Cogn Ther Res (2009) 33:585–596

DOI 10.1007/s10608-008-9223-9

ORIGINAL ARTICLE

Early Maladaptive Schemas and Adaptive/Maladaptive

Styles of Humor
David J. A. Dozois Æ Rod A. Martin Æ
Peter J. Bieling

Published online: 2 December 2008

Ó Springer Science+Business Media, LLC 2008

Abstract Early maladaptive schemas (EMSs) are domains and depressed mood was mediated by both
thought to act as templates for information processing self-enhancing and self-defeating humor styles.
which influence individuals’ emotional reactions to
life situations and their styles of interpersonal relating. Keywords Schemas
Humor

The association between EMSs and psychopathology Early maladaptive schemas
Depression

is also believed to be mediated by the use of Dysphoria
maladaptive compensatory coping and deficits in
adaptive coping (e.g., avoidance, surrender). As styles
of coping, humor may be such a mediator. This study
examined correlations between domains of the Young According to Beck’s (Beck 1967; Beck et al. 1979)
Schema Questionnaire-Short form and subscales of the cognitive theory of depression, a negative or de-
Humor Styles Questionnaire. A number of EMS pressogenic self-schema constitutes a vulnerability
domains were associated with reduced use of adaptive factor for depression. Self-schemas are well-orga-
affiliative and self-enhancing humor styles and nized perceptual sets that form the basis of core
increased use of maladaptive self-defeating humor. beliefs about self and have been conceptualized as
In addition, the maladaptive aggressive style of humor ‘‘relatively enduring internal structures of stored
was associated with the EMS involving insufficient generic or prototypical features of stimuli, ideas, or
self-control. The relationship between most EMS experience that are used to organize new information
in a meaningful way thereby determining how
phenomena are perceived and conceptualized’’ (Clark
et al. 1999, p. 79). Beck argued that the development
D. J. A. Dozois
R. A. Martin and structure of a maladaptive self-schema occurs in
The University of Western Ontario, London, ON, Canada early childhood but remains dormant until it is
activated by negative life circumstances (see Beck
P. J. Bieling et al. 1979). Once activated, schemas are thought to
St. Joseph’s Healthcare Hamilton and McMaster
University, Hamilton, ON, Canada influence information processing in a negative and
schema-congruent fashion (Beck et al. 1979; Ingram
D. J. A. Dozois (&) et al. 1998). There is now substantial support for the
Department of Psychology, University of Western idea that maladaptive schemas and core beliefs are
Ontario, Westminster Hall, 361 Windermere Road,
Rm. 313, London, ON, Canada N6A 3K7 contributory causes of unipolar depression (see
e-mail: [email protected] Dozois and Beck 2008, for review).

123
586 Cogn Ther Res (2009) 33:585–596

Although Beck’s model contends that early expe- Overvigilance and Inhibition are thought to be
riences contribute to the development of negative cultivated in early environments that are demanding
schemas and, in turn, depression, less theoretical and rigid and where abiding by rules and avoiding
emphasis has been allocated to these developmental mistakes are rewarded to the exclusion of explorative
vicissitudes. The work of Jeffrey Young (Young or pleasure-seeking behaviors.
1999; Young et al. 2003, 2008) has recently shifted A number of studies have recently explored the
the conceptual focus toward developmental factors specific core beliefs outlined in the YSQ. The
involved in cognitive vulnerability to depression and empirical research generally supports the differentia-
empirical work in this area has increased (e.g., Bruce tion among the EMSs (e.g., Glaser et al. 2002; Lee
et al. 2006; Gibb et al. 2006; Lumley and Harkness et al. 1999; Schmidt et al. 1995). The EMSs have also
2007; Mezulis et al. 2006; Oliver et al. 2007; been supported in factor analytic studies of the YSQ
Whisman and Kwon 1992). (Hoffart et al. 2005; Lee et al. 1999; Welburn et al.
Young (Young 1990, 1999; Young et al. 2003) 2002) although a consistent higher-order factor struc-
argued that the unique circumstances an individual ture has not emerged (see Oei and Baranoff 2007).
experiences in childhood contribute to the develop- Hoffart et al. (2005) recently evaluated the higher-
ment of a distinctive set of core beliefs about self and order structure of the 15 EMSs on the short-form of the
others which he termed early maladaptive schemas YSQ using confirmatory factor analysis. In addition to
(EMS).1 EMSs are believed to develop as a function replicating an earlier factor analysis (see Lee et al.
of thwarted, unmet, or inadequately met needs during 1999), these researchers tested a number of solutions
early development (see Young et al. 2003) and including the five domains purported by Young et al.
become self-perpetuating and resistant to change. A (2003). Hoffart et al. (2005) found that a four factor
number of different EMSs have been identified, model yielded the most parsimonious solution. The
which the Young Schema Questionnaire (YSQ) schema domains represented in this solution were
(Young 1990) or its short form (Young and Brown Disconnection (comprised of Emotional Deprivation,
2003) are designed to assess. An individual who Emotional Inhibition, Mistrust/Abuse, Social Isolation
experiences emotional detachment, rejection or and Defectiveness), Impaired Autonomy (which con-
abuse, for example, is purported to develop core sisted of EMSs of Subjugation, Dependence, Failure,
beliefs in the domain of Disconnection and Rejection Vulnerability, Abandonment, Enmeshment and Insuf-
(e.g., beliefs of being unwanted, inferior or unlov- ficient Self-control), Impaired Limits (made up of
able). Impaired Autonomy is believed to stem from an Insufficient Self-control and Entitlement) and Exag-
early environment that fails to reinforce a child gerated Standards (composed of Self-Sacrifice and
appropriately or that undermines a child’s perceived Unrelenting Standards).
competence. Beliefs and behaviors subsumed under According to Young et al. (2003), individuals with
the domain of Impaired Limits (e.g., entitlement, low EMSs tend to also display maladaptive coping
frustration tolerance or the refusal to maintain strategies that may perpetuate their schemas. Young
sufficient self-control) are thought to develop in the et al. (2003) discuss three strategies for coping with
context of parental over-permissiveness and lack of the threat of an activated schema. These coping
discipline. Other-directedness (e.g., beliefs pertaining strategies (i.e., overcompensation, avoidance, and
to the excessive need for affection and approval) is surrender) correspond to basic responses to threat
believed to derive from families of origin in which (i.e., fight, flight, or freeze, respectively). To our
children gain acceptance only by suppressing their knowledge, no previous research has examined the
own needs and pleasing their parents. Finally, relationship between compensatory or coping styles
and EMS.
1
Schemas have been defined in a number of ways, but most Humor styles—conceived as the ways people use
definitions incorporate the idea that they consist of both humor as a method of coping and a form of
structure (i.e., an organizational component) and content communication—may mediate the relationship
(Ingram et al. 1998). As such, our view is that ‘‘early
between EMSs and subsequent expressions of psy-
maladaptive schemas’’ (and other accessible beliefs) are really
proxies of the self-schema as they do not provide information chopathology (e.g., depressive symptomatology).
about the organization of this cognitive content. Many authors have noted that humor, because it

123
Cogn Ther Res (2009) 33:585–596 587

inherently involves incongruity and multiple inter- Martin et al. (2003) developed the Humor Styles
pretations, provides a way for individuals to shift Questionnaire (HSQ) to assess these four uses of
perspective on a stressful situation, reappraising it humor. A considerable amount of research using this
from a new and less threatening point of view, and measure has confirmed that the four styles are
thereby gaining a sense of mastery (e.g., Dixon 1980; differentially related to emotional and psychosocial
O’Connell 1976). Freud (1928) viewed healthy forms well-being in predicted ways (see Martin 2007, for
of humor as a mature defense mechanism, a view that review). In particular, self-enhancing humor corre-
has been enunciated more recently by Vaillant lates positively with such emotional well-being
(2000). Despite much research, it is unclear whether variables as self-esteem, optimism, positive moods,
humor moderates stress (for reviews, see Lefcourt and cheerfulness, and negatively with depression,
2001; Martin 2007). More recently, Martin et al. anxiety, rumination, perceived stress, and neuroti-
(2003) have argued that relatively weak and incon- cism. Affiliative humor, while somewhat less
sistent findings in previous humor research may be strongly related to emotional well-being, is particu-
due to a failure on the part of researchers to larly associated with positive relationship variables,
distinguish between potentially beneficial and detri- including intimacy, relationship satisfaction, social
mental styles of humor. Consistent with proposals of support, interpersonal competence, secure attach-
several earlier theorists (e.g., Allport 1961; Maslow ment, and extraversion, and negatively related to
1954), they suggested that humor may be used to loneliness and interpersonal anxiety.
cope in ways that are maladaptive as well as adaptive. On the other hand, self-defeating humor is posi-
Based on a review of the literature, Martin et al. tively correlated with anxiety, depression, psychiatric
(2003) identified four such styles of humor, two of symptoms, anxious attachment, and neuroticism, and
which were hypothesized to be beneficial for well- negatively associated with self-esteem and optimism.
being (affiliative and self-enhancing) and two poten- Finally, aggressive humor, while less strongly associ-
tially detrimental (self-defeating and aggressive). ated with emotional well-being, correlates negatively
Affiliative humor is characterized by making sponta- with relationship variables such as relationship satis-
neous humorous comments and telling jokes and faction, interpersonal competence, agreeableness, and
humorous anecdotes in order to amuse others, conscientiousness, and positively with hostility and
facilitate relationships, and minimize interpersonal neuroticism. Overall, then, greater use of self-enhanc-
tension. Self-enhancing humor refers to the use of ing humor and lower use of self-defeating humor seem
humor to regulate emotions and cope with stress by to be important for emotional well-being, whereas
maintaining a humorous and cheerful outlook on life. greater use of affiliative humor and lower use of
Aggressive humor involves the use of humor for the aggressive humor are predictive of more satisfactory
purpose of demeaning or manipulating others, as in interpersonal relationships.
sarcasm, teasing, or ridicule. Finally, self-defeating Individuals with early maladaptive schemas may
humor involves excessively self-disparaging humor, be less likely to engage in adaptive forms of humor
attempts to amuse others by doing or saying funny and more likely to use maladaptive types of humor in
things at one’s own expense as a means of ingrati- coping with schema-related stress and interacting
ating oneself or gaining approval, allowing oneself to with others. For example, individuals with core
be the ‘‘butt’’ of others’ humor, and using humor as a schemas involving themes of social disconnection
way of avoiding dealing constructively with one’s (e.g., emotional deprivation, mistrust/abuse, social
problems.2 isolation/alienation) may be less likely to develop the
playful, witty, and convivial interpersonal style
associated with affiliative humor, and more likely to
2
It is important to note that self-defeating humor is associated engage in the cynical, self-disparaging, ingratiating,
with low self-esteem and is conceptually distinct from the and avoidant forms of humor associated with the self-
concept of ‘‘self-deprecating humor’’, which involves not defeating humor style. In turn, these uses of humor
taking oneself too seriously and being able to make light of
may contribute to less satisfying interpersonal rela-
one’s faults and mistakes in a self-accepting manner. This is a
more healthy form of humor and is a component of affiliative tionships and greater dysphoria. Similarly, those with
humor. schemas containing themes of impaired autonomy

123
588 Cogn Ther Res (2009) 33:585–596

(e.g., dependence/incompetence, enmeshment, vul- between EMSs and dysphoria would be mediated by
nerability to harm) may be less likely to develop the Self-enhancing and Self-defeating humor (both of
self-enhancing style of humor which involves the which involve the use of humor in coping, but in
ability to find amusement and laugh at potentially adaptive and maladaptive ways, respectively).
threatening situations and thereby gain perspective
and distance oneself from sources of threat. In turn,
this reduced tendency to engage in humor as an Method
adaptive coping style may contribute to poorer
emotion regulation and greater emotional distress. Participants
Additionally, individuals with schemas associated
with impaired limits (entitlement, insufficient self- The sample was comprised of 305 first-year under-
control) may be more likely to engage in aggressive graduate students (77% female) recruited from the
humor styles involving the use of put-downs, sar- Psychology Department research participation pool at
casm, teasing, ridicule or disparaging humor, which the University of Western Ontario (UWO). The
in turn may contribute to less satisfactory relation- average age of participants was 18.95 (SD = 3.65).
ships with others. Seventy-percent of the sample identified themselves
This line of thinking would suggest that aggressive as Caucasian, 17% Asian, 2% African-Canadian, and
and self-defeating humor styles may be viewed as 1% of First Nations decent. Eleven percent of the
maladaptive ways of coping with activated schemas, sample listed ‘‘other’’ ethnic backgrounds. The
which may serve to perpetuate and maintain the majority of the students (97%) were single, 2%
schema-related negative beliefs. Thus, aggressive married and 1% divorced or separated.
humor may be used as an overcompensation strategy
to enhance oneself at the expense of one’s relation- Measures
ships with others. Similarly, self-defeating humor
may been viewed as a coping strategy that is Young Schema Questionnaire-Short Form (YSQ-SF;
consistent with Young et al.’s (2003) notion of Young and Brown 2003). The YSQ-SF is a self-report
surrender. measure designed to assess 15 different core beliefs and
The purpose of the present study was (1) to maladaptive schemas (emotional deprivation, abandon-
determine whether EMSs are associated with partic- ment, mistrust/abuse, social alienation, defectiveness,
ular humor styles, and (2) to determine whether incompetence, dependency, vulnerability to harm,
humor styles mediate the relationship between EMSs enmeshment, subjugation of needs, self-sacrifice, emo-
and dysphoria.3 It was predicted that, in general, tional inhibition, unrelenting standards, entitlement, and
higher EMS scores would be associated with lower insufficient self-control). The scale consists of 75 items
scores on Affiliative and Self-enhancing humor and each of which is rated from one (completely untrue of
higher Self-defeating humor. In addition, EMSs me) to six (describes me perfectly). The long form of this
involving Impaired Limits (i.e., Entitlement/Grandi- measure has received most of the empirical attention
osity and Insufficient Self-Control/Self-Discipline) and appears to be a valid measure of the verbal products
were expected to correlate positively with Aggressive (i.e., an indirect assessment) of core early maladaptive
humor. Finally, it was predicted that the association schemas (e.g., Riso et al. 2006). The psychometric
properties of the short form of this instrument also
appear to be on par with those of the full (205-item)
3
scale, demonstrating similar levels of reliability, valid-
We are not able to conclude that our findings speak to
ity, and clinical utility (Hoffart et al. 2005; Welburn
depression or its vulnerability because we assessed depressive
severity in an unselected sample of undergraduate students et al. 2002). The internal reliability (Cronbach’s alpha)
using the BDI-II. Nonclinical samples may simply capture the of the YSQ-SF in this study was .96. The average
experiences associated with general negative affect (e.g., internal consistency of the subscales was also excellent
neuroticism, fatigue, dysphoria, transient distress) rather than
(M = .86; SD = .04; range = .78–.93). In order to
depression per se. As such, we use the term ‘‘dysphoria’’ or
‘‘self-reported depressive severity’’ to describe this construct minimize Type I error in this study, the analyses focused
(see Kendall et al. 1987; Haaga and Solomon 1993). on the four schema domains (Disconnection, Impaired

123
Cogn Ther Res (2009) 33:585–596 589

Autonomy, Impaired Limits, and Exaggerated Stan- Procedure

dards) supported in a recent confirmatory factor analysis
(see Hoffart et al. 2005). Total scores for each of these Participants completed these measures in group
domains were computed by summing scores on the testing sessions. After providing informed consent,
particular YSQ subscales loading on the corresponding participants were administered the YSQ-SF, the
factor (Disconnection: emotional deprivation, mistrust/ HSQ, and the BDI-II in random order. They also
abuse, social isolation, defectiveness/shame, and emo- completed other measures which are not the focus of
tional inhibition; Impaired Autonomy: abandonment, this investigation. Participants received course credit
failure, dependence/incompetence, vulnerability, and for their involvement in the study.
enmeshment, subjugation; Impaired Limits: entitlement
and insufficient self-control; Exaggerated Standards:
self-sacrifice and unrelenting standards). All YSQ Statistical Analyses
domains intercorrelated between .25 and .48, except
for Disconnection and Impaired Autonomy, which Mediation analyses were conducted to test the
correlated .73. hypothesis that humor styles mediate the relationship
Humor Styles Questionnaire (HSQ; Martin et al. between the four YSQ domains and dysphoria. The
2003). The HSQ is a 32-item self-report measure that first step in these analyses was to examine simple
assesses four different styles of adaptive and mal- correlations between the predictor variables (i.e.,
adaptive humor (Affiliative, Self-Enhancing, YSQ domains), mediator variables (HSQ subscales),
Aggressive, and Self-Defeating). Respondents indi- and criterion variable (BDI-II). As noted by Baron
cate the extent to which they agree with each and Kenny (1986), a prerequisite for mediation is that
statement on a seven-point (1 = totally disagree; there must be significant correlations between pre-
7 = totally agree) Likert-type scale. Higher scores on dictor and criterion variables, between predictor and
this measure indicate that a particular style of humor mediator variables, and between mediator and crite-
is more descriptive of the respondent. Previous rion variables.
research has demonstrated good reliability for each Following these correlational analyses, multiple
of the subscales (Martin et al. 2003). Validity of the mediation analyses were conducted separately for
measures has been demonstrated by significant cor- each of the four YSQ domains, with a given YSQ
relations between each of the subscales and peer domain as the predictor variable and depression
ratings on the corresponding dimensions, as well as scores (BDI-II) as the criterion variable in each
theoretically predicted associations with other mea- analysis. Only those humor styles having significant
sures of humor, moods, self-esteem, relationship simple correlations with both the YSQ domain and
satisfaction, extraversion, agreeableness, hostility, BDI-II scores were examined as potential mediators
and other variables (Martin 2007; Martin et al. in a given analysis. To test for potential mediating
2003). The internal consistency (coefficient alpha) effects of several humor styles simultaneously, we
for the Affiliative, Self-Enhancing, Aggressive and used the recently developed bootstrap sampling
Self-Defeating subscales in the present sample were procedure described by Preacher and Hayes (2008).
.85, .87, .73, and .82, respectively. This procedure allows for the simultaneous exami-
Beck Depression Inventory-II (BDI-II). The BDI-II nation and statistical testing of the direct effect of the
(Beck et al. 1996) is well-established index of the predictor variable on the criterion variable, as well as
severity of unipolar depressive symptomatology. This the indirect (i.e., mediating) effect through the
21-item instrument is presented in a multiple-choice pathway of each of the mediator variables, control-
format (0–3), with summary scores ranging from 0 to ling for all the other mediators in the model. This is
63. Numerous studies have supported the reliability conceptually similar to multiple regression, in which
and validity of the BDI-II in both clinical (Beck et al. the effect of each of several predictors can be tested
1996) and undergraduate (Dozois et al. 1998) sam- while controlling for the effects of all the other
ples (see Dozois and Covin 2004, for review). The predictors in a model.
internal consistency of the BDI-II in this study was This bootstrap procedure uses sampling with
excellent (Cronbach’s alpha = .92). replacement to draw a large number of samples

123
590 Cogn Ther Res (2009) 33:585–596

(1,000 in the present study) from the data set and Table 1 Average scores on early maladaptive schema
calculate path coefficients for each sample. Then, domains, humor styles and dysphoria
using the estimates based on these 1,000 bootstrap Variable Mean (standard deviation)
samples, the mean direct and indirect effects and their
confidence intervals (CIs) are calculated. These CIs Disconnection 1.97 (.81)
are used to determine if each effect is statistically Impaired autonomy 1.87 (.70)
significant. For each effect, the 95% CI is examined Impaired limits 2.59 (.84)
and, if this range does not include zero, then this Exaggerated standards 3.46 (.88)
effect is statistically significant at P \ .05. In a Affiliative humor 45.99 (7.27)
similar fashion, CIs can also be set at 99%, resulting Self-enhancing humor 35.91 (9.47)
in significance levels of P \ .01 if zero does not fall Aggressive humor 29.22 (7.72)
within the calculated CI range. Preacher and Hayes Self-defeating humor 27.52 (8.84)
(2008) articulate several advantages of using this Beck depression inventory-II 11.46 (8.88)
bootstrap-driven CI statistical test for direct and Note: Mean scores for the YSQ-SF domains reflect the average
indirect effects, compared to product-of-coefficient across the number of items in each domain
approaches such as the Sobel test. One important
advantage of this approach is that it does not impose YSQ domains, the four scales of the HSQ, and the
the assumption of normally distributed variables. BDI-II scores are presented in Table 2. The Discon-
We conducted these analyses with SPSS 15.0 nection domain of the YSQ-SF was significantly
using the macro provided by Preacher and Hayes negatively correlated with both Affiliative and Self-
(2008) for carrying out the bootstrap procedure. In Enhancing humor and positively correlated with Self-
order to compare the results across analyses and Defeating humor. A similar pattern of correlations
variables, all variables used in these analyses were was found with the Impaired Autonomy domain. The
standardized (M = 0, SD = 1.0). Path coefficients Impaired Limits domain was negatively correlated
can therefore be interpreted in a manner similar to with Self-Enhancing humor and positively correlated
correlation coefficients. with both Aggressive and Self-Defeating humor.
Finally, the Exaggerated Standards domain was
weakly negatively correlated with Self-Enhancing
Results humor and positively correlated with Self-Defeating
humor.
For descriptive purposes, the means (SD) for the four All four of the YSQ-SF domains were significantly
Young Schema Questionnaire-Short Form (YSQ-SF) positively correlated with BDI-II scores, which is not
domains, the four scales of the Humor Styles inconsistent with the idea that these early maladap-
Questionnaire, and the BDI-II scores are presented tive schemas may be vulnerability factors for
in Table 1. Pearson correlations between the four depression (or at least dysphoria as assessed in this

Table 2 Correlations between Young Schema Questionnaire scale domains and scores on Humor Styles and Beck Depression
Inventory-II
Affiliative Self-enhancing Aggressive Self-defeating BDI-II
*** *** ***
Disconnection -.41 -.35 .09 .35 .61***
*** *** ***
Impaired autonomy -.29 -.37 .09 .42 .66***
Impaired limits -.09 -.24*** .27*** .27*** .39***
* ***
Exaggerated standards -.01 -.12 -.09 .26 .23***
*** *** ***
BDI-II -.22 -.48 .10 .30
Note: N = 305; BDI-II Beck depression inventory-II
* ***
P \ .05, P \ .001

123
Cogn Ther Res (2009) 33:585–596 591

Fig. 1 Mediating effects of

Affiliative
Humor Styles Questionnaire
Humor *
scales on relationship
between YSQ
Disconnection Factor and -.41 *** -.15 **
Self-
BDI-II Enhancing
-.35 *** -.36 ***
Humor **

YSQ .51 *** BDI-II

Disconnection Depression

Self-
Defeating .10 *
.35 ***
Humor (ns)

Note. *p < .05; **p < .01; ***p < .001

study). In addition, congruent with previous research, .37 to .24, indicating that the mediating effect of the
BDI-II scores were negatively correlated with Affi- two humor measures combined accounts for 13% of
liative and Self-Enhancing humor, and positively the variance in BDI-II scores, or about 35% (i.e., .13/
correlated with Self-Defeating humor. .37) of the total effect of Disconnection on BDI-II.4
Using the procedure described earlier, analyses In the mediation analysis for the YSQ Impaired
were then conducted to examine potential mediating Autonomy domain, Affiliative, Self-Enhancing, and
effects of the humor styles on the relationships Self-Defeating humor were again examined as
between each of the four YSQ domains and BDI-II potential mediators, based on the pattern of simple
scores. In the analysis using the YSQ Disconnection correlations found earlier. The results are shown in
domain as the predictor variable, the HSQ Affiliative, Fig. 2. In this analysis, a significant mediating effect
Self-Enhancing, and Self-Defeating humor styles was found only for Self-Enhancing humor (P \ .01).
were included as potential mediators, because only Higher scores on Impaired Autonomy predicted
these three humor scales were significantly correlated lower Self-Enhancing humor, which in turn predicted
with both Disconnection and BDI-II. The results of higher BDI-II scores. The direct effect of Impaired
this analysis are presented in Fig. 1. Significant Autonomy on dysphoria was also significant
mediating effects were found for Affiliative (c0 = .54, P \ .001), indicating that Self-Enhancing
(P \ .05) and Self-Enhancing humor (P \ .01), but humor only partially mediates this relationship.
the mediating effect for Self-Defeating humor was Regression analyses were again conducted to esti-
not significant. Thus, higher scores on Disconnection mate the strength of this effect. These results
were associated with lower Affiliative and Self-
Enhancing humor, which in turn predicted higher
4
BDI-II scores. In addition to the indirect effects of The reported data analyses were conducted on the combined
data from males and females. Further analyses were also
Disconnection on dysphoria through the two humor carried out on the data for females and males separately. The
styles, a direct effect was also found (c0 = .51, pattern of simple correlations among all the variables was very
P \ .001), indicating that the humor styles only similar for males and females as for the total sample. The
partially mediate this relationship. To obtain an mediation analyses for females revealed a pattern of significant
results that were nearly identical to those for the total sample.
estimate of the strength of this effect, multiple However, the mediational analyses for males revealed only one
regression analyses were conducted to compare the significant effect (Self-Enhancing humor mediating the rela-
variance in BDI-II scores accounted for by Discon- tionship between Disconnection and BDI-II). This was likely
nection when entered as the sole predictor, and after due to reduced power resulting from a much smaller sample
size for males than for females. However, further research with
first entering Affiliative and Self-Enhancing humor a larger sample of males is needed to explore the possibility of
scores. These analyses showed a reduction in R2 from a different pattern of mediating effects for males.

123
592 Cogn Ther Res (2009) 33:585–596

Fig. 2 Mediating effects of

Affiliative
Humor Styles Questionnaire
Humor (ns)
scales on relationship
between YSQ Impaired
Autonomy Factor and BDI- -.29 *** -.07
Self-
II
Enhancing
-.37 *** Humor ** -.31 ***

YSQ .54 *** BDI-II

Impaired Depression
Autonomy

Self-
.42 *** Defeating
.05
Humor (ns)
**
Note. p < .01; ***p < .001

Fig. 3 Mediating effects of

Self-
Humor Styles Questionnaire
Enhancing
scales on relationship
-.24 *** Humor ** -.41 ***
between YSQ Impaired
Limits Factor and BDI-II

YSQ BDI-II
Impaired .23 ***
Depression
Limits

.28 *** Self-

Defeating .21 ***
Humor **

**
Note. p < .01; ***p < .001

indicated that the R2 for Impaired Autonomy in associated with greater dysphoria. A significant direct
predicting BDI-II scores was reduced from .43 to .27 effect of Impaired Limits on BDI-II scores was also
when Self-Enhancing humor was entered first, indi- found (c0 = .23, P \ .001), indicating a partial medi-
cating that the mediating effect of Self-Enhancing ating effect of humor styles. Regression analyses
humor accounts for about 16% of the variance in revealed that the R2 for Impaired Limits in predicting
BDI-II scores, or about 37% of the total effect of BDI-II scores was reduced from .15 to .05 after
Impaired Autonomy. entering these two humor scales, indicating that the
In the analysis for the Impaired Limits domain, we mediating effect of humor accounts for about 10% of
included Self-Enhancing and Self-Defeating humor as the variance in BDI-II scores, or about 66% of the total
potential mediators, based on the pattern of simple effect of Impaired Limits.
correlations obtained with these scales. The results are Finally, Fig. 4 shows the results of the analysis for
presented in Fig. 3. Here, significant mediating effects the YSQ Exaggerated Standards domain, in which we
were found for both Self-Enhancing and Self-Defeat- again entered Self-Enhancing and Self-Defeating
ing humor (both P’s \ .01). Higher scores on Impaired humor as potential mediators. Again, significant
Limits predicted lower Self-Enhancing humor, which mediating effects were found for both Self-Enhancing
in turn predicted higher dysphoria. On the other hand, (P \ .05) and Self-Defeating humor (P \ .01).
higher Impaired Limits scores were associated with Higher scores on Exaggerated Standards predicted
higher Self-Defeating humor, which in turn was lower Self-Enhancing humor and higher Self-

123
Cogn Ther Res (2009) 33:585–596 593

Fig. 4 Mediating effects of

Self-
Humor Styles Questionnaire
Enhancing
scales on relationship
-.12 * Humor * -.45 ***
between YSQ Exaggerated
Standards Factor and BDI-II

YSQ BDI-II
Exaggerated .12 * Depression
Standards

.25 *** Self- .24 ***

Defeating
Humor **

Note. *p < .05; **p < .01; ***p < .001

Defeating humor, which in turn predicted higher to depression when congruent life events occur
BDI-II scores. The direct effect of Exaggerated (Abramson et al. 1997; Clark et al. 1992; Frewen
Standards on dysphoria was also significant and Dozois 2006; Kwon and Whisman 1998; Neitzel
(c0 = .12, P \ .05), although somewhat weaker than and Harris 1990; Robins and Block 1988).
in the preceding analyses, indicating a partial medi- Consistent with previous research (e.g., Chen and
ating effect of humor styles. Regression analyses Martin 2007; Frewen et al. 2008; Martin et al. 2003),
revealed a reduction in R2 from .06 to .01, indicating self-reported depressive symptoms were negatively
that the mediating effect of humor accounts for about correlated with the use of Self-Enhancing humor and
5% of the variance in BDI-II scores, or about 83% of (less strongly) with Affiliative humor, and positively
the total effect of Exaggerated Standards. related to Self-Defeating humor. The relationship
between dysphoria and Aggressive humor was not
statistically significant, a finding that is also congru-
Discussion ent with other research (see Martin 2007). The
relationship between aggressive humor and emotional
One objective of this study was to examine the well-being has not been established, although this
relationships among EMSs, adaptive/maladaptive form of humor has only been studied in the context of
styles of humor and dysphoria. Another purpose internalizing pathology. The lack of relationship
was to evaluate whether humor styles may mediate between an aggressive style of humor and dysphoria
the relationship between EMSs and dysphoria. Each is consistent with Beck’s (1967) early formulations of
of the EMS domains was positively and significantly cognitive theory in which he contended that depres-
related to self-reported depressive severity. Not sion is related more to themes of rejection, defeat,
surprisingly, the strongest relationships with BDI-II deprivation, and sensitivity to failure than a result of
scores were observed for Disconnection and Impaired anger turned inward. It is possible that a relationship
Autonomy. Disconnection has to do with core beliefs between aggressive humor and externalizing prob-
that one’s needs for love, nurturance, empathy, lems would emerge.
acceptance, and respect from others will not be met. An intriguing pattern of correlations was found
Impaired Autonomy involves beliefs that one lacks between EMSs and humor styles. Affiliative humor
the ability to separate and function independently was negatively related to both Disconnection and
from others. J. Beck (1995) contended that depress- Impaired Autonomy. As mentioned earlier, this
ogenic core beliefs typically fall into themes of humor style appears to be used to facilitate relation-
unlovability (‘‘I am unworthy’’; ‘‘I am undesirable’’) ships and minimize interpersonal tension. Both of
or helplessness (e.g., ‘‘I am inadequate’’; ‘‘I am these EMSs are characterized by distinctively inter-
incompetent’’). In addition, these variables corre- personal components. Young et al. (2003) defined
spond to two personality dimensions (sociotropy and Disconnection and Rejection as the expectation that
autonomy) which are believed to confer vulnerability an individual’s needs for nurturance, safety,

123
594 Cogn Ther Res (2009) 33:585–596

acceptance, and respect will not be reliably met. ability to see humor in potentially stressful circum-
Impaired autonomy, in contrast, refers to the antic- stances, perhaps due to a more serious and generally
ipation that one’s ability to function independently negative outlook on life. In turn, this coping deficit
will be impaired. Self-Enhancing humor (e.g., main- seems to result in poorer emotion regulation and
taining a humorous perspective even in the face of greater dysphoria. In addition, Self-Defeating humor
stress or adversity) was negatively correlated with all mediated the effect of Impaired Limits and Exagger-
four of the EMS domains and may serve a general ated Standards on BDI-II scores. Finally, Affiliative
protective function against dysphoria and negative humor mediated the relationship between Disconnec-
beliefs. tion and dysphoria.
Aggressive humor correlated significantly only A significant limitation, particularly in these
with the EMS dimension of Impaired Limits. This mediator analyses, is that our measures were admin-
finding makes intuitive sense when one considers the istered cross-sectionally. Recent conceptual work in
nature of this EMS domain. Impaired Limits (e.g., the area of moderator and mediator analysis demon-
entitlement, insufficient self-control) pertains to a strates that a priori assumptions about causal
deficiency in one’s direction, goal-setting, responsi- direction can lead to flawed conclusions that may in
bility to others and internal limits. The use of fact reverse the actual cause and effect relationships,
aggressive humor may represent a method for fail to identify critical third variables, or result in
elevating one’s social status in a situation by cutting modifying the wrong factor until the actual cause and
others down. We view this form of humor as a effect relationship is identified in subsequent studies
potential compensatory strategy for perceived defi- (Kraemer et al. 2008). This work also highlights the
ciencies in self-worth. Finally, Self-Defeating humor fact that temporal precedence is critical for identify-
was significantly and positively related to all EMS ing mediators and that this is important to consider in
domains. This excessively self-disparaging use of designing any study that aims to demonstrate causal
humor is likely used as a coping strategy to attempt to mediation. Given the cross-sectional nature of our
gain approval by amusing others. Rather than min- study, we cannot rule out that, for example, humor
imizing negative experiences and beliefs, however, styles precede the formation of a particular EMS, that
this form of humor may have an effect opposite to the the experience of prior depression produces EMS, or
intended one. In sum, early maladaptive schemas are even that a third variable is responsible for deter-
associated with less use of adaptive humor styles and mining both EMS and humor style. It is also possible,
greater use of maladaptive humor styles. for instance, that depressive symptomatology may
Another aspect of this investigation was to test affect the use of and/or reporting of humor styles.
whether styles of humor serve as potential mediators Although our correlational data cannot be used to
of the relationship between EMSs and dysphoria. determine the direction of causality, these results are
Young et al. (2003) had argued that particular coping consistent with the hypothesis that the link between
styles may serve to exacerbate or maintain core these EMS and depressed mood may be (at least
beliefs. We conceptualize the four humor styles as partially) mediated by reduced use of adaptive forms
coping and communication strategies that may be in of humor in coping with stress and regulating
part influenced by maladaptive schemas and may in emotions, and increased use of excessively self-
turn play a role in increasing or decreasing depres- disparaging and avoidant humor. Another limitation
sion. Mediator analyses revealed that the relationship in this study was that the sample was comprised
between each of the four YSQ domains and dysphoria mainly of Caucasian females and this limits the
was mediated by various combinations of the humor generalizability of these findings to other populations.
styles, with moderate to strong combined mediating For example, the level of endorsement of core belief
effect sizes (ranging from 35 to 83% of the total domains was generally low. To what extent these
effects). In particular, Self-Enhancing humor showed relations between EMS and humor styles would hold
a significant mediating effect on the relationships in individuals who endorse these core beliefs more
between BDI-II scores and all four YSQ domains. strongly (e.g., a clinical sample with a mood or other
Thus, individuals with these maladaptive schemas disorder requiring treatment) remains an open ques-
appear to be less likely to develop this adaptive tion. The focus of this investigation was also on

123
Cogn Ther Res (2009) 33:585–596 595

depressive symptoms and not other variables that Conceptual, strategic, and statistical considerations.
may have demonstrated different relationships with Journal of Personality and Social Psychology, 51, 1173–
1182. doi:10.1037/0022-3514.51.6.1173.
core beliefs and humor (e.g., externalizing pathology, Beck, A. T. (1967). Depression: Causes and treatment. Phil-
impaired social interactions). Future research should adelphia: University of Pennsylvania Press.
examine the relationships among EMS, humor styles Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979).
and depression using a prospective design in a sample Cognitive therapy of depression. New York: Guilford.
Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Beck
of patients requiring treatment. depression inventory manual (2nd ed.). San Antonio:
These findings suggest the possibility that humor, Psychological Corporation.
as a coping style, could be a useful target for Beck, J. S. (1995). Cognitive therapy: Basics and beyond. New
intervention in treatment for depression. Cognitive York: Guilford.
Bruce, A. E., Cole, D. A., Dallaire, D. H., Jacquez, F. M.,
therapy approaches to depression already emphasize Pineda, A. Q., & LaGrange, B. (2006). Relations of par-
the relationship between beliefs and use of compen- enting and negative life events to cognitive diatheses for
satory strategies—as well as how this relationship depression in children. Journal of Abnormal Child Psy-
between beliefs and behaviors leads to further chology, 34, 321–333. doi:10.1007/s10802-006-9019-x.
Chen, G., & Martin, R. A. (2007). A comparison of humor
distress. In the present analysis, humor style could styles, coping humor, and mental health between Chinese
be said to be a compensation for specific kinds of and Canadian university students. Humor: International
EMS’s. Therefore, as with other compensatory strat- Journal of Humor Research, 20, 215–234. doi:
egies, it might be useful for therapist and patient to 10.1515/HUMOR.2007.011.
Clark, D. A., Beck, A. T., & Alford, B. A. (1999). Scientific
collaboratively evaluate the function of humor, and foundations of cognitive theory and therapy of depression.
thereby establish whether the patient’s usual Philadelphia: Wiley.
approach to humor leads to intended and/or unin- Clark, D. A., Beck, A. T., & Brown, G. K. (1992). Sociotropy,
tended consequences. If patients were to recognize autonomy, and life event perceptions in dysphoric and
nondysphoric individuals. Cognitive Therapy and
that their use of humor may have some drawbacks Research, 16, 635–652. doi:10.1007/BF01175404.
that they had not previously considered, more adap- Dixon, N. F. (1980). Humor: A cognitive alternative to stress?
tive humor styles could be worked on or implemented In I. G. Sarason & C. D. Spielberger (Eds.), Stress and
on an experimental basis to discover what impact this anxiety (Vol. 7, pp. 281–289). Washington: Hemisphere.
Dozois, D. J. A., & Beck, A. T. (2008). Cognitive schemas,
has on mood. Certainly the results of this study beliefs and assumptions. In K. S. Dobson & D. J. A.
suggest that these processes ought to be better Dozois (Eds.), Risk factors in depression (pp. 121–143).
understood in a clinical population. Oxford: Elsevier/Academic Press.
The findings from this study demonstrate that Dozois, D. J. A., & Covin, R. (2004). The Beck depression
inventory-II (BDI-II), Beck hopelessness scale (BHS), and
putative cognitive vulnerabilities to depression may Beck scale for suicide ideation (BSS). In M. Hersen
be mediated by adaptive and maladaptive uses of (Series Ed.), D. L. Segal & M. Hilsenroth (Vol Eds.),
humor in coping. Additional research is needed to Comprehensive handbook of psychological assessment:
examine the specific mechanisms by which EMSs Volume 2. Personality assessment and psychopathology
(pp. 50–69). Hoboken, NJ: Wiley.
and humor styles confer risk to depression. Dozois, D. J. A., Dobson, K. S., & Ahnberg, J. L. (1998). A
psychometric evaluation of the Beck depression inven-
Acknowledgements This research was supported by a grant tory-II. Psychological Assessment, 10, 83–89. doi:
from the Social Sciences and Humanities Research Council of 10.1037/1040-3590.10.2.83.
Canada. This funding is gratefully acknowledged. Freud, S. (1928). Humour. The International Journal of Psy-
cho-Analysis, 9, 1–6.
Frewen, P. A., Brinker, J. K., Martin, R. A., & Dozois, D.
J. A. (2008). Humor styles and personality-vulnerabil-
References ity to depression. Humor: International Journal of
Humor Research, 21, 179–195. doi:10.1515/HUMOR.
Abramson, L. Y., Alloy, L. B., & Hogan, M. E. (1997). Cog- 2008.009.
nitive/personality subtypes of depression: Theories in Frewen, P. A., & Dozois, D. J. A. (2006). Self-worth appraisal of
search of disorders. Cognitive Therapy and Research, 21, life events and Beck’s congruency model of depression
247–265. doi:10.1023/A:1021870315058. vulnerability. Journal of Cognitive Psychotherapy: An
Allport, G. W. (1961). Pattern and growth in personality. New International Quarterly, 20, 231–240. doi:10.1891/jcop.20.
York: Holt, Reinhart & Winston. 2.231.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator Gibb, B. E., Alloy, L. B., Walshaw, P. D., Comer, J. S., Shen, G. H.
variable distinction in social psychological research: C., & Villari, A. G. (2006). Predictors of attributional style

123
596 Cogn Ther Res (2009) 33:585–596

change in children. Journal of Abnormal Child Psychology, Neitzel, M. T., & Harris, M. J. (1990). Relationship of
34, 425–439. doi:10.1007/s10802-006-9022-2. dependency and achievement/autonomy to depression.
Glaser, B. A., Campell, L. F., Calhoun, G. B., Bates, J. M., & Clinical Psychology Review, 10, 279–297. doi:10.1016/
Petrocelli, J. V. (2002). The early maladaptive schema 0272-7358(90)90063-G.
questionniare-short form: A construct validity study. O’Connell, W. E. (1976). Freudian humor: The eupsychia of
Measurement & Evaluation in Counseling & Develop- everyday life. In A. J. Chapman & H. C. Foot (Eds.),
ment, 35, 2–13. Humor and laughter: Theory, research, and applications
Haaga, D. A. F., & Solomon, A. (1993). Impact of Kendall, (pp. 313–329). London: Wiley.
Hollon, Beck, Hammen, and Ingram (1987) on treatment Oei, T. P. S., & Baranoff, J. (2007). Young Schema Ques-
of the continuity issue in ‘depression’ research. Cognitive tionnaire: Review of psychometric and measurement
Therapy and Research, 17, 313–324. doi:10.1007/ issues. Australian Journal of Psychology, 59, 78–86.
BF01177657. Oliver, J. M., Murphy, S. L., Ferland, D. R., & Ross, M. J.
Hoffart, A., Sexton, H., Hedley, L. M., Wang, C. E., Holthe, (2007). Contributions of the cognitive style questionnaire
H., Haugum, J. A., et al. (2005). The structure of mal- and the dysfunctional attitudes scale to measuring cogni-
adaptive schemas: A confirmatory factor analysis and a tive vulnerability to depression. Cognitive Therapy and
psychometric evaluation of factor-derived scales. Cogni- Research, 31, 51–69. doi:10.1007/s10608-006-9067-0.
tive Therapy and Research, 29, 627–644. doi:10.1007/ Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and
s10608-005-9630-0. resampling strategies for assessing and comparing indirect
Ingram, R. E., Miranda, J., & Segal, Z. V. (1998). Cognitive effects in multiple mediator models. Behavior Research
vulnerability to depression. New York: Guilford. Methods, 40, 879–891. doi:10.3758/BRM.40.3.879.
Kendall, P. C., Hollon, S. D., Beck, A. T., Hammen, C. L., & Riso, L. P., Froman, S. E., Raouf, M., Gable, P., Maddux, R.
Ingram, R. E. (1987). Issues and recommendations E., Turini-Santorelli, N., et al. (2006). The long-term
regarding the use of the Beck depression inventory. stability of early maladaptive schemas. Cognitive Therapy
Cognitive Therapy and Research, 11, 289–299. doi: and Research, 30, 515–529. doi:10.1007/s10608-006-
10.1007/BF01186280. 9015-z.
Kraemer, H. C., Kiernan, M., Essex, M., & Kupfer, D. J. Robins, C., & Block, P. (1988). Personal vulnerability, life events,
(2008). How and why criteria defining moderators and and depressive symptoms: A test of a specific interactional
mediators differ between the Baron & Kenny and Mac- model. Journal of Personality and Social Psychology, 54,
Arthur approaches. Health Psychology, 27(2), S101– 847–852. doi:10.1037/0022-3514.54.5.847.
S108. doi:10.1037/0278-6133.27.2(Suppl.).S101. Schmidt, N. B., Joiner, T. E., Young, J. E., & Telch, M. J.
Kwon, P., & Whisman, M. A. (1998). Sociotropy and auton- (1995). The schema questionnaire: Investigation of psy-
omy as vulnerabilities to specific life events: Issues in life chometric properties and the hierarchical structure of a
event categorization. Cognitive Therapy and Research, measure of maladaptive schemas. Cognitive Therapy and
22, 353–362. doi:10.1023/A:1018757128916. Research, 19, 295–321. doi:10.1007/BF02230402.
Lee, C. W., Taylor, G., & Dunn, J. (1999). Factor structure of Vaillant, G. E. (2000). Adaptive mental mechanisms: Their
the schema questionnaire in a large clinical sample. role in a positive psychology. The American Psychologist,
Cognitive Therapy and Research, 23, 441–451. doi: 55, 89–98. doi:10.1037/0003-066X.55.1.89.
10.1023/A:1018712202933. Welburn, K., Coristine, M., Dagg, P., Pontefract, A., & Jordan,
Lefcourt, H. M. (2001). Humor: The psychology of living S. (2002). The schema questionnaire-short form: Factor
buoyantly. New York: Kluwer. analysis and relationship between schemas and symptoms.
Lumley, M. N., & Harkness, K. L. (2007). Specificity in the Cognitive Therapy and Research, 26, 519–530. doi:
relations among childhood adversity, early maladaptive 10.1023/A:1016231902020.
schemas, and symptom profiles in adolescent depression. Whisman, M. A., & Kwon, P. (1992). Parental representation,
Cognitive Therapy and Research, 31, 639–657. doi: cognitive distortions and mild depression. Cognitive Ther-
10.1007/s10608-006-9100-3. apy and Research, 16, 557–568. doi:10.1007/BF01175141.
Martin, R. A. (2007). The psychology of humor: An integrative Young, J. E. (1990). Cognitive therapy for personality disor-
approach. Burlington: Elsevier. ders: A schema-focused approach. Sarasota: Professional
Martin, R. A., Puhlik-Doris, P., Larsen, G., Gray, J., & Weir, K. Resource Press.
(2003). Individual differences in uses of humor and their Young, J. E. (1999). Cognitive therapy for personality disor-
relation to psychological well-being: Development of the ders: A schema-focused approach (3rd ed.). Sarasota:
humor styles questionnaire. Journal of Research in Per- Professional Resource Press.
sonality, 37, 48–75. doi:10.1016/S0092-6566(02)00534-2. Young, J.E., & Brown, G. (2003). Young schema questionnaire:
Maslow, A. H. (1954). Motivation and personality. New York: Short form. Available at http://www.schematherapy.
Harper. com/id54.htm.
Mezulis, A. H., Shibley Hyde, J., & Abramson, L. Y. (2006). Young, J. E., Klosko, J. S., & Weshaar, M. E. (2003). Schema
The developmental origins of cognitive vulnerability to therapy: A practitioner’s guide. New York: Guilford.
depression: Temperament, parenting, and negative life Young, J. E., Rygh, J. L., Weinberger, A. D., & Beck, A. T.
events in childhood as contributors to negative cognitive (2008). Cognitive therapy for depression. In D. H. Barlow
style. Developmental Psychology, 42, 1012–1025. doi: (Ed.), Clinical handbook of psychological disorders (4th
10.1037/0012-1649.42.6.1012. ed., pp. 250–305). New York: Guilford.

123