express more gratitude
for interpersonal relationships and boys tending to express more gratitude for material objects.
Older children included more themes than younger children, and they tended to show less appre-
ciation for material objects in relation to their younger counterparts. Therefore, although
evidence suggests a differential relation between gender and gratitude themes, gender differences
in the experience and expression of gratitude in youth remains mysterious.

Gender differences in gratitude

While related to positive psychological outcomes, gratitude may be an emotion whose expres-
sion and experience men may seek to avoid because of its association with negative emotions and
cognitions (Naito, Wangwan, & Tani, 2005; Solomon, 1995; Sommers & Kosmitzki, 1988). Men
typically express emotions associated with power and status (Brody, 1997, 1999). Therefore,
because gratitude, indebtedness, and dependency are associated with each other in some ways
(Solomon, 1995) but not all (Watkins, Scheer, Ovnicek, & Kolts, 2006) men may view the expe-
rience and expression of gratitude as verification of weakness, which may threaten their masculin-
ity and hurt their social status (Levant & Kopecky, 1995). Consequently, to protect themselves
from any associated negative emotions or social consequences, men might avoid experiencing
and expressing gratitude.
Women, compared with men, seem more likely to experience and express gratitude (Becker &
Smenner, 1986; Gordon et al., 2004; Ventimiglia, 1982) and derive more benefit from it (Kashdan,
Mishra, Breen, & Froh, in press). Indeed, across three studies with various methodologies,
Kashdan et al. (in press) found support for this model. Women, compared with men, evaluated
the expression of gratitude to be less novel, complex, uncertain, and conflicting, and more inter-
esting and exciting (Study 1). When asked to describe a recent episode when they were the bene-
ficiary, women, compared with men, reported less burden and obligation, and greater gratitude
(Study 2). Finally, over the course of 3 months, women with greater gratitude, but not men,
were more likely to satisfy the psychological needs of belongingness and autonomy. Furthermore,
the willingness to openly express emotions, which was greater in women, mediated these gender
differences (Study 3). Taken together, women might be at an advantage compared with men to
 J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650 637

experience and derive benefit from gratitude. The current study is a novel attempt at investigating
adolescent gender differences in gratitude.

The present study

We investigated the interplay among gratitude and subjective well-being, social relationships,
prosocial behavior, physical symptoms, and gender in early adolescence. We expected gratitude
to be positively related to positive affect and life satisfaction. With gratitude being an attribu-
tion-dependent emotion (McCullough & Tsang, 2004; Weiner, 1985) and life satisfaction the
cognitive component of subjective well-being (Diener, Suh, Lucas, & Smith, 1999), we hypothe-
sized that gratitude compared with positive affect would demonstrate a stronger relation with
life satisfaction. We also predicted that negative affect and physical symptoms would be inversely
related to gratitude and gratitude would demonstrate positive relations with subjective well-being,
social support, relational satisfaction, and prosocial behavior. To address construct specificity, we
examined these same relations controlling for positive affect. We conducted an exploratory anal-
ysis of the position of gratitude within the universe of other positive affects. Developmentally,
emotions may show less differentiation in early childhood compared to adults. Thus, it seems
reasonable to presume that although gratitude is considered a ‘‘moral’’ and ‘‘sacred’’ emotion
with links to hope and forgiveness, these arbitrary categories may be less apparent in early child-
hood and instead, all high energy, positive affects may cohere onto a single, underlying factor.
Finally, we predicted that girls, compared with boys, would experience and derive more benefit
from gratitude, operationalized as stronger gratitude correlations with indices of well-being and
positive social functioning.

Method

Participants

Participants were 154 middle school students (mean age ¼ 12.14 years, SD ¼ 0.67, range ¼ 11e
13 years). Students were in grades 6 (29.2%) and 7 (70.8%) within an affluent district (district
median household income ¼ $94,339; state median household income ¼ $43,393). The majority
was male (53.9%) and Caucasian (79.9%).

Measures

Gratitude
The Gratitude Adjective Checklist (GAC; McCullough et al., 2002) was used to assess grati-
tude. It is the sum of three adjectives: grateful, thankful, and appreciative. A Likert scale from
1 (not at all) to 5 (extremely) followed each item. Internal consistency is strong (a ¼ 0.87), and
convergent and discriminant validity has been established in adolescent samples (Froh, Miller,
& Snyder, 2007; Froh et al., 2008). Students were asked to rate the amount they experienced
each feeling ‘‘since yesterday’’ (a ¼ 0.70).
638 J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650

Positive and negative affect

The following 22 affect adjectives were used: interested, distressed, excited, alert, irritable,
ashamed, upset, strong, nervous, guilty, determined, attentive, forgiving, hostile, hopeful,
enthusiastic, active, inspired, afraid, jittery, proud, and scared. A Likert scale from 1 (not
at all) to 5 (extremely) followed each item. Students were asked to rate the amount they
experienced each feeling ‘‘since yesterday’’ (a ¼ 0.82 for positive affect and 0.79 for negative
affect).

Life satisfaction
Students were asked to consider their satisfaction ‘‘during the past few weeks.’’ We interpreted
this as measuring contentment. Students were also asked, ‘‘How do you expect to feel about your
life next week?’’ We interpreted this as measuring optimism. Both items used a Likert scale
ranging from 3 (expecting the worst) to þ3 (expecting the best).
Students completed the Brief Multidimensional Students’ Life Satisfaction Scale (BMSLSS;
Seligson, Huebner, & Valois, 2003). The BMSLSS is a five-item scale that assesses satisfaction
with family life, friendships, school experience, self, and living environment. Response options
are on a seven-point Likert scale ranging from 1 (terrible) to 7 (delighted). Life satisfaction was
the sum of the five items (a ¼ 0.75); similar to prior studies of middle school and high school
students (Huebner, Suldo, & Valois, 2003).

Gratitude in response to aid

Students were asked to ‘‘check off how you handled the most serious problem you dealt with
over the past few weeks.’’ The options were accepted sympathy from someone, talked to someone
about how you were feeling, and got help or advice from someone. Students who indicated
engaging in one or more of these responses were asked, ‘‘Please rate how you felt toward the
person who helped you using the following feelings.’’ The affect adjectives provided were: grateful,
annoyed, embarrassed, surprised, understood, glad, frustrated, and appreciative. A Likert scale
ranging from 1 (not at all) to 5 (extremely) followed all adjectives. The four feelings of grateful,
appreciative, understood, and glad were summed (a ¼ 0.76).

Physical symptoms
Students were asked to check off ‘‘the following things you have experienced over the past 2
weeks.’’ The symptoms included: headaches, dizziness, stomach ache/pain, shortness of breath,
chest pain, runny nose, feeling chilly or really hot, not feeling hungry or not eating, coughing/
sore throat, stiff or sore muscles, nausea or felt like you were going to throw up, and other. A
check was coded as 1, and no check was coded as 0. Higher numbers indicated more physical
symptoms reported (a ¼ 0.72).

Prosocial behavior
Students were instructed to answer ‘‘Yes’’ or ‘‘No’’ to the following questions: ‘‘Have you
helped someone with a problem since yesterday?’’ and ‘‘Have you offered someone emotional
support since yesterday?’’ ‘‘Yes’’ was coded as 1, and ‘‘No’’ was coded as 0 (a ¼ 0.40).
Because this alpha is unacceptable (George & Mallery, 2003), we examined these items
independently.
 J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650 639

Social support
Students completed single items high in content validity assessing supportive peer and familial
relationships. A Likert scale from 1 (not very supportive) to 5 (very supportive) was used. The two
items were: ‘‘How supportive is your family?’’ and ‘‘How supportive are your friends?’’

Procedure

Students enrolled in mandatory curriculum were sought for participation to increase the odds
of obtaining a representative sample. One week prior to data collection, the first author reviewed
all measures and instructions with the teachers. Teachers were provided with a script to introduce
the study to the students to ensure uniformity and control for potential demand characteristics.
Teachers administered questionnaires in classrooms. Data packets were distributed by and
collected from each teacher on the same day. Measures were counterbalanced via a Latin square
to control for order effects.

Results

Data screening

Because interval data seemed to be missing at random (MAR), we imputed values via expecta-
tion maximization (EM) using EQS 6 (see Appendix). EM produces more precise estimates of
imputed scores compared with other methods (e.g., mean imputation) (Bentler, 2006), especially
when data are MAR (McKnight, McKnight, Sidani, & Figueredo, 2007).

Correlations and factor analysis with all students

The lower diagonal for Table 1 shows zero-order correlations among gratitude, subjective well-
being, and physical symptoms. With 12 correlations conducted, a Bonferroni adjustment indi-
cated that alpha needed to be below 0.004 (0.05/12 ¼ 0.004) in order to be significant. Gratitude
demonstrated a significant relation with familial satisfaction, school satisfaction, life satisfaction,
optimism, and positive affect.
To address construct specificity, we reexamined relations between gratitude, subjective well-
being, and physical symptoms using partial correlations controlling for trait positive affect. Using
a Bonferroni adjustment for 11 correlations, alpha needed to be below 0.005 (0.05/11 ¼ 0.005).
The relation between gratitude and family satisfaction was significant ( p ¼ 0.002) and reached
traditional significance for school satisfaction ( p ¼ 0.017), life satisfaction ( p ¼ 0.008), and
optimism ( p ¼ 0.045). The upper diagonal in Table 1 represents these partial correlations.
Zero-order correlations were conducted between gratitude and prosocial behavior, gratitude in
response to aid, family support, and friend support. The Bonferroni adjustment with six variables
indicated that alpha needed to be below 0.008 (0.05/6 ¼ 0.008). Gratitude was related with being
grateful in response to aid. Friend support approached the cutoff ( p ¼ 0.013), and family support
( p ¼ 0.026) and providing emotional support ( p ¼ 0.021) reached traditional significance.
Controlling for positive affect, gratitude was only related with gratitude in response to aid. In
 640
Table 1

J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650

Intercorrelations, partial correlations, means, and standard deviations among gratitude, subjective well-being, and physical symptoms.
Variable 1 2 3 4 5 6 7 8 9 10 11 12
c b b b
1. Gratitude e 0.25 0.12 0.19 0.04 0.12 0.21 0.03 0.16 e 0.09 0.12
2. Family LS 0.33c e 0.42c 0.44c 0.44c 0.30c 0.75c 0.48c 0.37c e 0.29c 0.30c
3. Friends LS 0.23b 0.45c e 0.40c 0.39c 0.23b 0.70c 0.29c 0.24c e 0.22b 0.13
4. School LS 0.30c 0.47c 0.43c e 0.32c 0.33c 0.77c 0.30c 0.32c e 0.42c 0.22b
5. Self LS 0.23b 0.48c 0.43c 0.37c e 0.21b 0.68c 0.39c 0.31c e 0.26c 0.18b
6. Community LS 0.22b 0.33c 0.26c 0.36c 0.26c e 0.57c 0.32c 0.32c e 0.15a 0.09
7. LS 0.37c 0.76c 0.71c 0.78c 0.71c 0.59c e 0.51c 0.45c e 0.40c 0.27c
8. Contentment 0.21b 0.51c 0.34c 0.34c 0.45c 0.36c 0.55c e 0.55c e 0.22b 0.15a
9. Optimism 0.35c 0.42c 0.29c 0.38c 0.39c 0.36c 0.51c 0.59c e e 0.21b 0.14a
10. Positive Affect 0.67c 0.22b 0.21b 0.24c 0.31c 0.20b 0.33c 0.29c 0.36c e e e
11. Negative Affect 0.09 0.29c 0.22b 0.41c 0.26c 0.15a 0.39c 0.22b 0.21b 0.03 e 0.23b
12. Physical Symptoms 0.16b 0.31c 0.15a 0.24c 0.20b 0.10 0.29c 0.18b 0.17b 0.11 0.23b e
M 12.01 6.21 6.23 5.49 6.18 6.47 30.60 6.13 6.21 44.39 17.04 4.28
SD 2.50 0.93 0.87 1.20 0.92 0.78 3.38 0.86 0.88 7.63 5.38 2.73
Note. LS ¼ life satisfaction. Correlations below the diagonal represent zero-order correlations, and correlations above the diagonal represent partial
correlations controlling for positive affect.
a
p < 0.10, bp < 0.05, cp < 0.004 is used for the lower diagonal, which was the alpha set by the Bonferroni adjustment for 12 correlations. cp < 0.005 is
used for the upper diagonal, which was the Bonferroni adjustment for 11 correlationsdbecause positive affect is controlled for here.
 J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650 641

Table 2
Intercorrelations, partial correlations, means, and standard deviations among gratitude, prosocial behavior, gratitude
in response to aid, and relational support.
Variable 1 2 3 4 5 6
c
1. Gratitude e 0.02 0.12 0.31 0.10 0.14a
2. Helped someone 0.09 e 0.23c 0.03 0.00 0.00
3. Gave someone emotional support 0.19b 0.25c e 0.01 0.01 0.13a
4. Gratitude in response to aid 0.55c 0.10 0.08 e 0.09 0.21b
5. Family support 0.18b 0.02 0.03 0.16a e 0.11
6. Friend support 0.20b 0.02 0.15a 0.26c 0.13a e
M 12.01 0.34 0.46 16.23 4.86 4.38
SD 2.50 0.47 0.50 3.11 0.35 0.62
Note. Correlations below the diagonal represent zero-order correlations, and correlations above the diagonal represent
partial correlations controlling for positive affect.
a
p < 0.10, bp < 0.05, cp < 0.008, which was the alpha set by the Bonferroni adjustment for six correlations.

Table 2, zero-order correlations are below the diagonal, and partial correlations are above the
diagonal.
A principal components factor analysis was conducted to determine gratitude’s place among
the positive affects. Three factors yielded eigenvalues greater than 1.0, with the eigenvalues drop-
ping markedly from the first to second factor (i.e., 4.7e1.1). One factor seemed present. We then
re-ran the factor analysis specifying that only one factor be extracted. The factor accounted for
35.87% of the variance. A scree plot provided more evidence for a one-factor solution because
an elbow appeared beyond one factor. To enhance interpretability, Table 3 presents the varimax
rotated component matrix for three factors. Using 0.40 as the factor loading cutoff, gratitude
loads on the first factor with proud, hopeful, excited, forgiving, and inspired (see Table 3).

Table 3
Summary of factor loadings for gratitude and related positive affects using varimax rotation.
Positive affect Factor loading
1 2 3 Communality
Proud 0.73 0.16 0.08 0.56
Gratitude 0.72 0.22 0.30 0.66
Hopeful 0.62 0.30 0.14 0.49
Excited 0.61 0.23 0.03 0.43
Forgiving 0.57 0.11 0.44 0.53
Inspired 0.46 0.41 0.15 0.40
Active 0.26 0.79 0.17 0.72
Determined 0.05 0.77 0.29 0.68
Strong 0.36 0.59 0.17 0.51
Enthusiastic 0.17 0.57 0.32 0.46
Alert 0.00 0.19 0.86 0.78
Attentive 0.39 0.12 0.61 0.54
Interested 0.33 0.21 0.39 0.31
Note. Boldface indicates highest factor loadings.
642 J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650

Relational fulfillment as a mediator of gratitude and physical symptoms

We tested the hypothesis that relational fulfillment would mediate the relationship between
gratitude and physical symptoms following the statistical specification of Baron and Kenny
(1986). Relational fulfillment was defined as the composite of four items (a ¼ 0.68): family satis-
faction, friend satisfaction, family support, and friend support. Three regression equations must
be run to show mediation. The first regression equation requires that the predictor (i.e., gratitude)
has a significant effect on the mediator (i.e., relational fulfillment), which was true, b ¼ 0.332,
R2 ¼ 0.11, F(1, 152) ¼ 18.82, p < 0.001. The second regression equation requires that the predictor
(i.e., gratitude) has a significant effect on the criterion variable (i.e., physical symptoms), which
also was true, b ¼ 0.162, R2 ¼ 0.03, F(1, 151) ¼ 4.08, p ¼ 0.045. The third regression requires
that the relation between the mediator and the criterion is significant when controlling for the
predictor. Therefore, we simultaneously entered both gratitude and relational fulfillment as the
predictors and physical symptoms as the criterion into a regression equation. As hypothesized,
relational fulfillment remained significant, b ¼ 0.246, p ¼ 0.003, h2p ¼ 0.05. The final condition
involved demonstrating a significant reduction in the effect of gratitude on physical symptoms
after accounting for variance attributable to relational fulfillment. The Sobel test of mediation
(Sobel, 1982) was used. A significant Sobel z indicates that the mediator accounts for the influence
of a predictor on a criterion. When the beta weight of the predictor is non-zero in the third regres-
sion equation, partial mediation exists; when it is zero, full mediation exists. Because the beta
weight for gratitude was 0.080, relational fulfillment partially accounted for the significant rela-
tion between gratitude and physical symptoms, z ¼ 2.64, p ¼ 0.008. These data support our
mediation model (see Fig. 1).

Alternative model
Relational fulfillment and physical symptoms were measured contemporaneously. Therefore,
rather than relational fulfillment mediating the relation between gratitude and physical symptoms,
gratitude could mediate the relation between relational fulfillment and life satisfaction. Thus, rela-
tional fulfillment might lead to more gratitude (Algoe, 2006) and consequently less physical

Relational
Fulfillment

.33** -.25*

-.08
Physical
Gratitude Symptoms

Fig. 1. Beta coefficients for the pathways among gratitude, relational fulfillment, and physical symptoms. *p ¼ 0.003.
**p < 0.001.
 J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650 643

symptoms (Emmons & McCullough, 2003, Study 1). There was no evidence in favor of this medi-
ation model, as gratitude was not significantly related with physical symptoms when controlling
for relational fulfillment, p ¼ 0.335.
Mediational models were also tested attempting to identify mechanisms linking gratitude and
positive affect, life satisfaction, prosocial behavior, and relational fulfillment. No support was
found for any of these models.

Gender differences in gratitude

An independent samples t-test was conducted to test the hypothesis that girls reported more
gratitude than boys. Although the results were in the hypothesized direction, we only found
support at a trend level, t(152) ¼ 1.86, p ¼ 0.065, d ¼ 0.30. Girls, on average, reported slightly
elevated levels of gratitude (M ¼ 12.41, SD ¼ 2.39) compared with boys (M ¼ 11.67, SD ¼ 2.55).

Subjective well-being and physical symptoms. The marginally significant difference between
boys and girls in the experience of gratitude, coupled with the early stage of youth gratitude
research (Bono & Froh, in press; Froh & Bono, in press), prompted us to examine if gender
moderated the relationship between gratitude and subjective well-being and physical symp-
toms. Gratitude was centered because it is a continuous predictor (see Aiken & West, 1991).
All subjective well-being variables and the physical symptoms composite served as criterions.
Gender was dummy coded (boys ¼ 0; girls ¼ 1). We constructed three separate hierarchical
regression models for each of the eleven dependent variables. At Step 1, gratitude was entered.
At Step 2, gender was entered. Finally, at Step 3, the gratitude  gender interaction term was
entered. Gender failed to significantly moderate any of the relations ( p values ranged from 0.19
to 0.73). The correlations in Table 4 below the diagonal represent data for girls, and the corre-
lations above the diagonal represent data for boys.

Prosocial behavior and social support. Gender differences were also examined investigating
gratitude’s function as a moral motive and its association with gratitude in response to aid
and social support. Gratitude was centered because it is a continuous predictor (see Aiken &
West, 1991). Helping someone with a problem since yesterday, offering someone emotional
support since yesterday, gratitude in response to aid, family support, and friend support
were the criterions. Gender was dummy coded (boys ¼ 0; girls ¼ 1). We constructed three sepa-
rate hierarchical regression models for each of the five dependent variables. At Step 1, gratitude
was entered. At Step 2, gender was entered. Finally, at Step 3, the gratitude  gender interac-
tion term was entered.
Gender only moderated the effects of gratitude on family support ( p values ranged from 0.13 to
0.95 for the other four dependent variables), as the two-way interaction between gratitude and
gender was significant, FD(1, 150) ¼ 4.00, R2D ¼ 0.03, p ¼ 0.047. This significant interaction effect
was explored with simple effect analyses (see Aiken & West, 1991). For boys, gratitude was posi-
tively related to family support, b ¼ .31, t (3, 150) ¼ 2.93, p ¼ .004, whereas for girls, gratitude was
not significantly related to family support, b ¼ .01, t (3, 150) ¼ .11, p ¼ .916. Simple effects are
plotted in Fig. 2. The correlations in Table 5 below the diagonal represent data for girls, and the
correlations above the diagonal represent data for boys.
 644
Table 4
Intercorrelations, means, and standard deviations among gratitude, subjective well-being, and physical symptoms for boys and girls.

J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650

Variable 1 2 3 4 5 6 7 8 9 10 11 12
c c b c a c c
1. Gratitude e 0.33 0.16 0.36 0.15 0.30 0.36 0.21 0.39 0.72 0.12 0.18
2. Family LS 0.30b e 0.37c 0.55c 0.52c 0.42c 0.79c 0.50c 0.44c 0.27b 0.28b 0.38c
3. Friends LS 0.27b 0.55c e 0.49c 0.37c 0.21a 0.67c 0.32c 0.26b 0.20a 0.36c 0.18
4. School LS 0.16 0.34b 0.31b e 0.46c 0.37c 0.82c 0.34c 0.45c 0.32c 0.48c 0.28b
5. Self LS 0.34b 0.41c 0.50c 0.23a e 0.29b 0.73c 0.43c 0.35c 0.33c 0.26b 0.27b
6. Community LS 0.09 0.17 0.30b 0.33b 0.20a e 0.60c 0.37c 0.37c 0.30b 0.05 0.17
7. LS 0.34b 0.72c 0.76c 0.70c 0.67c 0.55c e 0.53c 0.52c 0.39c 0.41c 0.36c
8. Contentment 0.20a 0.53c 0.34b 0.33b 0.46c 0.33b 0.58c e 0.50c 0.33c 0.16 0.16
9. Optimism 0.24b 0.36c 0.29b 0.20a 0.44c 0.32b 0.46c 0.76c e 0.38c 0.26b 0.24b
10. Positive Affect 0.63c 0.16 0.25b 0.14 0.29b 0.08 0.27b 0.24b 0.35c e 0.12 0.15
11. Negative Affect 0.04 0.30b 0.03 0.34b 0.25b 0.28b 0.36c 0.29b 0.12 0.08 e 0.30b
12. Physical Symptoms 0.15 0.20a 0.13 0.21a 0.08 0.00 0.20 0.22a 0.05 0.05 0.14 e
Girls
M 12.41 6.30 6.39 5.75 6.24 6.56 31.24 6.20 6.38 44.19 16.78 4.39
SD 2.39 0.85 0.82 1.12 0.85 0.71 2.98 0.75 0.74 7.61 5.48 2.35
Boys
M 11.66 6.14 6.10 5.28 6.13 6.40 30.05 6.07 6.06 44.55 17.26 4.19
SD 2.55 0.99 0.89 1.23 0.98 0.83 3.61 0.95 0.97 7.69 5.32 3.02
Note. LS ¼ life satisfaction. Correlations below the diagonal represent zero-order correlations for girls, and correlations above the diagonal represent
zero-order correlations for boys.
a
p < 0.10, bp < 0.05, cp < 0.004, which was the alpha set by the Bonferroni adjustment for 12 correlations.
 J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650 645

5
Boys
Girls

Family Support

4.5
Low Gratitude High Gratitude

Fig. 2. Family support as a function of gratitude and gender.

Note. The graph reflects simple slopes of gratitude predicting family support for boys and girls. One standard deviation
above the mean was used as the upper observed values of gratitude, and one standard deviation below the mean was
used as the lower observed values of gratitude.

Discussion

Adolescents who reported grateful moods indicated greater subjective well-being, optimism,
prosocial behavior, gratitude in response to aid, and social support. Thus, gratitude is likely an
important ingredient for adolescent flourishing. But contrary to prior research (McCullough
et al., 2002; Watkins et al., 2003), gratitude was unrelated to negative affect. Because pathology
and flourishing are separable constructs (Keyes, 2007) gratitude interventions may enhance
adolescent well-being, without necessarily relieving pain and distress. Froh et al. (2008), however,
found that counting blessings was unrelated with increased positive affect, but was related with
Table 5
Intercorrelations, means, and standard deviations among gratitude, prosocial behavior, gratitude in response to aid,
and relational support for boys and girls.
Variable 1 2 3 4 5 6
a a c c
1. Gratitude e 0.19 0.21 0.56 0.30 0.13
2. Helped someone 0.06 e 0.15 0.17 0.00 0.04
3. Gave someone emotional support 0.07 0.37c e 0.13 0.09 0.01
4. Gratitude in response to aid 0.54c 0.01 0.00 e 0.17 0.35c
5. Family support 0.01 0.05 0.16 0.14 e 0.19a
6. Friend support 0.22a 0.06 0.15 0.09 0.02 e
Girls
M 12.41 0.31 0.30 16.38 4.89 4.59
SD 2.39 0.47 0.46 2.93 0.32 0.50
Boys
M 11.66 0.36 0.60 16.11 4.83 4.20
SD 2.55 0.48 0.49 3.26 0.38 0.66
Note. Correlations below the diagonal represent zero-order correlations for girls, and correlations above the diagonal
represent zero-order correlations for boys.
a
p < 0.10, bp < 0.05, cp < 0.008, which was the alpha set by the Bonferroni adjustment for six correlations.
646 J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650

decreased negative affect. Therefore, although promoting gratitude may be a useful intervention
for youth, identifying theoretically meaningful moderators to understand the individual difference
variables influencing the enhancement of positive experiences seems essential for optimal treat-
ment outcomes.
This study helps identify gratitude’s relation with other positive emotions in early adoles-
cence. Gratitude was related with pride, hope, inspired, forgiveness, and excited. Pride is linked
with the locus dimension of causality, whereas gratitude is connected with the controllability
dimension (Weiner, 1985). Pride is experienced when attributing a positive outcome to the
self and is thus classified as a self-reflective emotion. Gratitude is elicited only if the beneficiary
perceives herself as the intended recipient of an intentionally bestowed benefit (McCullough
et al., 2002). Thus, gratitude in relation to pride seems to be an other-focused emotion. This,
however, does not necessarily mean that adolescents in grateful moods discount their own
causal effort. Rather, adolescents in grateful moods may, like dispositionally grateful adults,
‘‘stretch their attributions to incorporate the wide range of people who contribute to their
well-being’’ (McCullough et al., 2002, p. 113). In combination with showing students the impor-
tance of taking credit for one’s successes, it makes sense to also consider teaching them grateful
attributions. Doing so may provide positive self-esteem and the added bonus of relational
enhancement.
In-line with Fredrickson’s (1998, 2001) broaden-and-build theory of positive emotions, grati-
tude being grouped with hope and forgiveness could be explained in terms of broadening. The
experience of gratitude might coincide with constructive and positive appraisals of the future
or life circumstances, and thereby hope. This might be one reason why adolescents responded
similarly to these emotion measures. Furthermore, gratitude might be linked with forgiveness
because adolescents experiencing both emotions might be focused on being socially constructive.1
The mediational analysis also seems to support this theory. To the extent that gratitude in early
adolescence broadens one’s view about life’s circumstances, which might include relational fulfill-
ment, gratitude may help adolescents flourish by means of enhancing their physical health and
well-being. Experimental and longitudinal studies are needed to determine if gratitude causes
the broadening of thought-action repertoires, which then builds intellectual, social, or physical
resources. For example, does gratitude lead an adolescent to creatively repay their parents’ kind-
ness, which then bolsters and maintains strong parent-child bonds? If so, these social resources
can then be utilized in times of need. We encourage researchers to help define youth gratitude
within the context of other positive emotions.
Girls tended to report more gratitude than boys, but boys showed a stronger relation between
gratitude and a single positive outcome out of several under studydfamily support. Although
most of our results failed to find support for gender differences and these findings should be
considered exploratory, we provide tentative speculation about these relations in hopes of further
understanding gratitude in early adolescence (which may differ from the manifestation in adult
samples; Kashdan et al., in press). Considering the potential causal relation between gratitude
and family support might elucidate the positive relation found in boys, but not girls. Does family
support cause gratitude? If so, perhaps boys experience more gratitude if they have high levels of
familial support, but girls with less familial support might still experience gratitude because of
their tendency to be more dispositionally grateful compared with boys. Does gratitude cause
family support? This is possible because gratitude seems to promote relational strengthening
 J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650 647

(Algoe, 2006). Or perhaps a third variable, like religiosity, explains the relation between gratitude
and family support. Indeed, gratitude and religion go hand-in-hand (Emmons, 2005, 2007), and
people who self-identify as religious also report strong social bonds (Myers, 2000). Longitudinal
and experimental work is needed to clarify this relation. Furthermore, the trends favoring boys
might be partially explained by the androgyny hypothesis, which suggests that psychologically
androgynous individuals (i.e., those who possess both masculine and feminine characteristics in
almost equal measure) are optimally healthy compared with their sex-typed counterparts (i.e.,
those who posseses mainly masculine or feminine characteristics) (Lefkowitz & Zeldow, 2006).
Therefore, to the extent that boys view gratitude as feminine, they might derive more benefit
from its experience and expression, but only when predicting family support.

Strengths and limitations

This study adds to the gratitude literature in several ways. It is the first known study to examine
the following in early adolescence: gratitude as a mood; gender as a moderator on numerous
measures (e.g., prosocial behavior); and relational fulfillment as a potential mechanism
accounting for an inverse relation between gratitude and physical distress.
Several limitations are noted. First, the correlational design precludes causal statements.
Second, our study most clearly demonstrates that gratitude as a mood covaries with other posi-
tive emotions in early adolescence. Thus, any discussion of differential relations between gratitude
and specific positive affects, such as stronger ties to hope and forgiveness, should be considered
preliminary. To help elucidate the manifestation and developmental trajectory of gratitude
within youth, researchers may consider including trait gratitude measures such as the Gratitude
Questionnaire e 6 (GQ-6; McCullough et al., 2002) or the Gratitude Resentment and Appreci-
ation Test (GRAT; Watkins et al., 2003). Third, to enhance reliability, future researchers should
measure positive emotions via scales with multiple items (e.g., Dispositional Positive Emotions
Scale; Shiota, Keltner, & John, 2006), not the single item approach used in this study. Fourth,
demand characteristics could have been present; however, given that teachers used standardized
scripts to introduce the study, daily ratings were completed before the gratitude and hassles listings,
and the principal investigator and a school psychology intern conducted several random integrity
checks, we think the effects, if any, were negligible. Finally, significant findings should be considered
in context of the large number of tests conducted (and increased Type I error rate).

Conclusion

Gratitude in early adolescents was related with social, emotional, and physical benefits. Grat-
itude stems from and creates strong social bonds (Algoe, 2006), partly by encouraging people to
create meaningful experiences for others (Fredrickson, 2004). For instance, the benefactor may
experience happiness after being reinforced for behaving prosocially (McCullough et al., 2001),
onlookers may experience awe (Haidt, 2000), and future benefactors may experience gratitude.
‘‘This socioemotional cycle centered on gratitude could continue indefinitely’’ (Fredrickson,
2004, p. 159). Therefore, with the benefits of gratitude extending beyond the self to enhance
others’ lives, it makes sense for psychologists to foster it in youth.
648 J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650

Acknowledgment

Gratitude is extended to Andrew Greene, principal, and the teachers Lois Krawitz, Ellen
Gryszkin, and Susan Hart for assisting with data collection. Thanks go to Philip C. Watkins,
Christopher J. Fives, and Robert W. Motta for helpful comments on an earlier draft of this
manuscript and Christine M. White and Lisa Wajsblat for their assistance in the preparation
stages. Todd B. Kashdan was supported by National Institute of Mental Health Grant
MH-73937.

Appendix. Supplementary data

Supplementary data associated with this article can be found, in the online version, at
doi:10.1016/j.adolescence.2008.06.006.

References

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. London: Sage.
Algoe, S. B. (2006). A relational account of gratitude: A positive emotion that strengthens interpersonal connections.
Dissertation Abstracts International, 66, 5137, (Doctoral dissertation, University of Virginia, 2005).
Baron, R. M., & Kenny, D. A. (1986). The moderatoremediator variable distinction in social psychological research:
conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173e1182.
Bartlett, M. Y., & DeSteno, D. (2006). Gratitude and prosocial behavior: Helping when it costs you. Psychological
Science, 17, 319e325.
Baumgarten-Tramer, F. (1938). ‘‘Gratefulness’’ in children and young people. Journal of Genetic Psychology, 53, 53e66.
Becker, J. A., & Smenner, P. C. (1986). The spontaneous use of thank you by preschoolers as a function of sex, socio-
economic status, and listener status. Language in Society, 15, 537e546.
Bentler, P. M. (2006). EQS 6 structural equations program manual. Encino, CA: Multivariate Software, Inc.
Bono, G., & Froh, J. J. Gratitude in school: Benefits to students and schools. In R. Gilman, E. S. Huebner, & M.
Furlong (Eds.), Handbook of positive psychology in the schools: Promoting wellness in children and youth. Hillsdale,
NJ: Lawrence Erlbaum, in press.
Brody, L. R. (1997). Gender and emotion: Beyond stereotypes. Journal of Social Issues, 53, 369e394.
Brody, L. R. (1999). Gender, emotion, and the family. Cambridge: Harvard University Press.
Diener, E., Suh, E. M., Lucas, R. E., & Smith, H. L. (1999). Subjective well-being: three decades of progress. Psycho-
logical Bulletin, 125, 276e302.
Dunn, J. R., & Schweitzer, M. E. (2005). Feeling and believing: The influence of emotion on trust. Journal of Person-
ality and Social Psychology, 88, 736e748.
Emmons, R. A. (2004). Gratitude. In C. Peterson, & M. E. P. Seligman (Eds.), Character strengths and virtues: A hand-
book and classification (pp. 553e568). New York: Oxford University Press.
Emmons, R. A. (2005). Emotion and religion. In R. F. Paloutzian, & C. L. Park (Eds.), Handbook of the psychology of
religion and spirituality (pp. 235e252). New York: Guilford Press.
Emmons, R. A. (2007). Thanks! How the new science of gratitude can make you happier. Boston: Houghton Mifflin
Company.
Emmons, R. A., & McCullough, M. E. (2003). Counting blessings versus burdens: An experimental investigation of
gratitude and subjective well-being in daily life. Journal of Personality and Social Psychology, 84, 377e389.
Emmons, R. A., & Shelton, C. M. (2002). Gratitude and the science of positive psychology. In C. R. Snyder, &
S. J. Lopez (Eds.), Handbook of positive psychology (pp. 459e471). New York: Oxford University Press.
Fredrickson, B. L. (1998). What good are positive emotions? Review of General Psychology, 2, 300e319.
 J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650 649

Fredrickson, B. L. (2001). The role of positive emotions in positive psychology: the broaden-and-build theory of posi-
tive emotions. American Psychologist, 56, 218e226.
Fredrickson, B. L. (2004). Gratitude, like other positive emotions, broadens and builds. In R. A. Emmons, &
M. E. McCullough (Eds.), The psychology of gratitude (pp. 145e166). New York: Oxford University Press.
Froh, J. J., & Bono, G. The gratitude of youth. In S. J. Lopez (Ed.), Positive psychology: Exploring the best in people.
Westport, CT: Greenwood Publishing Company, in press.
Froh, J. J., Miller, D. N., & Snyder, S. (2007). Gratitude in children and adolescents: Development, assessment, and
school-based intervention. School Psychology Forum, 2, 1e13.
Froh, J. J., Sefick, W. J., & Emmons, R. A. (2008). Counting blessings in early adolescents: An experimental study of
gratitude and subjective well-being. Journal of School Psychology, 46, 213e233.
George, D., & Mallery, P. (2003). SPSS for Windows step by step: A simple guide and update. Boston: Allyn & Bacon.
Gleason, J. B., & Weintraub, S. (1976). The acquisition of routines in child language. Language in Society, 5, 129e136.
Gordon, A. K., Musher-Eizenman, D. R., Holub, S. C., & Dalrymple, J. (2004). What are children thankful for? An
archival analysis of gratitude before and after the attacks of September 11. Applied Developmental Psychology, 25,
541e553.
Haidt, J. (2000). The positive emotion of elevation [commentary on ‘‘cultivating positive emotions to optimize health
and well-being’’ by B.L. Fredrickson]. Prevention and Treatment, 3. http://journals.apa.org/prevention/volume3/
pre0030003c.html Article 3. Available from: Accessed 22.02.07.
Harpham, E. J. (2004). Gratitude in the history of ideas. In R. A. Emmons, & M. E. McCullough (Eds.), The
psychology of gratitude (pp. 19e36). New York: Oxford University Press.
Harris, P. L., Olthof, T., Terwogt, M. M., & Hardman, C. E. (1987). Children’s knowledge of the situations that
provoke emotion. International Journal of Behavioral Development, 10, 319e343.
Huebner, E. S., Suldo, S. M., & Valois, R. F. (2003). Psychometric properties of two brief measures of children’s life
satisfaction: the students’ life satisfaction scale and the brief multidimensional students’ life satisfaction scale
(BMSLSS). http://www.childtrends.org/files/huebnersuldovaloispaper.pdf. Accessed 26.12.05.
Kashdan, T. B., Mishra, A., Breen, W. E., & Froh, J. J. Gender differences in gratitude: Examining appraisals, narra-
tives, the willingness to express emotions, and changes in psychological needs. Journal of Personality, in press.
Keyes, C. L. M. (2007). Promoting and protecting mental health as flourishing: A complementary strategy for
improving national mental health. American Psychologist, 62, 95e108.
Komter, A. E. (2004). Gratitude and gift exchange. In R. A. Emmons, & M. E. McCullough (Eds.), The psychology of
gratitude (pp. 195e212). New York: Oxford University Press.
Laurent, J., Catanzaro, S., Joiner, T., Rudolph, K., Potter, K., Lambert, S., et al. (1999). A measure of positive
and negative affect for children: scale development and preliminary validation. Psychological Assessment, 11,
326e338.
Lazarus, R. S., & Lazarus, B. N. (1994). Passion and reason: Making sense of our emotions. New York: Oxford Univer-
sity Press.
Lefkowitz, E. S., & Zeldow, P. B. (2006). Masculinity and femininity predict optimal mental health: A belated test of
the androgyny hypothesis. Journal of Personality Assessment, 87, 95e101.
Leslie, A. M. (1987). Pretense and representation: The origins of ‘‘theory of mind. Psychological Review, 94, 412e426.
Levant, R. F., & Kopecky, G. (1995). Masculinity, reconstructed. New York: Dutton.
Lyubomirsky, S., King, L. A., & Diener, E. (2005). The benefits of frequent positive affect. Psychological Bulletin, 131,
803e855.
McCullough, M. E., Emmons, R. A., & Tsang, J. (2002). The grateful disposition: A conceptual and empirical topog-
raphy. Journal of Personality and Social Psychology, 82, 112e127.
McCullough, M. E., Kilpatrick, S. D., Emmons, R. A., & Larson, D. B. (2001). Is gratitude a moral affect? Psycholog-
ical Bulletin, 127, 249e266.
McCullough, M. E., & Tsang, J. (2004). Parent of the virtues? The prosocial contours of gratitude. In R. A. Emmons, &
M. E. McCullough (Eds.), The psychology of gratitude (pp. 123e144). New York: Oxford University Press.
McCullough, M. E., & Witvliet, C. V. (2002). The psychology of forgiveness. In C. R. Synder, & S. J. Lopez (Eds.),
Handbook of positive psychology (pp. 446e458). New York: Oxford University Press.
McKnight, P. E., McKnight, K. M., Sidani, S., & Figueredo, A. J. (2007). Missing data: A gentle introduction. New
York: The Guildford Press.
650 J.J. Froh et al. / Journal of Adolescence 32 (2009) 633e650

Myers, D. G. (2000). The American paradox: Spiritual hunger in an age of plenty. New Haven, CT: Yale University
Press.
Naito, T., Wangwan, J., & Tani, M. (2005). Gratitude in university students in Japan and Thailand. Journal of Cross
Cultural Psychology, 36, 247e263.
Russell, J. A., & Paris, F. A. (1994). Do children acquire concepts for complex emotions abruptly? International Journal
of Behavioral Development, 17, 349e365.
Seligson, J. L., Huebner, E. S., & Valois, R. F. (2003). Preliminary validation of the brief multidimensional students’ life
satisfaction scale (BMSLSS). Social Indicators Research, 61, 121e145.
Sheldon, K. M., & King, L. (2001). Why positive psychology is necessary. American Psychologist, 56, 216e217.
Shiota, M. N., Keltner, D., & John, O. P. (2006). Positive emotion dispositions differentially associated with big five
personality and attachment style. The Journal of Positive Psychology, 1, 61e71.
Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in structural equations models. In S. Leinhart
(Ed.), Sociological methodology (pp. 290e312). San Francisco: Jossey-Bass.
Solomon, R. C. (1995). The cross-cultural comparison of emotion. In J. Marks, & R. T. Ames (Eds.), Emotions in Asian
thought (pp. 253e294). Albany: State University of New York Press.
Sommers, S., & Kosmitzki, C. (1988). Emotion and social context: An AmericaneGerman comparison. British Journal
of Social Psychology, 27, 35e49.
Tsang, J. (2006). Gratitude and prosocial behavior: An experimental test of gratitude. Cognition & Emotion, 20, 138e148.
Tsang, J. (2007). Gratitude for small and large favors: A behavioral test. Journal of Positive Psychology, 3, 157e167.
Ventimiglia, J. C. (1982). Sex roles and chivalry: Some conditions of gratitude to altruism. Sex Roles, 8, 1107e1122.
Watkins, P. C., Scheer, J., Ovnicek, M., & Kolts, R. (2006). The debt of gratitude: Dissociating gratitude and indebt-
edness. Cognition & Emotion, 20, 217e241.
Watkins, P. C., Van Gelder, M., & Frias, A. Furthering the science of gratitude. In C. R. Snyder & S. J. Lopez (Eds.),
The handbook of positive psychology (2nd ed.). New York: Oxford University Press, in press.
Watkins, P. C., Woodward, K., Stone, T., & Kolts, R. (2003). Gratitude and happiness: Development of a measure of
gratitude, and relationships with subjective well-being. Social Behavior and Personality, 31, 431e452.
Weiner, B. (1985). An attributional theory of achievement motivation and emotion. Psychological Review, 92, 548e573.