Content uploaded by Jeffrey A. Hall
Author content
All content in this area was uploaded by Jeffrey A. Hall on Mar 19, 2018
Content may be subject to copyright.
Personal Relationships,(2017).PrintedintheUnitedStatesofAmerica.
Copyright © 2017 IARR; DOI: 10.1111/pere.12183
Humor in romantic relationships: A meta-analysis
JEFFREY A. HALL
University of Kansas
Abstract
This manuscript reports a meta-analysis of the correlation between humor and relationship satisfaction in romantic
relationships, combining 43 distinct samples from 39 manuscripts and including 15,177 participants (54.7% female)
with a mean age of 34.0 years. Drawing from 3 theoretical dimensions of humor (i.e., within-person/relational,
positive/negative, instrumental/content free), weighted mean effect sizes were estimated for 12 distinct subdimensions
of humor. All 6 positive types of humor were positively associated with relationship satisfaction, and 4 of 5 negative
types of humor were negatively associated with relationship satisfaction. Instrumental humor was unassociated with
satisfaction. In support of the proposed relational dimension, self-reported humor showed small associations with
satisfaction, while partner-perceived and relational humor showed medium and large effect sizes, respectively.
Humor is highly valued by single adults
seeking romantic relationships (Wilbur &
Campbell, 2011) and by committed romantic
partners reecting on what contributes to
relationship success (Lauer, Lauer, & Kerr,
1990; Ziv, 1988). Not only is humor a common
form of daily talk among romantic partners
(Alberts, Yoshimura, Rabby, & Loschiavo,
2005), romantic partners are probably one
of, if not the single most common, audience
to partners in expressing humor. Playfulness
between romantic partners is a crucial com-
ponent in bonding and establishing relational
security (Betcher, 1981). Laughter, particu-
larly shared laughter, is an important indicator
of romantic attraction between potential mates
(Hall, 2015) and is associated with relational
quality, closeness, and support in established
relationships (Kurtz & Algoe, 2015). Sharing
a humorous experience can reinforce bonds
and increase relationship satisfaction (Bazzini,
Stack, Martincin, & Davis, 2007).
For nearly 30 years (e.g., Ziv, 1988),
researchers have offered empirical evidence
Jeffrey A. Hall, Department of Communication Studies,
University of Kansas.
Correspondence should be addressed to Jeffrey A. Hall,
University of Kansas, Bailey Hall, 1440 Jayhawk Blvd, Rm
102, Department of Communication Studies, Lawrence,
KS 66045-7574, e-mail: hallj@ku.edu.
of humor’s role in relationships. Although
some researchers have identied a strong
association between humor and relation-
ship satisfaction (e.g., Weisfeld et al., 2011),
for others, large effects are elusive, leading
Barelds and Barelds-Dijkstra (2010) to claim
that “humor plays a limited role in intimate,
long-term relationships” (p. 458). These mixed
ndings suggest that the association between
humor and relationship satisfaction is highly
dependent on its conceptualization and oper-
ationalization. Many distinct measures of
humor in romantic relationships have been
employed, which presents another challenge
to clarifying humor’s role in romantic rela-
tionships. The multifaceted nature of humor
complicates a simple summary of its role in
romantic relationships (Martin, 1998). As
research on humor has matured and the num-
ber of studies increased, there is a growing
need to supplement qualitative reviews with
quantitative ones.
Ameta-analysisiswellequippedtoprovide
order and clarity when the results are multi-
faceted, moderated, or inconsistent (Lipsey
&Wilson,2001).Thegoalofthepresent
investigation is to identify the direction and
strength of the association between humor
and relationship satisfaction in romantic
relationships. Meta-analyses will be guided by
1
2J. A. Hall
theory on the primary dimensions of humor.
One relevant theoretical dimension of humor is
positive versus negative affect (Craik & Ware,
1998; Martin, 1998). In the context of romantic
relationships, there is also a relational versus
within-person dimension, which is the degree
to which humor—in production, affect, and
function—is about us, about you, or about
me. The strength and valence of humor’s
association with relationship satisfaction is
dependent on the subdimension’s arrange-
ment on two primary axes: positive versus
negative and relational versus within-person.
The association is predicted to be stronger
when the measure of humor is more relational
(and less within-person) and with a positive
sign when humor is affectively positive and
negative when affectively negative. To further
establish the distinction between relational
and within-person humor, one moderator of
the strength of association will be explored:
whether humor is self-reported or perceived in
one’s partner. Additionally, this investigation
will test moderation by content-specic versus
content-free measures (Martin, 1998) as well
as functional forms of humor (Graham, Papa,
&Brooks,1992).Finally,continuousmoder-
ators by study and sample characteristics will
be explored among types of humor that were
heterogeneous and had sufcient coverage.
The proposed relational versus within-person
dimension
Interpretations guided by theories of natural
selection (Darwin, 1859) suggest that humor
is advantageous for the survival of primates
inasmuch that it enhances pair bonding, eases
social interactions, increases group cohesion,
and engenders an approach response with other
primates (Caron, 2002). Bonding through con-
structive play is commonplace among many
mammals, and what can be called laughter
among primates signals playful intent. There-
fore, laughter and play for the sake of bond-
ing likely preceded humor in humans. For
humans, humor’s primary value as a trait, skill,
or resource is in the social realm (Caron,
2002; Craik & Ware, 1998), with the notable
exception of coping or adaptive humor. In the
social world, a sense of humor is valuable
because it is a “potent instrument for at once
forging indispensible social bonds and permit-
ting the individual a great deal of (self-serving)
maneuverability within them” (Storey, 2003,
p. 323). Although an adaptive resource for
the individual, humor’s value in the context
of romantic relationships is not so straightfor-
ward.
When a sense of humor refers to humor
production, its strongest benets are obtained
by the speaker. Being thought to be a humor-
ous person is a perception generated within
one’s social network rather than a product of
any given relationship (Craik & Ware, 1998).
Being perceived to be a humorous person may
or may not benet one’s romantic partnership.
In a romantic relationship, the value of humor
depends on how it is used (Graham et al., 1992;
Hall, 2013). Although humor production may
help to generate a general atmosphere of pleas-
antness, its association with relational satis-
faction is weak (Barelds & Barelds-Dijkstra,
2010; Cann, Zapata, & Davis, 2011). The bene-
ts of humor production may be valuable to the
humor-producing individual but not as much
for the joker’s partner or for the relationship
(Hall, 2013).
When humor is truly relational—that is, it
is cocreated and enjoyed by both partners—its
role in the relationship is greatly enhanced.
Joking and laughter, particularly shared laugh-
ter, are signs of mutual romantic interest (Hall,
2015; Kurtz & Algoe, 2015). Betcher (1981)
rst suggested that the evolutionary roots of
humor as play could be applied to cocreated
humor in intimate relationships and suggested
three possible benets. The rst is intrin-
sic to the value of laughter and mirth itself:
Humor enhances relationships by magnifying
the enjoyment of shared interactions (Bazzini
et al., 2007). Among humor’s functions, its use
in bringing about good cheer is most consis-
tently associated with relationship satisfaction
(Hall, 2013). Laughter, particularly shared
laughter, is associated with social bonding
(Caron, 2002), romantic interest (Hall, 2015),
and relationship satisfaction (Kurtz & Algoe,
2015). Second, private jokes and playfulness
create a shared space—a home where risks
can be taken, and departures from the mundane
are welcomed (Betcher, 1981). Engaging in
playful and humorous banter reafrms the
Humor meta-analysis 3
safety and intimacy of the relationship (Bip-
pus, 2000; Ziv, 1988). Third, shared humor
afrms each partner’s values and perspective
(Betcher, 1981). The feeling that someone
gets your jokes means they get you too, even
when those jokes are in poor taste (Hall &
Sereno, 2010). Martin (1998) concurs, “people
tend to enjoy and laugh at humor that reects
themes and attitudes that are in agreement with
their own attitudes, interests, and behaviors”
(p. 56). Shared humor bonds, makes secure,
and afrms the individual’s attitudes.
In terms of measurement, the within-person
versus relational dimension encapsulates
the distinction between one’s own sense
of humor and humor that is generated
or shared between romantic partners.
These two distinctions create four subdi-
mensions: production–appreciation and
shared–evaluated. The distinction between
humor production and appreciation is widely
accepted (Craik & Ware, 1998; Kohler & Ruch,
1996; Martin, 1998). Humor production is the
ability to produce humorous communication
or actions, to make jokes, tell funny stories,
and make others laugh (Thorson & Powell,
1993). There are many self-reported measures
of humor production (e.g., Booth-Buttereld
&Booth-Buttereld,1991:humororientation
[HO]; Martin, Puhlik-Doris, Larsen, Gray,
&Weir,2003:HumorStylesQuestionnaire
[HSQ]—afliative; Thorson & Powell, 1993:
Multidimensional Sense of Humor Scale
[MSHS]—creation; Ziv & Gadish, 1989:
creation). They measure the ability to create
or produce humor or elicit laughter and related
concepts, such as the perception that others
think one is funny or witty (e.g., HO, MSHS),
being expected to produce humor by friends
(e.g., HO), and enjoyment of producing humor
(e.g., HSQ). When administered simulta-
neously, self-reported measures of humor
production show convergent validity (Kohler
&Ruch,1996).Consideringthesemeasures
on face value, they are content free. Measures
of humor production do not ask how the indi-
vidual makes others laugh or the content of the
stories or jokes told. Notably, these measures
do not specify that the audience to humor
is one’s romantic partner; humor production
spans context and relationships.
Similar to humor production, humor appre-
ciation is content free. A person who appreci-
ates humor enjoys laughing, laughs easily, likes
jokes, and is generally a good and receptive
audience to others’ jokes (Thorson & Powell,
1993). Humor appreciation is strongly associ-
ated with trait cheerfulness and positive affect,
and behavioral measures of production and
appreciation suggest that they are distinct con-
structs (Kohler & Ruch, 1996). Like measures
of production, humor appreciation is not rela-
tionship or context specic—an appreciative
person will laugh easily when joking around
with many different conversational partners.
By contrast, shared humor is the degree to
which humor is shared, and laughing together
occurs in relationships. The Relational Humor
Inventory (RHI) self-positive scale (de Kon-
ing & Weiss, 2002) is the most commonly
used measure of shared humor. Other mea-
sures of relational humor (e.g., Jacobs, 1985;
Raniseski, 1998) focus on sharing stories or
events, sharing inside jokes, and joking with
one another. Couples often look for things that
they can laugh about to strengthen their bond
(Bippus, 2000; Lauer et al., 1990; Ziv, 1988).
The evaluative dimension of humor is a
judgment of another person’s sense of humor
in quality, capability, and value. It is nearly
always perceived in one’s relational partner,
not in oneself.1Evaluative measures assess
the degree to which one’s partner’s humor
is skillful, enjoyable, attractive, and capable
of eliciting laughter. Single-item measures of
humor evaluation are sometimes used (e.g.,
Rust & Goldstien, 1989: “I really appreciate
my partner’s sense of humor”). This dimension
does not evaluate what that sense of humor is
like (e.g., bawdy or good-natured) or does for
the couple (e.g., eases tensions); it is content
free.
When subdimensions of humor are more
relational (and less within-person), they are
likely to have a stronger association with rela-
tionship satisfaction. Cocreated humor reects
1. This is probably because an evaluative dimension of
one’s own humor is a tautology: Most people believe
they have a good sense of humor because what often
makes another person’s sense of humor good is that it
is funny to the perceiver.
4J. A. Hall
the perspective, experiences, and emotions that
overlap between people (Martin, 1998). As
such, when humor is relational in nature, it
is more likely to be reective of the underly-
ing relational state, such as satisfaction (Ziv,
1988) or dissatisfaction (Saroglou, Lacour, &
Demeure, 2010). Because humor’s effective-
ness is audience dependent, what people think
about their romantic partner’s sense of humor
is often more important than how a partner
actually is or perceives him- or herself to
be (Cann et al., 2011; Ziv & Gadish, 1989).
As such, humor perceived in one’s partner
is predicted to lie between the within-person
versus relational dimension of humor. There-
fore, partner-perceived humor is likely to have
astrongerassociationwithsatisfactionthan
self-reported humor. Formally stated:
H1: The effect size of the association
between the subdimension of humor and
relationship satisfaction will increase as
the subdimension becomes more relation-
ally oriented, moving from self-reported to
partner perceived and then to relationship
focused.
Perceptions of partners are distinct from
partner effects, which can be estimated when
data exist for both partners and is measured
by the degree to which one partner’s humor
is associated with the other partner’s relational
satisfaction. Thus:
RQ1: What will be the association between
partner effects and relational satisfaction?
The affective valence dimension
Humor can be used to demean, belittle, attack,
put down, or mock, and it can be a means
of expressing hostility toward one’s partner
(Jacobs, 1985; Ziv, 1988). The two negative
RHI subscales (de Koning & Weiss, 2002)
both assess partner-directed attacks. Whether
self-reported or evaluated in one’s partner, this
subscale focuses on one partner being made
the object of humorous attack or put down by
the other partner.2In a romantic relationship,
negative humor directed at one’s partner is
2. One problem with the RHI (de Koning & Weiss, 2002)
negative-partner subscale is that it includes reworded
very different than negative humor directed at
external targets. When the object of ridicule is
not one’s partner but a person or group outside
of the relationship, it is more characteristic
of an aggressive humor style (Martin et al.,
2003). Humor used in this fashion sometimes
expresses inappropriate, sexist, racist, or offen-
sive attitudes. However, the HSQ-aggressive
subdimension does not capture forms of nega-
tive humor that are simply inappropriate or in
poor taste. The earthy measure (Craik & Ware,
1998) and Hall and Sereno’s (2010) inap-
propriate joke telling are more specic types
of negative humor not directed at one’s rela-
tional partner. Finally, self-defeating humor is
another content-specic style of humor that is
not relationally focused (Martin et al., 2003).
This involves making oneself the object of oth-
ers’ humor and can have relational implications
for a couple’s joint face via embarrassment
(Hall, 2011). Although the association’s
strength is expected to be a function of the
within-person versus relational dimension, the
sign of the association is predicted to vary
based on the subdimension’s place on the
affective continuum:
H2a: Negative forms of humor will be
negatively associated with relationship
satisfaction.
H2b: Positive forms of humor will be
positively associated with relationship
satisfaction.
The content dimension: A functional approach
The multidimensional content dimension of
humor encompasses humor’s communica-
tive goal and purpose (Craik & Ware, 1998;
Martin, 1998). There is a long tradition of
items from the self-instrumental scale. From the per-
spective of one’s partner, a partner’s failed attempt
to make light of something, change the subject, or
avoid a distressing topic can be as hurtful as a direct
attack. Yet there is a distinction between direct attacks
and failed attempts to use humor to manage conicts
(Campbell et al., 2008; Jacobs, 1985). One is inten-
tionally hurtful, and the other is simply not effec-
tive at achieving communicative goals. Nonetheless,
this measure was categorized in analyses as negative,
not instrumental-functional to reect the intent of its
creators.
Humor meta-analysis 5
studying the functions of humor in romantic
relationships (Graham et al., 1992; Ziv, 1988).
It can be dened as the “intended use of a
humorous message or behavior to achieve a
communicative goal specic to the romantic
relationship” (Hall, 2013, p. 274). Dened
this way, the inuence of humor in a romantic
relationship depends on its ability to help
relational partners communicate an attitude or
emotion or achieve a relational goal. Although
there have been attempts to categorize and
separately measure the relational functions of
humor (Hall, 2013), commonly used measures
do not make such ne-tuned distinctions. For
example, the RHI-instrumental subscale (de
Koning & Weiss, 2002) includes items regard-
ing easing tensions and managing moods, and
other conceptualizations focus on problem
solving and conict management (e.g., Jacobs,
1985), attitude expression and apologizing
(e.g., Graham et al., 1992), and sexual expres-
sion (e.g., La France & Hall, 2012). The
functional dimension can be considered in
conjunction with the interpersonal and affec-
tive dimensions. Most functions are positive
(Bippus, 2000) and cluster together in factor
analyses (de Koning & Weiss, 2002).
Another functional use of humor that is not
relationship specic is the use of humor to
release stress and tension (Thorson & Powell,
1993). Within romantic relationships, the use
of humor to cope with stressful situations
and reframe life’s challenges is a commonly
identied function (Graham et al., 1992; Ziv,
1988). However, no commonly used functional
measure focuses on coping with one’s partner.
Rather, humor to cope is typically thought to
be a self-focused activity (Caird & Martin,
2014; Thorson & Powell, 1993). Although
self-enhancing coping is not necessarily
relationally oriented (Martin et al., 2003), it
appears to have relational implications as it is
positively associated with satisfaction (Cann
et al., 2011; Hall, 2013).
Afewnegativefunctionsofhumorarespe-
cic to relationships (Bippus, 2000; Graham
et al., 1992). Attacking, belittling, or making
fun of one’s partner could be conceived as a
relational and (dys)functional use of humor.
Another is distancing humor, used to push rela-
tional partners apart, to avoid one’s partner, or
to deect conversations about the relationship
(Raniseski, 1998).
Overall, using humor instrumentally may
not be a sign of a satisfying relationship.
Rather, it could be a sign that conict needs
to be managed, hidden thoughts expressed, and
apologies made (Campbell, Martin, & Ward,
2008; de Koning & Weiss, 2002; Hall, 2013),
all of which are more common in less sat-
isfying relationships. However, the functional
dimension is limited by the paucity of stud-
ies that independently measure the various
functions of humor rather than grouping them
together. Therefore:
RQ2: Will the a sso cia tio n bet ween instru-
mental/functional humor use and relation-
ship satisfaction be stronger or weaker
compared to content-free measures of
humor?
Each of the three theoretical dimensions is
expected to function independently. That is, the
strength of association on the within-person
versus relational dimension is predicted to
apply to humor whether it is positive, negative,
or content free. For example, negative humor
that is relationally oriented (i.e., distancing)
or partner oriented (i.e., attacking) should
be more strongly negatively associated with
relationship satisfaction than would nega-
tive humor that describes one’s own sense
of humor (e.g., inappropriate humor). The
affective dimension should also indepen-
dently inuence this association. For example,
content-free but positive dimensions (e.g.,
evaluation) will be positively associated with
satisfaction, and negative instrumental humor
(e.g., distancing) will be negatively associated
with satisfaction. To test the hypotheses and
answer the research questions, meta-analysis
will be used. In providing a thorough and
nuanced summary of this association, this
manuscript offers a summative point for future
research on humor in romantic relationships.
Method
Study identication
Study identication began by searching
PsycINFO, Communication & Mass Media
Complete, Dissertation Abstracts, and
6J. A. Hall
Education Resources Information Center
(ERIC) with the key terms “humor” and “satis-
faction,” which resulted in 320 articles. When
results were limited to relationship satisfaction
or quality, 136 articles remained. Studies
were then excluded when they did not include
measures of both humor and satisfaction, were
qualitative or did not report effect size, and
the effect sizes were not isolated (i.e., the
measure of humor was collapsed with other
measures, such as positive behaviors or per-
sonality traits). This resulted in a short list of
55 articles. Short-listed studies’ reference lists
were consulted to identify yet undiscovered
studies. Google Scholar was used to identify
studies that cited short-listed studies. The
nal list of 90 articles combined the short list
and those identied through citation chasing.
Studies were then excluded if they met any of
these criteria: (a) they were literature reviews
or book chapters not reporting results, (b)
they reported on relationship satisfaction in
nonromantic relationships (e.g., friendships,
employee–employer), (c) they used an experi-
mental design (e.g., Bazzini et al., 2007),3(d)
they measured intimacy or closeness in rela-
tionships in general but not specically with
romantic partners (e.g., Hampes, 1994), and
(e) correlations were not reported (e.g., only
regression results were reported), and attempts
to contact the authors were not successful
(Lipsey & Wilson, 2001).
The nal set of studies included 43 distinct
samples that were drawn from 39 manuscripts.
One manuscript reported two samples (i.e.,
Saroglou et al., 2010), and one manuscript
reported four samples (i.e., Caird, 2015).
Manuscripts were published between 1985
and 2016 (Myear =2006). Most manuscripts
were peer-reviewed publications (n=23), but
asubstantialminorityweredissertationsand
master’s theses of sufcient quality to merit
inclusion (n=16). The substantial number of
non-peer-reviewed publications decreases the
chances of publication bias when reporting
overall weighted mean effect sizes and allows
for direct moderator tests of publication bias
3. Experimental studies were excluded because they
reported group contrasts rather than associations
between variables (Lipsey & Wilson, 2001).
(Card, 2011; Lipsey & Wilson, 2001). Nearly
all effect sizes came from samples where
humor was self-reported and/or perceived
in one’s partner (n=39), but a few were
observed during couples’ interactions (n=4).
Observational studies were included because
they estimated a comparable effect size with
self-report measures, and their results help
to broaden the scope of the meta-analysis.
Humor was often reported in a relational
context (n=21) or in general (n=18) or both
(n=2). Two studies reported humor use in
conict, one after relational transgressions and
one in the bedroom.
Humor measures were coded for all stud-
ies: 29.5% of manuscripts used the HSQ,
11.4% of manuscripts used the RHI, 9.1% of
manuscripts used HO, 6.8% of manuscripts
used the MSHS, 4.5% used Ziv and Gadish’s
(1989) scale, and 38.6% used some other
measure. Measures of relational satisfaction
included the Relationship Assessment Scale
(36.4%; Hendrick, Dicke, & Hendrick, 1998)
and the Dyadic Adjustment Scale (29.5%;
Spanier, 1976). These measures use items
assessing satisfaction with one’s partner and
relationship, comparing one’s partner to an
ideal, reporting a willingness to work on the
relationship, and a lack of conict. Although
no other measure was used more than twice,
an additional 14% of studies used published
measures of marital satisfaction. The remain-
ing 19% used measures that were conceptually
similar to satisfaction (e.g., love, closeness,
marital well-being). The inclusion of these
related measures broadens the scope and
increases the generalizability of the ndings
reported here (Card, 2011).
Coding subdimensions of humor
At the conclusion of study identication,
we identied all instruments and items used
to measure humor by subdimension. In 39
manuscripts, there were 36 unique subdimen-
sions of humor measured. The three theoretical
dimensions (i.e., relationship/partner/within-
person, positive– negative, and functional–
content free) were used to categorize these
measures of humor, and a total of 12 subdi-
mensions were arrayed upon three dimensions.
Humor meta-analysis 7
Some cells yielded null sets. For example, the
evaluative dimension was always perceived
in one’s partner, not in oneself or in one’s
relationship. Other dimensions had more
than one subcategory of humor in the cell
(e.g., aggressive and inappropriate humor in
self-negative). To address the variety of func-
tional humor measures used across the studies,
auniquecategorywascreatedtogroupgeneral
measures of humor functions, while separate
categories were used for three functions: dis-
tancing, attacking, and self-enhancing/coping
humor. When humor evaluation, relational
humor, and/or humor production items were
all included in the same measure, the gen-
eral positive humor category was used. The
MSHS-attitude subdimension was not included
as it measures the degree to which one has a
positive evaluation of humor itself, which is
conceptually distinct from humor production
and appreciation (Kohler & Ruch, 1996).
We establi s h e d a n a l l i s t o f s u b c a t e g o r i e s .
Conceptual denitions were created in con-
sultation with the scale’s original authors’
denitions of each subdimension. Whenever
possible, all of the items used to measure each
subdimension were identied. To establish
intercoder reliability, two graduate students
then independently coded all 36 unique
measures of humor by matching the humor
subdimension and items with one of the
12 conceptually dened subdimensions of
humor. Reliability was calculated using Hayes
and Krippendorff’s (2007) alpha MACRO
for SPSS and was .85. Disagreements were
resolved by discussion. Table 1 provides a list
of 12 subdimensions, a conceptual denition,
and all of the measures or studies that were
coded as tting the subdimension.
Calculating effect sizes
When a single effect size was reported in
the original studies, one effect size was used.
When the study measured humor in more than
1ofthe12subdimensions,forself-reportedor
partner-perceived or for both actor and partner
effects, multiple effect sizes were calculated
from each sample. Although the practice of
using multiple effect sizes from the same study
violates the assumption of the independence
of samples, Card (2011) recommends this
practice to explore categorical moderators:
It is reasonable to expect multiple effect
sizes from the same study to be more similar
(i.e., positively correlated) than independent
effect sizes, the impact of this interdepen-
dence will be to attenuate between-group
differences. Therefore, violation of the
independence assumption in this case is
likely to impose a conservative bias (i.e.,
increase in Type II error rates). (p. 227)
Only one effect size per sample was
included for each subcategory of humor. For
example, when effect sizes were reported
separately by humor function (e.g., Cleaver,
1991) or by production (e.g., HSQ-afliation
and HO) in the same study, the aggregate
effect size for that subdimension was used
(Card, 2011). A combined effect size was not
included in the meta-analyses when subdimen-
sions were already represented in the separate
subcategories (e.g., MSHS-combined).
Once all effect sizes were identied, effect
sizes and standard errors were corrected for
the attenuation of measurement. Artifact cor-
rection for reliability of measurement permits
the estimation of effect sizes as they would
appear under ideal circumstances (Lipsey &
Wilson, 2001). In studies where reliability
was reported, either by Cronbach’s alpha
or Cohen’s intercoder kappa, measurement
attenuation was corrected. In studies where
single-item measures were used and stud-
ies where reliability was not reported, but a
multiple-item measure was used, there was
no correction. To see a table summarizing all
effect sizes by study, follow this link: http://
hdl.handle.net/1808/21559.
Heterogeneity
After conducting an overall meta-analysis of
the association between humor use and rela-
tionship satisfaction, a separate meta-analysis
was performed for each of the 12 subcate-
gories of humor. Both procedures used sample
weighted mean effect size estimates (Lipsey
&Wilson,2001).Arandomeffectsmodelwas
tested for the overall meta-analysis and each
8J. A. Hall
Tabl e 1 . Denitions and categorization of measures
Subcategory Conceptual denition Scales or studies
Positive
Relational The degree to which humor is
shared, intentionally created, or
sought out and laughing together
occurs in relationships
RHI—self-positive; Jacobs (1985),
Cleaver (1991), Raniseski (1998)
Evaluative Judgment of a person’s sense of
humor in quality, capability, and
value
RHI—partner positive; Rust and
Goldstien (1989), Weisfeld and
colleagues (2011)
General positive Positive humor use in a relationship
setting, including measures
reporting on a mix of positive,
evaluative, appreciative, and
relational humor in the same
measure
RHI—couple; Germaine (2010),
Williams (1991), La France and
Hall (2012)
Self-positive
Production The ability to produce humorous
communication or actions, to
make jokes, tell funny stories,
and make others laugh
HO; MSHS—production;
HSQ—afliative; Hall and
Sereno (2010), Ziv and Gadish
(1989)—creation
Appreciation Someone who enjoys laughing,
laughs easily, likes jokes, and is
generally a good and receptive
audience to others’ jokes
MSHS—appreciation; Ziv and
Gadish (1989)—appreciation
Functional-positive
Relationship The intended use of a humorous
message or behavior to achieve a
communicative goal specic to
the romantic relationship,
including apologizing,
conict/stress mitigation, sexual
expression, and problem solving
RHI—self-instrumental,
RHI—partner instrumental;
Jacobs (1985), Smith (1993),
Cleaver (1991), Hall (2013)
Coping Humor used to cope, to release
stress and tension, to make light
of things, or nd amusement in
life
MSHS—coping/adaptive;
HSQ—self-enhancing;
SHRQ—sense of humor
Functional-negative
Attack partner Humor used as a means of
expressing hostility toward one’s
partner, including to demean,
belittle, put down, tease, or mock
RHI—self-negative; RHI— partner
negative; Campbell and
colleagues (2008), Jacobs (1985),
La France and Hall (2012),
Val l ade , B oo t h -B u t te r el d , a nd
Vel a ( 201 3 )
Distancing Humor used to push relational
partners apart, to avoid one’s
partner, or to deect
conversations about the
relationship
Butzer and Kuiper (2008),
Raniseski (1998)
Humor meta-analysis 9
Tabl e 1 . Continued
Subcategory Conceptual denition Scales or studies
Self-negative
Aggressive Humor can be used to demean, belittle,
attack, put down, tease, or mock
others outside of the relationship
HSQ—aggressive
Earthy
inappropriate
Negative humor that is inappropriate
or in poor taste
Craik and Ware (1998) — earthy;
Hall and Sereno (2010)
Self-defeating Making oneself the object of others’
humor or making fun of oneself,
self-deprecating humor
HSQ—self-defeating
Note.HSQ=Humor Styles Questionnaire; HO =humor orientation; MSHS =Multidimensional Sense of Humor Scale;
RHI =Relational Humor Inventory; SHRQ =Situational Humor Response Questionnaire.
of the 12 subdimensions, and the weighted
mean effect and condence interval were
estimated for each subdimension. Effect sizes
were evaluated for heterogeneity. When the
variance between effects was attributable to
sampling error alone (i.e., nonsignicant Q
statistic), a lack of variation among effect
sizes was assumed, and the moderators were
not explored. When effect sizes were het-
erogeneous (i.e., signicant Qstatistic), the
self-reported/partner-perceived moderating
analysis was conducted. Subsequently, sev-
eral continuous moderators were explored
by sample characteristics (i.e., age, percent
non-White), relationship characteristics (i.e.,
length, percent married), and publication
source (i.e., published/not published).
When conducting a meta-analysis that tests
multiple moderators of an association, includ-
ing multiple subdimensions or categories, it
is important to establish reasonable coverage.
Although there is no clear guidance on what
constitutes adequate coverage in terms of the
number of effect sizes, Card (2011) recom-
mends that ve studies is adequate. As the goal
of this meta-analysis is to offer a summative
point of reference for future research, a lack of
coverage is in itself illuminating. Therefore,
weighted mean effect sizes for inappropriate
and distancing humor, humor appreciation,
and four of the partner effects (i.e., relational,
attack partner, aggressive, inappropriate)
should be interpreted with caution due to less
than adequate coverage.
Results
The nal sample included 15,177 partici-
pants (54.7% female), and the mean sample
size was 366. Mean age was reported in all
but 3 studies. Participants were on average
34.0 years of age (Msample range =18.6–58
years). Sample race/ethnicity was reported in
28 samples, and 31.4% of participants were
non-White. Six samples recruited participants
from countries exclusively or primarily outside
of the United States. The average length of
relationship was reported in 32 samples and
was 11.44 years (Msample range =1–33 years).
All studies required participants to report on
romantic relationship partners. Of the 40 sam-
ples reporting marital status of participants,
82.8% of participants were married (n=17:
100% married, n=13: 0% married).4
Overall meta-analysis
The weighted mean effect size for humor
and relationship satisfaction for the entire
sample was r=.22, 95% CI [.218, .220],
k=43, N=15,177, with substantial het-
erogeneity, Q=4,954.5, df =42, p<.001.
All three theoretical moderators were ana-
lyzed simultaneously using sample weighted
regression procedures in SPSS (Card, 2011),
4. The high proportion of participants who were married
was partly due to one very large study of married
participants measuring humor evaluation (i.e., Weisfeld
et al., 2011).
10 J. A. Hall
and each uniquely moderated the weighted
mean effect sizes, Q=1,293, df =4, p<.001.
The within-person versus relational dimension
moderated the effect, β=.04, SE =.02, p<.01,
wherein relational humor was more strongly
associated with relationship satisfaction (H1).
The positive versus negative dimension mod-
erated the effect, β=.32, SE =.02, p<.001,
wherein the positive dimensions were posi-
tively associated with relationship satisfaction
(H2a–b). The functional/instrumental ver-
sus content-free dimension moderated the
effect, β=−.32, SE =.03, p<.001, wherein
content-free dimensions were more strongly
associated with satisfaction (RQ2).
Categorical meta-analysis
To explore associations by category, a categor-
ical meta-analysis was conducted (Table 2).
Results offered support for H1, in that the three
relational forms of positive humor showed the
largest effect sizes— each outside of the 95%
condence interval of the within-person or
self-focused forms of positive humor. The
only exception to this trend was evaluative
humor, which was exclusively perceived in
one’s partner and showed the strongest associ-
ation with relational satisfaction. In response
to H2a, all six positive forms of humor (i.e.,
production, appreciation, relational evaluation,
general positive, coping) showed positive
associations with relationship satisfaction.
In response to H2b, four of the ve nega-
tive types of humor (i.e., distancing, attack
partner, aggressive, self-defeating) showed
negative associations with relationship satis-
faction. Earthy or inappropriate humor was
unassociated with relationship satisfaction.
Furthermore, relational negative humor types
(i.e., attack partner, distancing) were more
strongly associated with relationship satisfac-
tion than humor reported within the respondent
(i.e., aggressive, self-defeating). In response
to RQ2, instrumental humor was unassociated
with relationship satisfaction. The association
between distinct functions (e.g., coping, dis-
tancing) and satisfaction were consistent with
the affective and relational dimensions.
Partner effects measure the association of
one partner’s humor with the other partner’s
relationship satisfaction when data are col-
lected from paired dyads. In response to
RQ1, four of the seven partner effects were
signicantly associated with satisfaction:
humor production, relational humor, aggres-
sive humor (negatively), and self-enhancing or
coping humor (Table 3).
Moderators
Self versus partner perception
When there was signicant heterogeneity in
effect sizes, a self-reported, partner-perceived
moderator analysis was conducted using
sample weighted regression procedures
on actor effects (Lipsey & Wilson, 2001).
Signicant differences exist when the hetero-
geneity of the model was signicant at df =1.
Partner-perceived humor was more strongly
associated with relationship satisfaction than
self-reported humor for all forms of positive
humor: humor production, humor apprecia-
tion, relational/shared humor, general positive
humor, and self-enhancing or coping humor.
Partner-perceived aggressive humor was more
negatively associated with relationship satis-
faction than self-reported aggressive humor
(Table 3).
Continuous moderators
Several continuous moderators were
explored by sample characteristics (i.e., age,
race/ethnicity), relationship characteristics
(i.e., length, percent married), and publica-
tion source (i.e., published/not published).
Using SPSS procedures (Card, 2011), sig-
nicant moderators were explored for all
heterogeneous weighted mean effect sizes
where coverage was adequate. Moderators
were detected in four subcategories: humor
production, evaluation of partners’ humor,
relational humor, and coping humor.
Moderator analyses demonstrated a publi-
cation bias and, to a lesser degree, a bias by
year of publication. Published studies reported
larger mean effect sizes for humor produc-
tion, β=.19, SE =.07, Qpub =45.38, df =1,
p<.001; relational humor, β=.37, SE =.09,
Qpub =110.08, df =1, p<.001; and coping
humor, β=.16, SE =.05, Qpub =20.50, df =1,
Humor meta-analysis 11
Tabl e 2 . Meta-analyses results by humor category
95% condence
Humor type NkQAdj SE r Lower Upper
Positive
Relational
Relational 2,758 16 191.5.020 .456 .418 .494
Evaluative 7,377 6 102.9.012 .647 .623 .671
General 2,630 8 39.8.021 .437 .396 .478
Self
Production 3,760 17 98.4.017 .234 .200 .267
Appreciation 1,006 3 23.8.032 .227 .163 .291
Functional
Relationship 2,036 9 176.3.020 .022 −.016 .060
Self-coping 2,825 14 61.6.019 .146 .108 .183
Negative
Relational
Attack Partner 3,144 15 118.9.016−.258 −.290 −.226
Distancing 416 3 31.2.048−.261 −.354 −.167
Self
Aggressive 2,123 9 30.9.024−.159 −.206 −.111
Inappropriate 538 3 24.3.031 .049 −.011 .109
Self-defeating 2,351 11 32.8.022−.114 −.156 −.071
Note. All heterogeneity estimates (Q)signicantatp<.001; bolded rsignicant at p<.05.
p<.001. More recent studies, compared to
older studies, reported larger mean effects
for humor production, β=.001, SE =.001,
Qyear =40.38, df =1, p<.001; evaluation
of partners’ humor, β=.001, SE =.001,
Qyear =98.65, df =1, p<.001; and coping
humor, β=.001, SE =.001, Qyear =26.25,
df =1, p<.001.
For humor production, percent of sample
married, β=−.001, SE =.001, Q%m =14.84,
df =1, p<.001, was negatively related to
weighted mean effect size, wherein samples
with more married couples showed smaller
effect sizes than samples with more dating
couples. Mean sample age, β=−.01, SE =.00,
Qage =70.16, df =1, p<.001, was related
to weighted mean effect size, wherein older
samples were more likely to show weaker
effect sizes than younger samples. Mean
sample length of relationship, β=−.01,
SE =.00 Qage =19.76, df =1, p<.01, was
related to weighted mean effect size, wherein
more established couples were more likely
to show weaker weighted mean effect sizes
than recently involved couples. Multiple
regression of all three moderators suggested
that mean sample age was the best pre-
dictor of the moderation of the association
between humor production and relationship
satisfaction, β=−.02, SE =.01, Qage =9.09,
df =1, p<.001, controlling for the other two
moderators.
For humor evaluation, mean sample
age, β=.02, SE =.01, Qage =35.89, df =1,
p<.001, was related to weighted mean effect
size, wherein older samples were more likely to
show larger effect sizes than younger samples.
Mean sample length of relationship, β=.03,
SE =.01, Qlength =37.32, df =1, p<.01, was
related to weighted mean effect size, wherein
more established couples were more likely to
show larger effect sizes than recently formed
couples. When both moderators were entered
into a single regression model, neither was sig-
nicant, probably due to considerable shared
variance.
Coping or adaptive humor showed a
very similar pattern of results as humor
12 J. A. Hall
Tabl e 3 . Moderators of actor effects, self-reported, partner-perceived, and partner effects
Actor effects Partner effects
Humor type Qr-Self r-Partner kN SE r Upper Lower
Positive
Relational
Relational 14.1** .454 .642 2187.035 .348 .279 .416
General 13.3** .329 .522
Self
Production 10.3** .208 .334 61226.032 .152 .089 .215
Appreciation 9.7** .109 .334
Functional
Relationship 3.8
Coping 5.9* .121 .228 5788.037 .089 .016 .161
Negative
Relational
Attack partner .52187.036−.047 −.118 .024
Distancing .4
Self
Aggressive 6.1* −.126 −.270 3456.062−.139 −.261 −.037
Inappropriate .01 3 538 .043 .048 −.037 .133
Self-defeating .01 4 560 .053 −.043 −.147 .061
Note. All heterogeneity estimates (Q)signicantat*p<.05, **p<.01; bolded rsignicant at p<.05.
production. Percent sample married was neg-
atively related to weighted mean effect size,
β=−.001, SE =.001, Q%m =7.68, df =1,
p<.01, wherein samples with more married
couples showed weaker weighted mean effect
sizes than those with more dating couples.
Mean sample age, β=−.005, SE =.002,
Qage =8.87, df =1, p<.01, was related to
weighted mean effect size, wherein older
samples were more likely to show weaker
weighted mean effect sizes than younger
samples. When both moderators were entered
into a single regression model, neither was
signicant.
Discussion
Through a series of categorical meta-analyses,
this manuscript provides a nuanced and thor-
ough investigation of the role of humor in
romantic relationships and offers empirical
support for the relevance of the relational
versus within-person dimension of humor.
Results suggest that humor’s role in romantic
relationships can be understood primarily as a
function of two dimensions: relational versus
within-person and positive versus negative.
As illustrated in Figure 1, when humor is
more relationship oriented, its association
with relationship satisfaction increases, often
demonstrating large sample weighted mean
effect sizes. Self-rated humor was associated
with relationship satisfaction, but the strength
of that association was signicantly larger
when the same type of humor was evaluated in
romantic partners and when the type of humor
was directly relevant to the relationship itself.
Positive humor
Both of the content-free and self-rated positive
forms of humor analyzed herein showed small
weighted mean effect sizes with relationship
satisfaction: humor production, r=.21, and
humor appreciation, r=.11. There are sev-
eral reasons why individuals who are able to
produce and are appreciative of humor might
be more satised in romantic relationships.
These measures may tap into an underlying
social facility (Storey, 2003), ability to connect
or develop intimacy with others (Hampes,
1994), or cheerful disposition (Kohler & Ruch,
Humor meta-analysis 13
Mean Effect Size
-0.35 -0.175 0 0.175 0.35 0.525 0.7
Distancing
SR Appreciation
PP General Positive
PP Relational
Instrumental Functional Relational
Attack
Partner
SR Aggressive
Inappropriate
PP Evaluative
SR General Positive
SR Self
Defeating
SR Self-
EnhancingSR Production SR Relational
PP Appreciation
PP Production
PP Self-
Enhancing
PP Aggressive
Figure 1. Results of meta-analyses by weighted mean effect size. Size of circle corresponds to
total sample size of subdimension. SR =self-reported; PP =partner perceived.
1996). As such, individuals who can create
and appreciate humor may nd relationships
more pleasant and satisfying in general.
When these two types of humor were
perceived in participants’ romantic partners,
effects were medium sized: humor production,
r=.33, and humor appreciation, r=.32. This
trend was also reected in the general positive
humor category. This category included mea-
sures that combined elements of production,
appreciation, and/or evaluation. When eval-
uated in oneself, there was a medium-sized
effect, r=.33, but when evaluated in one’s
partner, the effect size was large, r=.52.
This points to an important trend, noted by
several past scholars (e.g., La France & Hall,
2012; Rust & Goldstien, 1989), that humor
perceived in one’s partner is a better predictor
of one’s relationship satisfaction than one’s
own self-reported humor. As the primary audi-
ence to and cocreator of humor (Alberts et al.,
2005; Hall & Sereno, 2010), perceptions of
one’s partner and his or her behavior matters
greatly when understanding humor’s associ-
ation with relationship satisfaction (Saroglou
et al., 2010).
Large weighted mean effect sizes were
detected for relational humor, r=.64, which
is humor that is created and shared between
partners, and humor evaluation, r=.65, which
is one partner’s judgment of the other partner’s
sense of humor. Having fun, being playful, and
being open to joking around are key functions
of humor in romantic relationships (Ziv,
1988). Because evaluative measures of humor
are notoriously vague, Martin (1998) notes
that saying that someone has a good sense of
humor may reect a general glow of positivity
about a person. When one believes his or her
romantic partner has a good sense of humor, he
or she may be making a blanket judgment that
the partner is simply a good person, a good
spouse, or the relationship is good. Top-down
positive processing, called the “halo error,”
may have articially inated the association
between relationship satisfaction and partner
humor reported here (Feeley, 2002). This inter-
pretation is supported by the strongly negative
evaluations of partners’ sense of humor for
divorced couples (Saroglou et al., 2010).
The importance of the within-person ver-
sus relational dimension can be seen in the
partner effects. While infrequently reported, a
revealing pattern of results emerged. Speci-
cally, more self-directed types of humor (i.e.,
production, r=.15; aggressive, r=−.14) were
weakly associated with the other partners’
satisfaction. By contrast, one partners’ rela-
tional humor showed a medium effect size
with the other partners’ satisfaction, r=.35.
Consistent with its conception as a primarily
self-focused process (Caird & Martin, 2014),
self-enhancing humor was expected to exert
asmalleffectonone’spartner,andresults
demonstrate that humorous coping was weakly
associated with partners’ satisfaction, r=.09.
14 J. A. Hall
Although partner effects are typically weaker
than actor effects in dyadic studies of humor
(e.g., Hall, 2013), results support the conclu-
sion that self-directed and self-reported humor
is less relationally consequential than rela-
tional or partner-perceived humor.
Negative humor
Self-reported self-defeating humor, r=−.12,
and aggressive humor, r=−.13, were both
negatively associated with relationship sat-
isfaction and showed small effect sizes.
Inappropriate or earthy humor was unassoci-
ated with satisfaction. Consistent with positive
humor ndings, when aggressive humor
was perceived in one’s partner, it showed a
medium-sized negative association with satis-
faction, r=−.27. Aggressive humor reects a
willingness to attack and make fun of others
(Martin et al., 2003). When perceived in one’s
partner, this may reect a judgment of his/her
character or disposition, and it has implica-
tions for the couple’s relationship-specic face
(Hall, 2011). The two types of negative humor
directed at one’s partner specically showed
medium negative weighted mean effects as
well: attacking humor, r=−.25, and distanc-
ing humor, r=−.26. Whether these types of
humor were self-reported or perceived in one’s
partner, the effects were similar sized. Attack-
ing and distancing humor probably reect
dissatisfaction with a relationship partner or
the relationship itself. At least in terms of
effect size, the deleterious effect of attacking
one’s partner with humor or using humor
to distance oneself is indistinguishable from
perceiving that one’s partner has an aggressive
style of humor. Perhaps functional uses of
humor emerge from one’s internal humor style
(Betcher, 1981; Hall, 2013).
Functional humor
Functional or instrumental humor was the only
type of relational humor unassociated with
satisfaction. Several relationship functions
are rolled into commonly used measures of
instrumental humor use, which likely attenu-
ated effect sizes. Additionally, humor used to
manage conict, express attitudes, or apologize
may be indicative of prevalent relationship
problems (Campbell et al., 2008). Betcher
(1981) suggests that the function humor serves
is extrinsic to the value of relational humor.
That relational partners would use humor
instrumentally is an outgrowth of successfully
cultivating humor in a relationship, but the use
of humor in this fashion does not positively
inuence the relationship when considered in
isolation. This explanation is supported by the
mediating effect of positive humor functions
in explaining the relationship between humor
style and satisfaction (Hall, 2013).
Self-enhancing humor, or humor used to
cope or adapt, was weakly associated with sat-
isfaction when self-reported, r=.12, and more
strongly so when perceived in romantic part-
ners, r=.23. This supports past assertions that
this type of humor could be particularly valu-
able in managing stressors naturally arising out
of life’s challenges, including stressors arising
from one’s relationship (Cann et al., 2011).
Continuous moderators
When effect sizes were reported in older,
longer established, and married samples—
compared to younger, newly dating, or
unmarried samples—the association between
satisfaction and humor production and coping
humor were attenuated. By contrast, as age
and relationship length of samples increased,
the association between humor evaluation and
satisfaction was larger. Without precedent for
these ndings, two speculative explanations
are offered. The stronger association between
humor and satisfaction found in young,
unmarried samples points to the importance
of producing humor in the initial and devel-
oping stages of romance. Due to its ability
to facilitate bonding and develop attraction
(Hall, 2015), humor may be more valuable
in keeping newly formed couples satised.
In enduring, married relationships, producing
humor and using it to cope are still important
predictors of relationship satisfaction (Lauer
et al., 1990), just not as strongly as they are in
younger, unmarried relationships. By contrast,
evaluating humor positively in one’s partner
becomes a stronger predictor of satisfaction
for older adults in long-lasting relationships.
Humor meta-analysis 15
Perhaps humor production and coping humor
matter more in developing a relationship, but
evaluating it positively in one’s partner matters
more once the relationship is well established.
Longitudinal studies have demonstrated
that a sense of humor is benecial in early
life, especially for building bonds with others,
but is not predictive of longevity and health
(Friedman & Martin, 2011). Perhaps the detri-
mental correlates of humor production (e.g.,
alcohol and tobacco use; Friedman & Martin,
2011) only become problematic at older ages,
particularly for health and well-being. Future
longitudinal studies could explore if either of
the above explanations stand up to scrutiny.
Limitations and future directions
This manuscript was unable to test a reverse
causal model, namely, that relationship satis-
faction results in different types of humor use.
This meta-analysis was also limited by the rel-
ative lack of coverage for certain categories of
humor, especially humor appreciation, distanc-
ing humor, and earthy or inappropriate humor.
Additionally, there is a long history of studying
humor in marriage, but only very recently have
gay and lesbian couples been allowed to marry
in the United States. Only one study included
here (Germaine, 2010) examined the role of
humor in same-sex romantic relationships.
Future research should disentangle the
role of humor from the role of personality
in bringing about relationship satisfaction.
Recent work using repeated measures of
humor and satisfaction (Caird & Martin,
2014) and observations of couple’s interac-
tions (Kurtz & Algoe, 2015) provide valuable
information on the unique role of humor in
explaining changes in satisfaction. Although
it is widely understood that humor is a clus-
ter of behaviors and traits (Kohler & Ruch,
1996), it is not well understood if positive
humor is merely a signal of underlying char-
acteristics that are protable in relationships
or whether humor itself has unique value.
Is coconstructed relational humor’s strong
association with relationship satisfaction a
fortunate consequence of partners possessing
valuable traits or a consequence of humor
itself? Similarly, aggressive or self-defeating
humor may not in itself diminish satisfaction
but may be indicative of personality traits that
result in less satisfying relationships.
The studies used in these meta-analyses
are overwhelmingly from self-reported and
partner-perceived measures of humor, with
only 9.5% of samples from observational
studies. However, humor can be measured in a
variety of other reliable and valid ways. Humor
production and appreciation are best differen-
tiated in behavioral rather than self-reported
measures (Kohler & Ruch, 1996). Yet, solely
behavioral, non-self-reported measures of
humor in the context of romantic relationship
initiation, development, or longevity were not
located. Furthermore, 21% of studies included
here used measures of humor that combined
production, appreciation, and evaluation,
which reects a similar measurement problem
as the combination of the various functions
of humor into an omnibus measure. The null
results of functional humor here may have
more to do with inadequate measurement
than with the absence of the benet of humor
used instrumentally. Objective measures could
also attenuate the halo error when measuring
partners’ perceptions of one another (Feeley,
2002). The conclusion of this investigation
concurs with the suggestions of Kohler and
Ruch (1996):
Research should be focused on at least two
goals: (1) theoretical and empirical work
aimed at a more precise outline and de-
nition of the construct; that is, identica-
tion of a more precise outline and deni-
tion of the construct, (2) application of a
more sophisticated technology of construct-
ing instruments for the assessment of the
sense of humor and/or the components of
this construct. (p. 392)
Both suggestions could help advance
the future study of humor in romantic
relationships.
References
Studies denoted with an asterisk were included in this
meta-analysis.
Alberts, J. K., Yoshimura, C. G., Rabby, M., & Loschiavo,
R. (2005). Mapping the topography of couples’ daily
16 J. A. Hall
conversation. Journal of Social and Personal Relation-
ships,22, 299– 322. doi:10.1177/0265407505050941
*Ammerman, M. C. (2002). Gender differences in the use
of humor in intimate relationships and marital adjust-
ment (Unpublished doctoral dissertation). Pepperdine
University, Malibu, CA.
*Aune, K. S., & Wong, N. C. H. (2002). Antecedents
and consequences of adult play in romantic rela-
tionships. Personal Relationships,9,279–286.
doi:10.1111/1475-6811.00019
*Bacanli, H., ˙
Ilhan, T., & Gungor, H. C. (2012). ˙
Ili¸skisel
mizah envanterinin Türkçe’ye uyarlanması: Geçer-
lik ve güvenirlik calı¸sması [Adaptation of relational
humor inventory to Turkish: A validity and reliability
study]. Turk Psikolojik Danisma ve Rehberlik Dergisi,
4,11–23.
*Bacon, C. M. (2010). Humor as a communication strat-
egy in military marriages: The inuences of its uses
and perceptions on marital satisfaction (Unpublished
doctoral dissertation). Retrieved from ProQuest Disser-
tations and Theses database. (UMI No. 3421079)
*Barelds, D. P. H., & Barelds-Dijkstra, P. (2010). Humor
in intimate relationships: Ties among sense of humor,
similarity in humor, and relationship quality. Humor,
23, 447– 465. doi:10.1515/humr.2010.021
Bazzini, D. G., Stack, E. R., Martincin, P. D., & Davis, C.
P. (2007). The effect of reminiscing about laughter on
relationship satisfaction. Motivation and Emotion,31,
25– 34. doi:10.1007/s11031-006-9045-6
Betcher, R. W. (1981). Intimate play and marital adapta-
tion. Psychiatry,44,13–33.
Bippus, A. M. (2000). Making sense of humor
in young romantic relationships: Understand-
ing partners’ perceptions. Humor,13, 395 –417.
doi:10.1515/humr.2000.13.4.395
Booth-Buttereld, S., & Booth-Buttereld, M. (1991). The
communication of humor in everyday life: The use of
humorous messages. Southern Communication Jour-
nal,56, 205–218.
*Butzer, B., & Kuiper, N. A. (2008). Humor use in roman-
tic relationships: The effects of relationship satisfaction
and pleasant versus conict situations. Journal of Psy-
chology,142, 245– 260.
*Caird, S. (2015). An examination of daily humour styles
and relationship satisfaction in dating couples (Paper
No. 3372). Electronic Thesis and Dissertation Reposi-
tory. Retrieved from http://ir.lib.uwo.ca/etd/3372
*Caird, S., & Martin, R. A. (2014). Relationship-focused
humor styles and relationship satisfaction in dating
couples: A repeated-measures design. Humor,27,
227– 247. doi:10.1515/humor-2014-0015
*Campbell, L., Martin, R. A., & Ward, J. R. (2008). An
observational study of humor use while resolving con-
ict in dating couples. Personal Relationships,15,
41– 55. doi:10.1111/j.1475-6811.2007.00183.x
*Cann, A., Norman, M. A., Welbourne, J. L., & Calhoun,
L. G. (2008). Attachment styles, conict styles and
humor styles: Interrelationships and associations with
relationship satisfaction. European Journal of Person-
ality,22, 131– 146. doi:10.1002/per.666
*Cann, A., Zapata, C. L., & Davis, H. B. (2009). Positive
and negative styles of humor in communication.
Evidence for the importance of considering both
styles. Communication Quarterly,57, 452 – 468.
doi:10.1080/01463370903313398
*Cann, A., Zapata, C. L., & Davis, H. B. (2011). Humor
style and relationship satisfaction in dating cou-
ples: Perceived versus self-reported humor styles
as predictors of satisfaction. Humor,24,1–20.
doi:10.1515/humr.2011.001
Card, N. A. (2011). Applied meta-analysis for social sci-
ence research.NewYorkCity,NY:Guilford.
Caron, J. E. (2002). From ethology to aesthetics: Evolu-
tion as a theoretical paradigm for research on laugh-
ter, humor, and other comic phenomenon. Humor,15,
245– 281. doi:10.1515/humr.2002.015
*Carstensen, L. L., Gottman, J. M., & Levenson, R. W.
(1995). Emotional behavior in long-term marriage.
Psychology and Aging,10, 140– 149.
*Cleaver, D. R. (1991). Humor and marital adjustment.
(Unpublished master’s thesis), California State Univer-
sity, Fullerton.
Craik, K. H., & Ware, A. P. (1998). Humor and person-
ality in everyday life. In W. Ruch (Ed.), The sense
of humor: Explorations of a personality characteristic
(pp. 63– 94). New York City, NY: Mouton de Gruyter.
Darwin, C. (1859). On the origin of species. London,
England: Murray.
*de Koning, E., & Weiss, R. L. (2002). The relational
humor inventory: Functions of humor in close relation-
ships. American Journal of Family Therapy,30,1–18.
doi:10.1080/019261802753455615
Feeley, T. H. (2002). Comments on halo effects in rat-
ing and evaluation research. Human Communication
Research,28, 578– 586.
Friedman, H. S., & Martin, L. R. (2011). The longevity
project. London, England: Hudson Street Press.
Graham, E. E., Papa, M. J., & Brooks, G. P. (1992).
Functions of humor in conversation: Conceptualization
and measurement. Western Journal of Communication,
56, 161–183. doi:10.1080/10570319209374409
Hall, J. A. (2011). Is it something I said? Sense of
humor and partner embarrassment. Journal of
Social and Personal Relationships,28,383–405.
doi:10.1177/0265407510384422
*Hall, J. A. (2013). Humor in long-term romantic rela-
tionships: The association of general humor styles
and relationship-specic functions with relationship
satisfaction. Western Journal of Communication,77,
272– 292. doi:10.1080/10570314.2012.757796
Hall, J. A. (2015). Sexual selection and humor in courtship:
A case for warmth and extraversion. Evolutionary Psy-
chology,13, 1– 10. doi:10.1177/1474704915598918
*Hall, J. A., & Sereno, K. (2010). Offensive jokes: How
do they impact long-term relationships? Humor,23,
351– 373. doi:10.1515/humr.2010.016
Hampes, W. (1994). Relation between intimacy
and the Multidimensional Sense of Humor
scale. Psychological Reports,74, 1360– 1362.
doi:10.2466/pr0.1994.74.3c.1360
Humor meta-analysis 17
Hayes, A. F., & Krippendorff, K. (2007). Answering the
call for a standard reliability measure for coding data.
Communication Methods & Measures,1,77–89.
Hendrick, S. S., Dicke, A., & Hendrick, C. (1998). The
Relationship Assessment Scale. Journal of Social and
Personal Relationships,15, 137–142.
*Jacobs, E. C. (1985). The functions of humor in marital
adjustment (Dissertation, New School for Social
Research). Dissertation Abstracts International,
46(5B), 1688. (AAT No. 8514188)
*Johari, M. (2004). Humor and marital quality: Is humor
style associated with marital success? (Unpublished
master’s thesis). Universityof Western Ontario, Wilfrid
Laurier University, London, Canada.
Kohler, G., & Ruch, W. (1996). Sources of variance in cur-
rent sense of humor inventories: How much substance,
how much method variance. Humor,9,363–397.
doi:10.1515/humr.1996/9/3-4/363
*Kurtz, L. E., & Algoe, S. B. (2015). Putting laughter
in context: Shared laughter as behavioral indicator of
relationship well-being. Personal Relationships,22,
573– 590. doi:10.1111/pere.12095
*La France, B. H., & Hall, J. A. (2012). Sex, humor, and
intimacy: An examination of sexual humor use in close
relationships. In J. Meyer & R. L. DiCioccio (Eds.),
Humor communication: Theory, impact, and outcomes
(pp. 119– 140). Dubuque, IA: Kendall Hunt.
Lauer, R. H., Lauer, J. C., & Kerr, S. T. (1990). The
long-term marriage: Perception of stability and satis-
faction. International Journal on Aging and Human
Development,31, 189– 195.
Lipsey, M. W., & Wilson, D. B. (2001). Practical
meta-analysis.ThousandOaks,CA:Sage.
*Maki, S. M., Booth-Buttereld, M., & McMullen, A.
(2012). Does our humor affect us?: An examination of
a dyad’s humor orientation. Communication Quarterly,
60, 649– 664. doi:10.1080/01463373.2012.725006
*Manne, S., Sherman, M., Ross, S., Ostroff, J., Heyman, R.
E., & Fox, K. (2004). Couples’ support-related com-
munication, psychological distress, and relationship
satisfaction among women with early stage breast can-
cer. Journalof Consulting and Clinical Psychology,72,
660– 670. doi:10.1037/0022-006X.72.4.660
Martin, R. A. (1998). Approaches to the sense of humor:
A history review. In W. Ruch (Ed.), The sense of
humor: Explorations of a personality characteristic
(pp. 15– 62). New York City, NY: Mouton de Gruyter.
Martin, R. A., Puhlik-Doris, P., Larsen, G., Gray, J., &
Weir, K. (2003). Individual differences in uses of
humor and their relation to psychological well-being:
Development of the Humor Styles Questionnaire.
Journal of Research in Personality,37,48–75.
doi:10.1016/S0092-6566(02)00534-2
*Raniseski, J. M. A. (1998). Exploring the relationships
between humor and marital well-being. Dissertation
Abstracts International, 59, 3124.
*Rhoads, M. E. (2014). The effects of demand-withdraw
and instrumental humor on relationship satisfaction: A
comparison of individuals in same-and cross-sex rela-
tionships (Unpublished doctoral dissertation). Avail-
able from PsycINFO. (UMI No. 3636899)
*Rust, J., & Goldstien, J. (1989). Humor in marital adjust-
ment. Humor,2, 217–224.
*Saroglou, V., Lacour, C., & Demeure, M.-E. (2010). Bad
humor, bad marriage: Humor styles in divorced and
married couples. Europe’s Journal of Psychology,6,
94– 121. doi:10.5964/ejop.v6i3.210
*Smith, E. L. (1993). Humor creation and humor appre-
ciation as predictors of marital satisfaction (Unpub-
lished master’s thesis). Auburn University, Auburn,
AL.
Spanier, G. B. (1976). Measuring dyadic adjustment: New
scales for assessing the quality of marriage and similar
dyads. Journal of Marriage and the Family,38,15–28.
*St. Germaine, C. (2010). Predicting relationship satis-
faction: An analysis of romantic partners’ percep-
tions (Unpublished doctoral dissertation, AAI341
7158). Available from PsycINFO. (862786982; 2011-
99040-437)
Storey, R. (2003). Humor and sexual selection. Human
Nature,14, 319– 336. doi:10.1007/s12110-003-1009-x
Thorson, J., & Powell, F. (1993). Development and
validation of a multidimensional sense of humor
scale. Journal of Clinical Psychology,49,13–23.
doi:10.1002/1097-4679(199301)49:1
*Vallade, J. I., Booth-Buttereld, M., & Vela, L. E. (2013).
Taking back power: Using superiority theory to pre-
dict humor use following a relational transgression.
Western Journal of Communication,77, 231 – 248.
doi:10.1080/10570314.2012/669018
*Vela, L. E., Booth-Buttereld, M., Wanzer, M. B., &
Vallade, J. I. (2013). Relationships among humor,
coping, relationship stress, and satisfaction in
dating relationships: Replication and extension.
Communication Research Reports,30,67–75.
doi:10.1080/08824096.2012.746224
*Weisfeld, G. E., Nowak, N. T., Lucas, T., Weisfeld, C. C.,
Imamoglu, E. O., Butovskaya, M., …Parkhill, M. R.
(2011). Do women seek humorousness in men because
it signals intelligence? A cross-cultural test. Humor,
24, 435– 462. doi:10.1515/humr.2011/025
Wilbur, C. J., & Campbell, L. (2011). Humor in roman-
tic contexts: Do men participate and women evalu-
ate? Personality and Social Psychology Bulletin,37,
918– 929. doi:10.1177/0146167211405343
*Wilkinson, L. L. (2003). Love and laughter …amagi-
cal cure?: The effects of humor and relationship sat-
isfaction on chronic pain (Unpublished master’s the-
sis, Lamar University, Beaumont, TX). Retrieved from
ProQuest Dissertations and Theses database.
*Williams, K. M. (1991). Operative factors that contribute
to long-term satisfying marriages (Unpublished doc-
toral dissertation). Available from PsycINFO. (6181
96023; 1992-76794-001) Retrieved from http://search
.proquest.com/docview/618196023?accountid=14556
Ziv, A. (1988). Humor’s role in married life. Humor,1,
223– 230. doi:10.1080/00224545.1989.9712084
*Ziv, A., & Gadish, O. (1989). Humor and marital satis-
faction. Journal of Social Psychology,129,759–768.
doi:10.1515/humr.1990.3.3.247
