You are on page 1of 10

Differential Emotions Theory: Relative

Contribution of Emotion, Cognition,


and Behavior to the Prediction
of Depressive Symptomatology
in Non-referred Adolescents
Ä

Tracy C. Carey and Michael P. Carey


Medical College of Ohio
Ä

Mary L. Kelley
Louisiana State University

This study examined the relative contribution of emotions, cognition, and


behavior to depressive symptomatology with 80 non-referred adoles-
cents. Differential emotions theory emphasized emotions but pointed to
the importance of behavioral and cognitive variables. Subjects completed
self-rating scales, theoretically related to depression and differential emo-
tions theory, assessing depressive symptomatology, emotions, automatic
dysfunctional thoughts, and activity. Two hierarchical multiple regression
models entered the emotional variables first but varied the entry of behav-
ioral and cognitive variables. The findings emphasized the interplay between
emotion and behavior, and provided equivocal support for the role of
cognition. © 1997 John Wiley & Sons, Inc.

Differential emotions theory provides a structure for understanding the role of emotions in
depressive symptomatology (Izard, 1972). According to the theory, the 10 fundamental emo-
tions (interest, enjoyment, surprise, sadness, anger, disgust, contempt, fear, shyness, and guilt)

Address correspondence to Tracy C. Carey, Ph.D., Medical College of Ohio, Department of Pediatrics, 3000 Arlington
Avenue, P.O. Box 10008, Toledo, OH 43699.

JOURNAL OF CLINICAL PSYCHOLOGY, Vol. 53(1), 25–34 (1997)


© 1997 John Wiley & Sons, Inc. CCC 0021-9762/97/010025-10
26 Journal of Clinical Psychology, January 1997

each have motivational characteristics. Each emotion has its own neuromotor program and
emotional expression is activated by neurochemical changes. Both internal and external events
can trigger neurochemical change and thus emotional expression. As an emotion is experi-
enced, it may become associated with other emotions such that certain emotions tend to occur
together or may influence the expression of other emotions. For example, the concurrent expe-
riences of joy and sadness reduced the facial expression of sadness in boys, whereas the com-
bined experience of sadness and anger increased the facial expression of sadness in the children
(Blumberg & Izard, 1991). Life events can trigger multiple emotions (Blumberg & Izard,
1991), and once activated, a pattern of emotions can impact on cognition and behavior in a
reciprocal manner (Buechler & Izard, 1980; Izard & Malatesta, 1987).
Differential emotions theory states that a combination of emotions highlighted by sadness,
and components of hostility (anger, disgust, contempt) and shame (guilt, shyness) contribute to
depressive symptomatology (Izard, 1972). The findings of Blumberg and Izard (1985, 1986)
revealed that the emotions of sadness, self-directed hostility, anger, shame, interest, and enjoy-
ment contributed to the prediction of depressive symptomatology in non-referred 10- and 11-year
old children. Moreover, research with reliably DSM-III-R (American Psychiatric Association,
1987) diagnosed adolescent psychiatric patients indicated that the emotional configuration of
depressed adolescents differed from non-depressed youths on the emotions of enjoyment, sur-
prise, sadness, anger, shame, shyness, guilt, and self-directed hostility (Carey, Finch, & Carey,
1991). These studies provide support for the role of emotions in depressive symptomatology.
Izard and Schwartz (1986) further stated that emotions are not perceived as the cause of
depression, but one’s lack of skills to protect oneself from negative emotions may have an
impact on negative emotions. According to Izard and Schwartz (1986), “Continual vulnerabil-
ity to these emotions may in turn lead to detrimental cognitive styles (negative views of the
self, world, and future) as described by Beck (1967) and Seligman and Peterson (1986). . .”
(p. 49). Several cognitive components including attributional styles and negative automatic
thoughts have been related to depressive symptomatology in the literature (Hops, Lewinsohn,
Andrews, & Roberts, 1990; Kazdin, 1990a). Research on differential emotions theory with
children has been limited largely to the integration of differential emotions theory and refor-
mulated learned helplessness theory (Abramson, Seligman, & Teasdale, 1978).
Differential emotions theory also suggests that life events impact on emotions. Blumberg
and Izard (1986) indicate that emotions can be related to Lewinsohn’s theory of depression.
Lewinsohn (1974) proposed that a reduced frequency of positive reinforcement may affect
emotional responses and this in turn can impact on behavior. Although the behavioral compo-
nent has not been examined in the differential emotions literature, research on activity and
depression in adolescents has demonstrated that individuals reporting a higher frequency of
unpleasant activities report higher levels of depressive symptoms (Carey, Kelley, Buss, & Scott,
1986; Cole, Kelley, & Carey, 1988). In contrast to the adult literature, the frequency of involve-
ment in pleasant activities for adolescents was not correlated with level of depressive symp-
toms (Carey et al., 1986; Kanner, Feldman, Weinberger & Ford, 1987).
Izard and Schwartz (1986) state that maladaptive behavior may result from problems in the
links between emotion, cognition, and behavior. The research to date on differential emotions
theory has just begun to examine the role of emotion and cognition in the prediction of depres-
sive symptoms from a reformulated learned helplessness perspective. Blumberg and Izard (1985),
in one of the few studies examining the linkage between emotion and cognition, found that a
child’s attributional style accounted for additional variance beyond that of emotions in boys’
depression scores. The literature has not yet addressed other cognitive components, the role of
behavior, or the combined role of these two modalities with emotion. Furthermore no study as
yet has examined how well emotions in combination with cognitive processes and behavior
predict the occurrence of depressive symptomatology.
Adolescent Depression 27

Thus, the present investigation was designed to build on the available literature based on
differential emotions theory. This study focused on adolescents because research on differential
emotions theory as applied to adolescents is limited. Differential emotions theory stresses the
importance of emotions while also acknowledging the reciprocal relations between emotions,
cognitions, and behavior in the onset and maintenance of depressive symptomatology. This
study examined whether behavioral and cognitive variables would add a significant amount of
variance in the prediction of self-reported depression scores above and beyond that which was
explained by emotional variables. Differential emotions theory emphasizes the importance of
emotions but does not explicitly define the relative importance of behavioral and cognitive
variables to the prediction of self-reported depressive symptoms. Therefore, two differential
emotions theory models were examined. The models focused primarily on emotions but varied
the order of entry of behavioral and cognitive variables to assess whether these variables could
account for additional variance in depression scores. The following hypotheses were explored
in this study:

1. Since differential emotions theory suggests the presence of reciprocal relations between
emotions, behavior, and cognition (Izard & Schwartz, 1986), the assessment of behav-
ioral and cognitive variables was expected to contribute variance above the variance
explained by emotions in the prediction of depressive symptomatology.
2. The emotions of sadness, self-directed hostility, anger, fear, interest, enjoyment, shame,
shyness, and guilt were expected to be significant in the prediction of depression. These
factors were selected based on their proposed relation to depression cited in differential
emotions theory and findings from previous studies on depressive symptoms with chil-
dren and adolescents (Blumberg & Izard, 1985, 1986; Carey, Finch, & Carey, 1991;
Izard, 1972).
3. The frequency of engagement in pleasant and unpleasant activities was proposed to
contribute a significant amount of variance in addition to emotions in the prediction of
depressive symptomatology. These expectations were based on the literature on activity
(Carey et al., 1986; Cole et al., 1988), and Lewinsohn’s (1974) hypothesis of the reduc-
tion in the rate of response contingent reinforcement and Blumberg and Izard’s (1986)
hypothesis regarding pleasant events.
4. According to Beck’s Cognitive Theory of Depression (1967), negative automatic thoughts
were expected to contribute to the prediction of depressive symptoms in addition to
emotions since the literature to date has provided support for the independent contri-
butions of negative automatic thoughts (Kazdin, 1990a; Stark, 1990) to depressive symp-
tomatology.
5. Finally, it was hypothesized that the combination of emotions, dysfunctional automatic
thoughts and behavior would correctly classify a substantial number of adolescents
whose score on a measure of depressive symptomatology fell above and below the
established cut-score.

METHOD
Subjects
Eighty non-referred adolescents from grades 7 through 12 in public schools participated in the
study. The sample contained 40% males (n 5 32) and 60% females (n 5 48) with a mean age
of 14.45 years (SD 5 1.58, range 12–17 years). Fifty-five percent of the sample was Caucasian
(n 5 44) and 45% (n 5 36) was Black or other minorities (n 5 36). According to Hollings-
head’s (1965) Two Factor Index of Social Status, the subjects primarily came from lower
28 Journal of Clinical Psychology, January 1997

(SES 5 1 or 2; n 5 35), middle (Mdn 5 3.0, n 5 23) and upper SES backgrounds (SES 5 4 or
5; n 5 22). Chi square analyses were conducted to determine whether the non-referred sample
differed in terms of over or under-representation of demographic characteristics and revealed
no significant differences in age and gender. However, the SES of the sample differed in regard
to race (x 2 5 18.42, p , .001). Caucasian subjects generally came from higher socioeconomic
backgrounds than minorities.

Instrumentation
The assessment measures employed in this study were selected based on their theoretical rela-
tion to depression. Since research has indicated that children/adolescents are better able to
report on their own subjective personal experiences, and the depressive syndrome includes
subjective states, self-rating scales were selected for this study (Kendall, Cantwell, & Kazdin,
1989; Morretti, Fine, Haley, & Marriage, 1985). Specifically, 9 of the 12 emotions from the
Differential Emotions Scale, shown to be related to depressive symptomatology in the litera-
ture, were used to assess the emotional domain (Blumberg & Izard, 1985; Izard & Schwartz,
1986). The cognitive domain was evaluated using a measure of dysfunctional automatic thoughts.
Overt behavior was assessed using a measure of reported engagement in pleasant and unpleas-
ant events. Adolescents completed the self-report measures in a group setting with the mea-
sures being presented in a counterbalanced order to control for response sets.

Reynolds Adolescent Depression Scale (RADS). The RADS is a 30-item self-report scale used
to assess depressive symptoms with adolescents in grades 7 through 12 (Reynolds, 1987). The
RADS has demonstrated high internal consistency and test–retest reliability (Reynolds, 1987).
The concurrent and predictive validity of the RADS has been evaluated in a number of inves-
tigations. Reynolds (1986) reported a high correlation between the Hamilton Depression Rat-
ing Scale (HDRS; Hamilton, 1960) and the RADS (r 5 .83). Convergence between the RADS
and HDRS using an established cut score of 77 on the RADS has yielded a hit rate of 89%
(Reynolds, 1987). Previous studies also have shown that the RADS is sensitive to changes in
treatment outcome of depressed adolescents (Reynolds & Coats, 1986).

Differential Emotions Scale (DES). The DES-IV is a 37-item self-report measure based on
Izard’s (1972) Differential Emotions Theory. The DES-IV assesses the following 12 fundamen-
tal emotions: interest, joy, surprise, sadness, anger, disgust, contempt, fear, shame, shyness,
guilt, and self-directed hostility. Nine DES factors (sadness, hostility, anger, fear, interest, enjoy-
ment, shame, guilt, and shyness) were chosen a priori for this study. The available studies on
the DES with children and adolescents have demonstrated adequate reliability and validity
(Blumberg & Izard, 1985, 1986; Carey, Finch, & Carey, 1991).

Automatic Thoughts Questionnaire (ATQ). The ATQ is a self-report instrument developed to


assess the frequency of automatic negative self-statements related to depression (Hollon &
Kendall, 1980). This 30-item instrument has been used with both referred and non-referred
children and adolescents. The ATQ has demonstrated good internal consistency, test–retest
reliability and validity (Carey, Lubin, & Brewer, 1992; Kazdin, 1990b). Validation of the ATQ
with several measures of depression has demonstrated moderate correlations in both referred
and non-referred samples (Garber, Weiss, & Shanley, 1990; Kazdin, 1990b).

Adolescent Activities Checklist (AAC). The AAC consists of items generated by adolescents
assessing 50 pleasant and 50 unpleasant activities for use with adolescents (Carey et al., 1986).
Two subscale scores are obtained by summing the frequency of pleasant (PES) and unpleasant
(UPES) activities. Standardization of the AAC has been established with 563 adolescents (Cole
et al., 1988). The AAC has demonstrated high internal consistency and good test–retest relia-
Adolescent Depression 29

bility (Carey et al., 1986; Cole et al., 1988). Studies investigating the validity of the AAC have
provided support for its content, construct, and concurrent validity.

RESULTS
Preliminary Analyses
Previous studies with the DES had indicated that several of the subscales differed by gender
(Blumberg & Izard, 1991). Therefore, preliminary analyses were conducted to determine whether
the RADS, ATQ, or AAC (PES, UPES) differed in terms of gender, race, age, or SES using
one-way ANOVAs. No significant differences were obtained on the RADS, ATQ, or AAC for
the four demographic variables investigated. Since gender had been found to differ on several
of the DES subscales, gender was entered into the multiple regression equations first to statis-
tically control for its effect.

Primary Analyses
Two multiple regression analyses were conducted to examine the relative contribution of vari-
ables assessing emotion, overt behavior, and dysfunctional automatic thoughts in the prediction
of self-reported depressive symptomatology. The analyses involved predicting the RADS total
score as a function of the emotional variables (DES factors: Sadness, Self-Directed Hostility,
Anger, Fear, Interest, Enjoyment, Shame, Guilt, and Shyness), behavioral variables (PES, UPES)
and cognitive variable (ATQ). Given the prospect of high multi-collinearity, each independent
variable was regressed on all other independent variables (Berry & Feldman, 1985). Self-
Directed Hostility and Sadness demonstrated a high degree of multi-collinearity as evidenced
by an R 2 greater than .70. Thus, the Self-Directed Hostility and Sadness subscales were elim-
inated from the two primary multiple regression analyses.
In the two regression analyses, the variables assessing emotion, overt behavior, and cog-
nition were entered by blocks into the regression equation. After gender, the emotion subscales
were entered since differential emotions theory emphasizes the importance of emotions. The
theory does not explicitly define the relative importance of behavioral and cognitive variables
to the prediction of self-reported depressive symptoms. Therefore, two hierarchical multiple
regression analyses were conducted between blocks of variables to determine whether cogni-
tive and/or behavioral variables added to the prediction of depressive symptoms.

Emotion, Behavior, and Cognition Model


In the first model, gender was entered first, then the block of emotional variables was entered
in the second step. The behavioral and cognitive blocks were entered in the third and fourth
steps. Table 1 summarizes the hierarchical multiple regression analysis for this model. Gender
did not contribute a significant amount of variance. The block of emotional variables, as expected,
added significantly to the prediction of self-reported depressive symptoms. The emotional
variables accounted for 67% ( p , .0005) of the variance. The behavioral variables explained
an additional 4% of the variance ( p , .01). The cognitive variable accounted for approximately
1% of the variance but did not significantly ( p , .06) increase the amount of variance accounted
for by the regression equation.

Emotion, Cognition, and Behavior Model


The second model again entered gender first followed by the block of emotional variables. The
order of the cognitive and behavioral blocks was switched with cognitive variable being entered
30 Journal of Clinical Psychology, January 1997

Table 1 (Model 1). Hierarchical Multiple Regression: Emotion, Behavior, and Cognition

Variable Beta Weight Partial R F R 2 Change F Change

Gender .10 .15 1.55 .01 .82


Emotion .67 20.81*
Shame .41 .33 8.24*
Interest −.01 .02 .04
Anger .09 .02 .88
Enjoyment −.66 −.40 12.82*
Guilt .29 .26 4.73**
Fear −.02 −.09 .56
Shyness .03 .03 .08
Behavior .04 4.56***
PES .04 −.16 1.82
UPES .29 .30 6.71***
Cognition .01 3.47
ATQ .18 .22 3.47

*p < .005. **p < .05. ***p < .01.

in the third step and overt behavioral variables being entered in the fourth step. The second
model revealed that all three blocks of variables added significantly to the prediction of RADS
scores (see Table 2). The block of emotional variables as expected accounted for the same
amount of variance as reported in the first model. In contrast to the findings in the first model,
a cognitive variable significantly added to the prediction of self-reported depressive symptoms
(2%, p , .02). As in the first model, the block of behavioral variables again added to the
prediction of depressive symptoms (3%, p , .04).
In both models the variables that were most influential were Enjoyment ( p , .0005),
Shame ( p , .006), frequency of involvement in Unpleasant Events ( p , .01), and Guilt ( p ,
.03). Dysfunctional automatic thoughts was less influential only approaching statistical signif-

Table 2 (Model 2). Hierarchical Multiple Regression: Emotion, Cognition, and Behavior

Variable Beta Weight Partial R F R 2 Change F Change

Gender .10 .15 1.55 .01 .82


Emotion .67 20.81*
Shame .41 .33 8.24*
Interest −.01 .02 .04
Anger .09 .02 .88
Enjoyment −.66 −.40 12.82*
Guilt .29 .26 4.73**
Fear −.02 −.09 .56
Shyness .03 .03 .08
Cognition .02 5.66**
ATQ .23 .22 3.47
Behavior .03 3.43**
PES .03 −.16 1.81
UPES .26 .30 6.71***

*p < .005. **p < .05. ***p < .01.


Adolescent Depression 31

icance ( p , .06). Thus, several emotions, one component of overt behavior and to a lesser
extent dysfunctional automatic thoughts contributed to the prediction of depressive symptoms.

Discriminant Function Analysis


A discriminant function analysis was conducted to determine how accurately the seven emo-
tions (anger, enjoyment, shame, fear, guilt, interest, and shyness) along with the two overt
behavioral variables (PES and UPES), and the cognitive variable (ATQ) could correctly clas-
sify individuals with high or low depressive symptoms. The high depressive symptomatology
group (n 5 18; 22.8%) consisted of youths with RADS scores greater than or equal to the
established cut score of 77. The low depressive symptomatology group (n 5 61; 77.2%) was
composed of adolescents with RADS scores below 77. Prior probabilities were set proportional
to the number of youths within each group so as not to take advantage of chance. The findings
indicated that the predictor variables were able to significantly distinguish the two groups and
accounted for 52.7% of the variance, x 2 (10) 5 53.88, p , .0001. As seen in Table 3, the
discriminant function analysis correctly classified 92.4% of the subjects with 27.8% of the
non-depressed group being misclassified.

DISCUSSION
This study was conducted to further evaluate Izard’s (1972) differential emotions theory with
adolescents, an under-studied population with respect to Izard’s theory. This study evaluated
whether depressive symptomatology in a non-referred adolescent sample could be predicted by
integrating emotional, cognitive, and behavioral variables. Partial support was found for the
first hypothesis relating to the contribution of the block of emotional variables, behavioral, and
a cognitive variable to the prediction of depressive symptoms. The findings emphasize the
importance of the interplay between emotion and behavior, although the impact of cognition
remains in question since its degree of importance was affected by its order of entry into the
hierarchical multiple regression analyses.
Hypotheses concerning the contribution of variables within a specific domain (i.e., emo-
tional, behavioral, cognitive) to the prediction of depressive symptomatology also was exam-
ined. Partial support was found for the second hypothesis relating to the emotional domain. The
findings of the present study suggested that the emotions of joy, shame, guilt, anger, fear,
interest, and shyness did contribute to the prediction of depressive symptomatology. The emo-
tions of joy, shame, and guilt accounted for the largest amount of variance in depression scores
as evidenced by their respective beta weights and partial correlation coefficients. Unfortu-

Table 3. Discriminant Analysis Classification Results

Predicted
Group Membership

Actual Group N Depressed Non-Depressed

Depressed 61 60 1
(RADS $ 77) (.77) (98.4%) (1.6%)
Non-Depressed 18 5 13
(RADS # 77) (.23) (27.8%) (72.2%)
Overall Hit Rate = 92.4%
32 Journal of Clinical Psychology, January 1997

nately, the emotions of sadness and self-directed hostility were unable to be evaluated due to
the high degree of multi-collinearity with the other variables.
The research on differential emotions theory with children (Blumberg & Izard, 1985) has
emphasized the role of anger in depression; whereas research with severely emotionally dis-
turbed adolescents (Carey et al., 1991) has indicated that shyness, anger, low levels of enjoy-
ment, and shame were prominent in diagnosed depressed youth. Interestingly, the importance
of the emotions of joy and shame from this study’s non-referred adolescent sample paralleled
the findings of Carey et al. (1991) with severely emotionally disturbed adolescents. Yet the
emotions of anger and shyness were not found to be as influential with our sample of non-
referred adolescents. Thus, the findings of this study in combination with previous studies may
suggest that the relative importance of specific emotions to depressive symptomatology may be
affected by gender, age, and severity of depressive symptomatology.
The third hypothesis that reported engagement in pleasant and unpleasant events would
contribute to the prediction of depressive symptoms was supported by both hierarchical mul-
tiple regression analyses. Thus, despite of the order of entry, this study found that behavior was
an important component in the prediction of depressive symptoms. However, the findings from
this study did not conform to Blumberg and Izard (1986) and Lewinsohn’s (1974) hypotheses
concerning the relation between depression and frequency of engagement in various types of
activities or events.
Blumberg and Izard (1986) hypothesized that based on Lewinsohn’s (1974) theory that a
reduced frequency of positive reinforcement may affect emotional responses and may influ-
ence behavior. The findings of low levels of enjoyment may reflect a loss of positive reinforce-
ment or a greater frequency of unpleasant events. In contrast to Lewinsohn (1974) and Blumberg
and Izard’s (1986) hypothesis, the frequency of involvement in pleasant activities for non-
referred adolescents was not as important of a predictor of depressive symptoms as the fre-
quency of involvement in unpleasant events.
This finding was not surprising since studies with adolescents investigating the association
between pleasant events and depression have failed to find a significant relationship (Carey
et al., 1986; Garber & Hilsman, 1992). In contrast, the research on the relation between activity
and depression in adolescents has demonstrated that individuals who report a higher frequency
of engagement in unpleasant activities report higher levels of depressive symptoms (Carey
et al., 1986; Cole et al., 1988). This suggests a possible phenomenological difference between
the relation between activity and depressive symptoms in adolescents as compared to adults.
Concerning the fourth hypothesis, which proposed that dysfunctional automatic thoughts
would contribute to the prediction of depressive symptoms, the findings were mixed, contin-
gent on the order of entry of the cognitive variable in the hierarchical multiple regression
analyses. It is possible that the mixed findings concerning the importance of the dysfunctional
automatic thoughts may have been an artifact of the selection of the cognitive variable or its
shared variance with the overt behavioral variables. Alternatively, dysfunctional automatic
thoughts may provide a vulnerability to depression, but not predict subsequent depression. This
hypothesis is supported by the work of Hammen, Adrian, and Hiroto (1988) who in a longitu-
dinal study of children at risk for depression, reported that depression was not predicted by a
depressogenic attribution style. Furthermore, Blumberg and Izard (1985) found that attribu-
tional style was not predictive of girls’ depression scores.
Finally, the fifth hypothesis proposed that the combination of emotional, behavioral, and
cognitive variables would correctly classify a substantial number of adolescents whose RADS
score fell above and below the established cut-score of 77. This hypothesis was supported as
evidenced by correct classification of 92.4% of adolescents. This study used a non-referred
sample of adolescents. The findings lend support for utilizing emotional, behavioral, and cog-
nitive variables when attempting to assess for depressive symptomatology, given the high hit
Adolescent Depression 33

rate (i.e., 89%) between the RADS and the Hamilton Depression Rating Scale (Reynolds,
1987). Future studies are needed to determine whether similar findings can be obtained when
comparing clinically depressed and non-depressed adolescents.
Despite the favorable findings from this study, there were several limitations which deserve
mention. First, the findings are based solely on information obtained from adolescents. Thus,
future research needs to make use of alternate methods (i.e., structured diagnostic interviews)
and multiple informants. Second, our findings also are not necessarily generalizable to prepu-
bertal children; therefore, future research examining the presence of developmental differences
needs to be conducted.

REFERENCES
Abramson, L .Y., Seligman, M.E.P., & Teasdale, J.P. (1978). Learned helplessness in humans: Cri-
tique and reformulation. Journal of Abnormal Psychology, 46, 478–488.
American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders
(Third edition-revised). Washington, DC: American Psychiatric Association.
Beck, A.T. (1967). Depression. New York: Harper & Row.
Berry, W.D., & Feldman, S. (1985). Multiple regression in practice. Beverly Hills, CA: Sage Publi-
cations, Inc.
Blumberg, S.H., & Izard, C.E. (1985). Affective and cognitive characteristics of depression in 10- and
11-year old children. Journal of Personality and Social Psychology, 49, 194–202.
Blumberg, S.H., & Izard, C.E. (1986). Discriminating patterns of emotions in 10- and 11-year-old
children’s anxiety and depression. Journal of Personality and Social Psychology, 51, 852–857.
Blumberg, S.H., & Izard, C.E. (1991). Patterns of emotion experiences as predictors of facial expres-
sions of emotion. Merrill–Palmer Quarterly, 37, 183–195.
Buechler, S., & Izard, C.E. (1980). Anxiety in childhood and adolescence. In I.L . Kutash, L .B.
Schlesinger, & Associates (Eds.), Handbook on stress and anxiety (pp. 385–297). San Francisco:
Jossey-Bass.
Carey, T.C., Finch, A.J., & Carey, M.P. (1991). Relation between differential emotions and depression
in emotionally disturbed children and adolescents. Journal of Consulting and Clinical Psychology,
59, 594–597.
Carey, M.P., Kelley, M.L ., Buss, R.R., & Scott, W.O. (1986). Relationship of activity to depression
in adolescents: Development of the Adolescent Activities Checklist. Journal of Consulting and Clin-
ical Psychology, 54, 320–322.
Carey, M.P., Lubin, B., & Brewer, D.H. (1992). Measuring dysphoric mood in pre-adolescents and
adolescents: The Youth Depression Adjective Checklist (Y-DACL). Journal of Clinical Child Psy-
chology, 21, 332–338.
Cole, T.L ., Kelley, M.L ., & Carey, M.P. (1988). Adolescent Activities Checklist: Reliability, factorial
validity and standardization data. Journal of Abnormal Child Psychology, 16, 475–484.
Garber, J., & Hilsman, R. (1992). Cognitions, stress, and depression in children and adolescents.
Psychiatric Clinics of North America, 1, 129–167.
Garber, J., Weiss, B., & Shanley, N. (1990). Cognitions, depressive symptoms and development in
adolescents. Manuscript submitted for publication.
Hamilton, M. (1960). A rating scale for depression. Journal of Neurological and Neurosurgical Psychi-
atry, 35, 837–844.
Hammen, C., Adrian, C., & Hiroto, D. (1988). A longitudinal test of the attributional vulnerability
model in children at risk for depression. British Journal of Clinical Psychology, 27, 37–46.
Hollingshead, A.B. (1965). Two factor index of social status. Unpublished manuscript, Yale University,
New Haven, CT.
34 Journal of Clinical Psychology, January 1997

Hollon, S.D., & Kendall, P.C. (1980). Cognitive self-statements in depression: Development of an
Automatic Thoughts Questionnaire. Cognitive Therapy and Research, 4, 383–395.
Hops, H., Lewinsohn, P.M., Andrews, J.A., & Roberts, R.E. (1990). Psychosocial correlates of
depressive symptomatology among high school students. Journal of Clinical Child Psychology, 19,
211–220.
Izard, C.E. (1972). Patterns of emotions: A new analysis of anxiety and depression. New York: Aca-
demic Press.
Izard, C.E., & Malatesta, C. (1987). Perspectives on emotional development: 1. Differential Emo-
tions of early emotional development. In J. Osofsky (Ed.), Handbook of infant development (Vol. 2,
pp. 13–21). New York: Wiley.
Izard, C.E., & Schwartz, G.M. (1986). Patterns of emotion in depression. In M. Rutter, C. Izard, & P.
Read (Eds.), Depression in young people: Developmental perspectives (pp. 33–70). New York: Guil-
ford Press.
Kanner, A.D., Feldman, S.S., Weinberger, D.A., & Ford, M.E. (1987). Uplifts, hassles, and adap-
tational outcomes in early adolescents. Journal of Early Adolescence, 7, 371–394.
Kazdin, A.E. (1990a). Evaluation of the Automatic Thoughts Questionnaire: Negative cognitive pro-
cesses and depression among children. Psychological Assessment: A Journal of Consulting and Clin-
ical Psychology, 2, 73–79.
Kazdin, A.E. (1990b). Childhood depression. Journal of Child Psychology and Psychiatry, 31, 121–160.
Kendall, P.C., Cantwell, D.A., & Kazdin, A.E. (1989). Depression in children and adolescents:
Assessment issues and recommendations. Cognitive Therapy and Research, 13, 109–146.
Lewinsohn, P. (1974). Clinical and theoretical aspects of depression. In K. Calhoun, H. Adams, & K.
Mitchell (Eds.), Innovative treatment methods in psychopathology (pp. 63–120). New York: John
Wiley & Sons.
Morretti, M., Fine, S., Haley, G., & Marriage, K. (1985). Childhood and adolescent depression:
Child-report versus parent-report information. Journal of the American Academy of Child and Ado-
lescent Psychiatry, 24, 298–302.
Reynolds, W.M. (1986). A model for the screening and identification of children and adolescents in
school settings. Professional School Psychology, 1, 117–129.
Reynolds, W.M. (1987). Assessment of depression in adolescents: Manual for the Reynolds Adolescent
Depression Scale (RADS). Psychological Assessment Resources.
Reynolds, W.M., & Coats, K.I. (1986). A comparison of cognitive-behavioral therapy and relaxation
training for the treatment of depression in adolescents. Journal of Consulting and Clinical Psychol-
ogy, 54, 653–660.
Seligman, M.E., & Peterson, C. (1986). A learned helplessness perspective on childhood depression:
Theory and research. In M. Rutter, C. Izard, & P. Read (Eds.), Depression in young people: Clinical
and developmental perspectives (pp. 223–249). New York: Guilford Press.
Stark, K. (1990). Childhood depression: School-based intervention. New York: Guilford Press.

You might also like