School Psychology Review

ISSN: (Print) 2372-966x (Online) Journal homepage: https://www.tandfonline.com/loi/uspr20

Looking Beyond Psychopathology: The Dual-Factor

Model of Mental Health in Youth

Shannon M. Suldo & Emily J. Shaffer

To cite this article: Shannon M. Suldo & Emily J. Shaffer (2008) Looking Beyond
Psychopathology: The Dual-Factor Model of Mental Health in Youth, School Psychology Review,
37:1, 52-68

To link to this article: https://doi.org/10.1080/02796015.2008.12087908

Published online: 27 Dec 2019.

Submit your article to this journal

Article views: 19

View related articles

Citing articles: 29 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=uspr20
School Psychology Review,
2008, Volume 37, No. 1, pp. 52– 68

FEATURED ARTICLE

Looking Beyond Psychopathology: The Dual-Factor

Model of Mental Health in Youth

Shannon M. Suldo and Emily J. Shaffer

University of South Florida

Abstract. In a dual-factor model of mental health (cf. Greenspoon & Saklofske,

2001), assessments of positive indicators of wellness (i.e., subjective well-be-
ing—SWB) are coupled with traditional negative indicators of illness (i.e., psy-
chopathology) to comprehensively measure mental health. The current study
examined the existence and utility of a dual-factor model in early adolescence.
The SWB, psychopathology, academic functioning, social adjustment, and phys-
ical health of a general sample of 349 middle school students was assessed via
self-report scales, school records, and teacher reports regarding students’ exter-
nalizing psychopathology. The existence of a dual-factor model was supported
through the identification of four mental health groups: 57% of the sample had
complete mental health, 13% was vulnerable, 13% was symptomatic but content,
and 17% was troubled. The means of the four groups differed significantly in
terms of academic outcomes, physical health, and social functioning. Results
support the importance of high SWB to optimal functioning during adolescence,
as students with complete mental health (i.e., high SWB, low psychopathology)
had better reading skills, school attendance, academic self-perceptions, academic-
related goals, social support from classmates and parents, self-perceived physical
health, and fewer social problems than their vulnerable peers also without clinical
levels of mental illness but with low SWB. Among students with clinical levels
of psychopathology, students with high SWB (symptomatic but content youth)
perceived better social functioning and physical health.

This research was supported in part by the Pediatric Clinical Research Center of All Children’s Hospital
and the University of South Florida, and the Maternal and Child Health Bureau, R60 MC 00003– 01,
Department of Health and Human Resources, Health Resources and Services Administration.
The authors acknowledge the assistance of the following members of their university research team: Jessica
Michalowski, Jennie Farmer, Kristen Riley, Allison Friedrich, Julie Dixon, Nandelyne Metellus, and
Heather Merrifield.
Correspondence regarding this article should be addressed to Shannon Suldo, Department of Psychological
and Social Foundations, University of South Florida, 4202 East Fowler Avenue, EDU 162, Tampa, FL,
33620; E-mail: suldo@coedu.usf.edu
Copyright 2008 by the National Association of School Psychologists, ISSN 0279-6015

52
 Dual-Factor Model of Mental Health

Mental health is increasingly viewed as fect, and negative affect (Diener, 2000). Life
a complete state of being, consisting not satisfaction includes both global and domain-
merely of the absence of illness or disorder but specific (e.g., family, school) cognitive ap-
also the presence of positive factors such as praisals of one’s happiness. Affect entails
life satisfaction, self-acceptance, and social fairly stable emotions and mood states. An
contribution (Ryff & Singer, 1998; Keyes, individual reporting high SWB would make a
2003). Although there is budding attention to positive judgment of the overall quality of his
positive development, the majority of research or her life, and experience more frequent pos-
conducted to date on adolescents’ mental itive affect (e.g., joy, elation, delight) relative
health has focused exclusively on psycholog- to negative affect (sadness, guilt, anger; Die-
ical disorders (Evans et al., 2005). The mental ner, Lucas, & Oishi, 2002).
health concerns of vulnerable youth who may Cowen (1994) has proposed life satis-
be at risk for developing future problems— faction as a key indicator of psychological
those who are not detected on screening mea- wellness, asserting “wellness is something
sures of psychopathology, but, when given the more than/other than the absence of disease,
opportunity, report diminished life satisfaction that is, it is defined by the ‘extent of presence’
or happiness (cf. Greenspoon & Saklofske, of positive marker characteristics” (p. 154).
2001) or who are “languishing” in life (cf. Park (2004) echoed the sentiment that “posi-
Keyes, 2006)— have been overlooked to a tive indicators such as life satisfaction should
large degree. Similarly, not all youth with be included in any assessment battery to cap-
clinical levels of psychopathology experience ture comprehensively what is meant by the
poor quality of life (Bastiaansen, Koot, & Fer- psychological well-being of youth” (p. 27).
dinand, 2005). The integration of positive and Notably, there is little empirical support for
negative indicators of well-being into mental the additive information that may be gleaned
health assessment yields a more comprehen- from including SWB in assessments of mental
sive picture of functioning (Huebner, Gilman, health. Moreover, the assertion that “youth
& Suldo, 2007; Snyder et al., 2003). reporting low levels of pathological symptoms
could still experience diminished psychologi-
Traditional Negative Indicator of Mental cal well-being” (Park, 2004, p. 26) is intuitive
Health—Psychopathology but has little data behind it. This lack of em-
pirical investigation and support may contrib-
Psychopathology refers to both internal-
ute to the continued emphasis on psychopa-
izing psychological disorders (e.g., depres-
thology, a focus that has persevered even in
sion, anxiety) and externalizing disorders
the face of the positive psychology and posi-
(e.g., conduct disorder, oppositional defiant
tive youth development movements (Seligman
disorder). Traditionally, mental health diagno-
& Csikszentmihalyi, 2000).
sis is defined simply by the presence or ab-
Compared to the substantial body of lit-
sence of disorders or associated negative out-
erature on indicators of psychopathology, far
comes. If criteria are not met for a certain
fewer studies have focused on positive indica-
disorder, the patient is termed subclinical and
tors of well-being in youth. The traditional
no subsequent intervention would likely fol-
model of wellness places happiness on a con-
low. By using traditional assessments, mental
tinuum with psychopathology, each being at
health becomes an inferred by-product in the
opposite ends. Diener et al. (2002) assert that
absence of mental illness.
the absence of disease is not an adequate cri-
Positive Indicator of Mental Health— terion to describe a person as mentally
Subjective Well-Being (SWB) healthy, particularly not as possessing high or
even average levels of SWB or optimal adjust-
SWB is the scientific term for happiness. ment. In contrast to a one-dimensional view of
SWB is comprised of three related but sepa- mental health, a more encompassing view
rate constructs: life satisfaction, positive af- would define distress and well-being as sepa-
53
School Psychology Review, 2008, Volume 37, No. 1

Table 1
Groups Yielded from a Dual-Factor Model of Mental Health

Level of SWB
Level of
Psychopathology Low Average to High

Low II. Vulnerable I. Complete mental health

SWB composite ⱕ 30th percentile SWB composite ⬎ 30th percentile
and Internalizing T score ⬍ 60 and and Internalizing T score ⬍ 60
Externalizing T score ⬍ 60 and Externalizing T score ⬍ 60

High IV. Troubled III. Symptomatic but content

SWB composite ⱕ 30th percentile SWB composite ⬎ 30th percentile
and Internalizing T score ⱖ 60 or and Internalizing T score ⱖ 60 or
Externalizing T score ⱖ 60 Externalizing T score ⱖ 60

Note. SWB ⫽ subjective well-being.

rate yet interrelated constructs. Such a two- identified as at least minimally psychologi-
factor framework of psychological health is cally disordered. The latter group of children,
supported by a factor-analytic study with ad- who experienced diminished life satisfaction
olescents in which a well-being factor loaded even though they were not clinically dis-
on by life satisfaction and positive affect was rupted, had relatively low self-concepts, low
distinguishable from a distress factor loaded perceived academic competence, and poor in-
on by anxiety and negative affect (Wilkinson terpersonal relationships. Traditionally, such
& Walford, 1998). children would not be targeted for intervention
because of the absence of psychopathology.
Dual-Factor Model of Mental Health This study illustrates the utility of a dual-
in Youth factor model of mental health, that is, psycho-
To date, only one published study has pathology and SWB as separable constructs
explored the integration of psychopathology that make unique contributions to predictions
and SWB in children with respect to implica- of children’s functioning. Findings suggest
tions for outcomes. Specifically, Greenspoon that the construct of SWB is as worthy a target
and Saklofske (2001) isolated groups of ele- of study as the disorders on which psychology
mentary school children who demonstrated has historically focused.
that psychopathology could co-occur with Aims of Current Study
high life satisfaction and, conversely, that the
absence of psychopathology and low life sat- The overarching purpose of the project
isfaction could exist simultaneously. The was to further explore the utility of the dual-
former group, children who would likely be factor model of mental health in youth (cf.
identified on screeners of pathological behav- Greenspoon & Saklofske, 2001) by (a) extend-
ior, possessed a distinct set of traits (e.g., high ing the generalizability of existing research to
sociability and good interpersonal relations) U.S. adolescents (findings from the aforemen-
relative to other children characterized by high tioned study were based on a sample of chil-
psychopathology and low life satisfaction. dren in Grades 3– 6 in western Canada) and (b)
Thus, knowledge of children’s life satisfaction examining a wider variety of important out-
was helpful in predicting their functioning and come measures on which students with vari-
adjustment, even among children already ous mental health profiles may differ. Table 1
54
 Dual-Factor Model of Mental Health

summarizes the subgroups yielded from a the first examination of comprehensive SWB
dual-factor model. Complete mental health re- (i.e., life satisfaction and affect) relative to
fers to youth with optimal wellness. Vulnera- adolescents’ health. Previous research with
ble youth are typically excluded from study youth has linked psychopathology and physi-
and services because of their lack of psycho- cal health problems (Pine, Cohen, Brook, &
pathology, but may need assistance given their Coplan, 1997; Rohde, Lewinsohn, & Seeley,
low perceived quality of life. Symptomatic but 1994). Studies of adults suggest a relationship
content youth, although identified as deviant, between SWB and physical health (Michalos
may not suffer to the same extent because of & Zumbo, 2002). Preliminary research with
their positive cognitive judgments of life. youth links reduced life satisfaction to worse
Troubled youth have mental health problems perceived overall health and activity limita-
and appraise the quality of their lives as poor. tions because of poor health (Zullig, Valois,
The specific classification criteria included in Huebner, & Drane, 2005). Specific indicators
Table 1 are described in the Results section. of physical health measured in the current
The current study aimed to answer two study include perceptions of overall health,
primary questions. First, what proportion of bodily pain, and the extent to which health
children in a general population of middle problems limit children’s activities with their
school students have complete mental health, families.
are vulnerable, symptomatic but content, or
troubled? Second, do the subgroups of adoles- Method
cents yielded from a dual-factor model of
mental health differ in terms of educational Participants
functioning, social relationships, and physical Participants consisted of 349 students
health? The identification of adjustment dif- from a middle school (grades 6 – 8) in the
ferences between groups of students with low southeastern United States and 44 teachers
psychopathology and between groups with from the same school familiar with the student
high psychopathology will provide mental participants. Student participants ranged in
health professionals with an empirical ratio- age from 10 to 16 years (M ⫽ 12.96;
nale for attending to mental health beyond the SD ⫽ 0.97), with 60% female and 40% male.
presence or absence of illness—a goal consis- Fifty-five percent of student participants were
tent with a commitment to increase wellness in Caucasian, 14% African American, 12% His-
all children. Outcome variables in the current panic or Latino, 10% multiracial, and 8%
study were selected because of their salience other ethnicities. Twenty-six percent of stu-
to healthy development during youth. dents participated in the free or reduced-cost
Educational outcomes selected for study school lunch program (used as an indicator of
include academic achievement, in-school con- low socioeconomic status [SES]). Subgroups
duct, and attitudes toward learning and school. slightly underrepresented in the sample, as
A growing body of research identifies impor- compared to the school-wide demographics,
tant links between academic functioning and included African American students (19% of
life satisfaction (for a review, see Suldo, Ri- the student population) and students from low
ley, & Shaffer, 2006). Previous research with SES backgrounds (32% of student popula-
youth has also found that high life satisfaction tion). Multiracial students were overrepre-
is associated with social interactions with sented in the sample (5% of student
peers (Martin & Huebner, 2007) and higher population).
perceptions of social support from multiple
sources (Suldo & Huebner, 2006). Therefore, Measures and Indicators
the current study examined students’ social
problems and perceived social support from Students’ Life Satisfaction Scale
teachers, peers, and parents. The current study (SLSS; Huebner, 1991). The SLSS is a
is unique in its focus on physical health, and is 7-item measure of global life satisfaction de-
55
School Psychology Review, 2008, Volume 37, No. 1

signed for use with students in Grades 3–12. degree to which a feeling or behavior is true
Respondents indicate on a Likert scale, rang- for themselves currently (i.e., in the past 6
ing from 1 (strongly disagree) to 6 (strongly months) using a Likert scale that ranges
agree), the degree to which they endorse gen- from 0 (not true) to 2 (very true). A composite
eral statements about their life (e.g., “life is of the three subscales assessing (a) with-
going well,” “I wish I had a different life”). An drawn– depressed, (b) somatic complaints, and
overall life satisfaction score is obtained by (c) anxious– depressed was used as the indica-
reverse-scoring negatively worded items, then tor of internalizing psychopathology. Reliabil-
summing the responses and dividing by 7. For ity and validity of the YSR is well established
a summary of the respectable psychometric (see Achenbach & Rescorla, 2001). In the
properties of the SLSS (e.g., high test–retest current study, coefficient alpha for the inter-
and internal consistency reliability, validity nalizing composite was .89.
established via significant relationships with
other measures of life satisfaction), see Gil- Teacher Report Form of the Child
man and Huebner (1997). In the current study, Behavior Checklist (TRF; Achenbach &
coefficient alpha was .89. Rescorla, 2001). The TRF is a 113-item
rating scale completed by teachers and other
Positive and Negative Affect Scale school personnel who are familiar with the
for Children (PANAS-C; Laurent et al., child’s behavior in the school setting. The
1999). The PANAS-C consists of a 12-item TRF has norms for youth ages 5–18 years, and
and a 15-item self-report scale that assess fre- taps the same dimensions of psychopathology
quency of positive affect and negative affect, as the YSR. Respondents rate students’
respectively. Respondents rate 27 words that present behavior (i.e., behavior occurring over
describe feelings and emotions on a Likert the past 2 months) using the same 0 –2 agree-
scale, ranging from 1 (very slightly or not at ment scale as in the YSR. A composite of the
all) to 5 (extremely), to indicate the extent to two subscales assessing (a) rule-breaking be-
which they have experienced each mood or havior and (b) aggressive behavior was used
feeling in the past few weeks. Positive affect as the indicator of externalizing psychopathol-
descriptors include interested, excited, and ogy. The reliability and validity of the TRF is
proud. Negative affect adjectives include jit- well established; the test manual includes a
tery, lonely, and sad. Ratings within each scale thorough summary of the high test–retest re-
are averaged to generate overall scores of pos- liability and construct validity (Achenbach &
itive and negative affect. The correlation be- Rescorla, 2001). In this study, coefficient al-
tween the positive and negative subscales is pha for the externalizing composite was .92.
low (Laurent et al., 1999). Internal consistency Social Problems subscale of the
is high for both scales, and construct validity YSR. The 11-item Social Problems subscale
has been supported through the magnitude of of the YSR assesses students’ perceptions of
relationships with measures of constructs with the degree to which they have negative inter-
which affect is theoretically related and differ- actions with peers or exhibit socially awkward
ent (Laurent et al. 1999; Seligson, Huebner, & behaviors. See Achenbach and Rescorla
Valois, 2005). In this study, coefficient alphas (2001) for a summary of the high test–retest
were .88 (positive affect) and .93 (negative reliability and empirical support for construct
affect). validity. In the current study, coefficient alpha
for the social problems scale was .70.
The Youth Self-Report Form of the
Child Behavior Checklist (YSR; Achen- Child and Adolescent Social Sup-
bach & Rescorla, 2001). The YSR is com- port Scale (CASSS; Malecki & Demaray,
prised of 112 items designed to measure eight 2002). The CASSS is a self-report measure
dimensions of psychopathology among ado- of children’s perceptions of social support
lescents ages 11–18. Participants indicate the from five sources: parents, teachers, class-
56
 Dual-Factor Model of Mental Health

mates, a close friend, and school administra- and dividing by the total number of classes
tors. Using a Likert response metric, ranging attempted. Participants’ year-to-date class
from 1 (never) to 6 (always), students rate how grades through the grading period in which
often they perceive receiving four types of self- and teacher-report data were collected
support (emotional, instrumental, appraisal, (mean GPA for the first, second, and third
and informational) offered by each source 9-week segments of the current school year)
(e.g., parent, teacher). Scores within each sub- were obtained from school records.
scale were averaged so that higher scores re-
Standardized assessment scores. The
flect higher perceived social support from each
Florida Comprehensive Assessment Test
source. Only three subscales (Parent, Teacher,
(FCAT; Florida Department of Education,
Classmate) were administered in the current
2005) is a statewide, norm-referenced test ad-
study. The CASSS has high test–retest reli-
ministered to all students in Florida in Grades
ability and strong correlations with other es-
3–11 for purposes of educational assessment
tablished measures of social support (Malecki,
and accountability. The FCAT includes crite-
Demaray, & Elliot, 2000). In the current study,
rion-referenced tests that measure student
coefficient alpha was .95 for each support
progress toward statewide benchmarks. Scores
scale (Teacher Support, Classmate Support,
are assigned on a five-level grading criteria
and Parent Support).
(1–5), with Levels 3–5 constituting a passing
score. Criterion-referenced FCAT math and
School Attitude Assessment Survey—
reading scores were analyzed in the current
Revised (SAAS-R; McCoach & Siegle, study. More information about the FCAT is
2003). Students’ beliefs relevant to educa- available at http://fcat.fldoe.org/.
tional functioning were measured via four scales
of the 35-item self-report SAAS-R: (1) Aca- Attendance. Attendance history was
demic Self-Perceptions (perceived academic operationalized as the total number of school
abilities), (2) Motivation and Self-Regulation days missed during the school year through
(ability to regulate daily behavior to attain aca- the grading period in which the study was
demic goals), (3) Valuing of School (importance conducted. Therefore, higher attendance
placed on academic achievement in order to scores reflect more time absent from school.
meet long-term goals), and (4) Attitude Toward
School (feelings of pride and belonging to one’s Child Health Questionnaire—Child
specific school). Respondents indicated agree- Form 87 (CHQ-CF87; Landgraf, Abetz, &
ment with each item using a Likert scale, rang- Ware, 1999). The CHQ-CF87 is an 87-item
ing from 1 (strongly disagree) to 7 (strongly self-report measure for use with children ages
agree). Each scale of the SAAS-R has adequate 10 –18, designed to assess 12 physical and
reliability (McCoach & Siegle, 2003). See Mc- psychosocial concepts. The following three
Coach and Siegle (2001, 2003) for a summary of scales assessing physical health were analyzed
criterion-related validity. Validity is also sup- in the current study: (1) General Health Per-
ported by significant correlations between scales ceptions (subjective assessment of overall
of the SAAS-R and indicators of similar con- health and frequency of illness), (2) Bodily
structs (Suldo, Shaffer, & Shaunessy, in press). Pain/Discomfort (intensity and frequency of
In the current study, coefficient alpha ranged physical pain and discomfort), and (3) Limi-
from .90 (Academic Self-perceptions) to .95 (At- tations in Family Activities (frequency of dis-
titudes Towards School). ruption in usual family activities because of
child’s health or behavior). Although two ad-
Grade-point average (GPA). GPA ditional scales (Physical Functioning and So-
was calculated by assigning numerical values cial/Physical Role) were originally intended to
to letter grades received for academic perfor- be examined, scores from these scales were
mance in each class (e.g., A ⫽ 4.0, B ⫽ 3.0) not retained for data analyses because of ex-
57
School Psychology Review, 2008, Volume 37, No. 1

treme non-normality (i.e., values for skew ⬎ 4 formation about participants’ behavior by
and kurtosis ⬎ 19 in the current sample). completing the TRF. Incentives ($5 gift cer-
Moderate to high internal consistency and tificates) were provided upon completion of
test–retest reliability has been established for each teacher rating scale. Information relevant
the CHQ (see Schmidt, Garratt, & Fitzpatrick, to participants’ GPAs, FCAT scores, and at-
2002; Waters, Salmon, Wake, Wright, & Hes- tendance was obtained from school records.
keth, 2001). Validity has been supported in
previous research of the CHQ scores in youth Overview of Analyses
with physical health conditions (see Landgraf
et al., 1999). Coefficient alpha for the CHQ- All analyses were conducted using SAS
CF87 subscales used in the current study ranged version 9.1. Eight multivariate outliers were
from .81 (General Health) to .88 (Limitations in excluded from the final data set; thus, data
Family Activities). from 341 participants were used in all subse-
quent analyses. Students were classified into
Procedures mental health groups based on national norms
provided for commercially available instru-
Written approval to conduct the research ments and sample norms for all other instru-
was secured from the principal investigator’s ments. Because student SES and family com-
university institutional review board and the position (i.e., parent marital status) were con-
participating school district. The principal in- sistently differentially represented in mental
vestigator explained study goals and proce- health groups, these variables were entered as
dures to all regular education teachers and covariates in all subsequent analyses to control
students in the participating school. Although for influences of the demographic variables on
not every student was compensated for partic- the dependent variables. Three between-
ipation, incentives such as gift certificates groups multivariate analyses of covariance us-
were provided to students selected at random ing General Linear Model Method I (Type
from the larger body of student participants. III), which adjusts for unequal sample sizes
The principal investigator distributed letters within cells, were conducted to determine if
that outlined the goals and procedures of the mental health groups differed on educational,
project to parents of all students and requested social, and physical health variables, respec-
active consent for student participation. Ap- tively, after controlling for the influences of
proximately 40% of students returned signed covariates on the outcomes. Univariate analy-
parent consent forms. In the spring of 2006, ses of covariance were examined for each
students with parental consent and student as- indicator within the area of functioning to
sent to participate were administered a brief determine which indicators were driving the
demographics questionnaire (i.e., questions effect. Results of Tukey-Kramer tests and
about students’ ethnic identity, receipt of free group means were examined to identify dif-
or reduced-cost lunch, and marital status of ferences between mental health groups on out-
biological or adoptive parents) and each of the come measures. Comparisons between two
self-report instruments described above in sets of groups were of particular interest. First,
groups of 50 –75 students during the regular differences between youth with complete
school day. Measures were counterbalanced to mental health and their vulnerable peers would
control for order effects. The authors of the support the notion that an absence of psycho-
current study and several research assistants pathology is insufficient to ensure wellness.
were present at data collection sessions to Second, differences between troubled and
ensure valid and confidential completion and symptomatic but content youth would support
collection of all measures. After the collection the notion that intact SWB protects children
of self-report data, teachers familiar with par- with mental illness from experiencing the
ticipants were asked to provide additional in- most deleterious outcomes.
58
 Dual-Factor Model of Mental Health

Results Similar to the method used for psychopathol-

ogy, all students above the 30th percentile on
Mental Health Groups Yielded Through SWB (z score ⱖ 0.76; raw scores of life sat-
a Dual-Factor Model isfaction ⱖ 4.10, positive affect ⱖ 3.34, neg-
ative affect ⱕ 1.22) were classified as average
SWB scores and psychopathology
to high SWB, whereas students below the 30th
scores were used to determine the existence
percentile (z score ⬍ 0.76; raw scores of life
and sample size (n) of the four proposed men-
satisfaction ⬍ 4.10, positive affect ⬍ 3.34,
tal health status groups (see Table 1). Consis-
negative affect ⬎ 1.22) were classified as low
tent with previous research (Kasser & Shel-
SWB.
don, 2002; Sheldon, Kasser, Houser-Marko,
Of note, previous research has dichoto-
Jones, & Turban, 2005), an aggregate SWB
mized global life satisfaction (one component
variable was created by standardizing and
of SWB) using a cut point that corresponds to
summing scores for life satisfaction and posi-
a raw score of 4.0 (on the 1– 6 scale), in which
tive affect, then subtracting standardized neg-
scores above 4.0 are considered high life sat-
ative affect scores. All subsequent analyses
isfaction and scores below 4.0 are considered
were conducted using the aggregate SWB
low life satisfaction (e.g., Suldo & Huebner,
variable.
2004a). When applying the raw score cut point
Regarding group assignments, partici- of 4.0 for life satisfaction to the sample, ap-
pants were first classified according to mental proximately 25% of participants reported
health problems. High psychopathology was global life satisfaction scores below 4.0 and
defined according to published, gender-spe- would be considered to have low life satis-
cific norms for the Achenbach System of Em- faction. Thus, it is plausible that roughly
pirically Based Assessment (Achenbach & 30% of the sample would indeed be consid-
Rescorla, 2001). Scores within the “at-risk” or ered low on the SWB composite variable as
“clinically significant” range on either self- well. Because of the paucity of research on
reported internalizing symptoms or teacher- children’s affect, similar classifications of
rated externalizing symptoms were grouped as high levels of positive or negative affect
high psychopathology. Selection of infor- have not been published.
mants was based on the notion that children Finally, a new variable representing
may not be highly reliable reporters of their mental health groups was created based on
externalizing or acting out behaviors, but are participants’ dichotomized scores on SWB
the most informed reporters of their internal- and psychopathology. A summary of the cut
izing symptoms. To be considered at risk or points used to assign participants to mental
clinically significant, participant scores had to health groups is included in Table 1. Demo-
be at or above a T score of 60 on one or both graphic characteristics for each of the four
groups of symptoms (Achenbach & Rescorla, mental health groups are presented in Table 2.
2001). Using these criteria, 103 of the 441
participants (30%) met criteria for high psy- Group 1: Youth with complete men-
chopathology. The remaining 70% of partici- tal health. This subgroup of 194 children
pants were in the normal range of symptoms (57% of the sample) scored in the low to
based on T scores below 60, and were thus average range on self-reported internalizing
classified as low psychopathology. symptoms and teacher-rated externalizing
Because no published norms for SWB symptoms and reported satisfactory levels of
are available in the current literature, decision SWB. Z ratios and associated two-tailed prob-
points for high and low SWB corresponded abilities were calculated to test the signifi-
with the proportion of students classified as cance of the differences between proportions
having high or low psychopathology. A raw of participants in each subgroup compared to
SWB composite score corresponding to the the total sample on each demographic vari-
30th percentile was chosen as the cut point. able. As shown in Table 2, group composition
59
School Psychology Review, 2008, Volume 37, No. 1

Table 2
Demographic Characteristics of Participants in Mental Health Groups
(N ⴝ 341)
Mental Health Group

Complete Symptomatic
Mental Health Vulnerable but Content Troubled Total Sample
Demographic (n ⫽ 194) (n ⫽ 44) (n ⫽ 44) (n ⫽ 59) (N ⫽ 341)
Variable % % % % %

Gender
Male 41.75 36.36 45.45 37.29 40.76
Female 58.25 63.64 54.55 62.71 59.24
Ethnicity
American Indian 0.52 — — 6.78* 1.47
Asian 6.70 — 2.27 6.78 5.28
African American 9.28 11.36 25.00 23.73 14.08
Hispanic or Latino 12.37 20.45 9.09 10.17 12.61
White 60.31 56.82 45.45 45.76 55.43
Multiracial 8.76 11.36 15.91 6.78 9.68
Unknown 2.06 — 2.27 — 1.47
Socioeconomic status
Low 16.49* 34.09 31.82 38.98* 24.63
Average or high 83.51* 65.91 68.18 61.02* 75.37
Family composition
Married parents 70.10* 50.00 59.09 37.29* 60.41
Parents not married 29.90* 50.00 40.91 62.71* 39.59
Grade level
6 32.99 36.36 27.27 33.90 32.84
7 37.11 36.36 45.45 25.42 36.07
8 29.90 27.27 27.27 40.68 31.09

Note. The z tests were employed to test the significance of the difference between proportions of participants in each
mental health group and the total sample.
*p ⬍ .05.

approximates the makeup of the total sample Group 3: Symptomatic but content
with respect to gender, ethnicity, and grade youth. Approximately 13% of participants
level. Students with low SES and/or whose (n ⫽ 44) had high psychopathology scores and
parents are not married are significantly un- reported average to high SWB. This group
derrepresented, and high SES students and/or approximates the composition of the total
students with married parents are overrepre- sample on all demographic characteristics.
sented ( p ⬍ .05).
Group 4: Troubled youth. These 59
Group 2: Vulnerable youth. For children (17% of sample) had high psychopa-
roughly 44 children (13% of the sample), al- thology and low SWB. Group composition
though psychopathology scores were not high, approximates the total sample with respect to
SWB scores were low. Group composition gender and age. Youth who are American In-
approximates the total sample with respect to dian, low SES, and/or have unmarried parents
all demographic characteristics. are overrepresented, whereas youth from high
60
 Dual-Factor Model of Mental Health

Table 3
Mean Levels of Educational Functioning by Mental Health Group
(N ⴝ 341)

Mental Health Group

Complete Symptomatic
Mental Health Vulnerable but Content Troubled
(n ⫽ 194) (n ⫽ 44) (n ⫽ 44) (n ⫽ 59)

Dependent Variable M SD M SD M SD M SD

Grade point average1,2 3.62a 0.52 3.31a,b 0.71 3.26b 0.64 3.15b 0.71
(3.57) (3.36) (3.29) (3.26)
FCAT—reading level1 3.61a 1.11 2.95b 1.03 3.21a,b 1.04 2.91b 1.13
(3.53) (3.07) (3.27) (3.09)
FCAT—math level1,2 3.93a 1.11 3.36a,b 1.03 3.23b 1.23 3.05b 1.31
(3.82) (3.49) (3.30) (3.28)
School absences2 3.98a 4.01 6.25b 7.63 5.72 5.17 5.61 4.75
(4.14) (6.31) (5.68) (5.21)
Academic self-perception 6.11a 0.74 5.31b 1.17 5.63b 1.05 5.08b 1.32
Motivation and self-
regulation 5.94a 0.88 5.12b 1.19 5.39b 1.24 4.87b 1.33
Valuing of school1 6.74a 0.60 6.31b 1.02 6.42a,b 1.18 6.34b 0.80
(6.73) (6.33) (6.43) (6.35)
Attitudes toward school 5.92a 1.21 5.36a,b 1.50 5.09b,c 1.49 4.45c 1.84

Note. Tukey-Kramer comparisons were employed to analyze group means in cases of significant F tests. Significant
differences between group means are indicated by different letters. Means having the same subscript are not significantly
different. Means not marked by letters are not significantly different from any group means. In the cases of dependent
variables that were significantly related to covariates, adjusted means are presented in parentheses.
1
Dependent variables significantly related to the covariate of socioeconomic status.
2
Dependent variables significantly related to the covariate of parent marital status.

SES and/or married parent families are tion) were significantly affected by group
underrepresented. membership, F(24, 929) ⫽ 3.96, p ⬍ .001.
After adjustment by covariates, univariate
Differences in Adjustment Among tests (analyses of covariance) for each school-
Mental Health Groups ing variable reached statistical significance
Educational functioning. A between- ( p ⬍ .05), indicating that academic function-
subjects multivariate analysis of covariance ing differs among children with different men-
tested the main effect of mental health group tal health profiles even after the relationships
on academic functioning. Adjustment was between demographic variables (SES, parent
made for the two covariates (SES and parent marital status) and academic functioning are
marital status) that accounted for dispropor- controlled for statistically. Results of fol-
tionality in two of the groups, as described low-up analyses with Tukey-Kramer tests are
earlier. With use of Wilks’s criterion, the com- shown in Table 3, along with unadjusted
bined dependent variables (FCAT—reading, means and standard deviations for each group
FCAT—math, GPA, attendance, academic on each aspect of educational functioning. Ad-
self-perceptions, attitudes towards school, val- justed means are also presented for dependent
uing of school, and motivation or self-regula- variables that were significantly related to one
61
School Psychology Review, 2008, Volume 37, No. 1

Table 4
Mean Levels of Social Adjustment by Mental Health Group (N ⴝ 341)

Mental Health Group

Complete Symptomatic
Mental Health Vulnerable but Content Troubled
(n ⫽ 194) (n ⫽ 44) (n ⫽ 44) (n ⫽ 59)

Dependent Variable M SD M SD M SD M SD

Social problems 2.07a 1.96 3.57b 2.37 3.68b 2.81 6.46c 3.51
SS: Classmates1 4.52a 1.04 3.80b 1.13 4.43a 1.11 3.81b 1.39
(4.56) (3.74) (4.41) (3.71)
SS: Parents2 5.16a 0.81 4.23b 1.23 4.87a 0.93 3.63c 1.28
(5.13) (4.26) (4.87) (3.70)
SS: Teachers 5.02a 0.89 4.68a 1.03 4.80a 1.08 4.05b 1.26

Note. SS ⫽ social support. Tukey-Kramer comparisons were employed to analyze group means in cases of significant
F tests. Significant differences between group means are indicated by different letters. Means having the same subscript
are not significantly different. In the cases of dependent variables that were significantly related to covariates, adjusted
means are presented in parentheses.
1
Dependent variables significantly related to the covariate of socioeconomic status.
2
Dependent variables significantly related to the covariate of parent marital status.

or more covariate. The educational function- Social functioning. A second be-

ing of the youth with complete mental health tween-subjects multivariate analysis of covari-
was superior to that of their vulnerable peers ance tested the main effect of mental health
on five of eight indicators. Regarding objec- group on social functioning after adjusting for
tive indicators of academic achievement, the two covariates. With use of Wilks’ crite-
youth with complete mental health scored rion, the combined dependent variables (social
higher on the reading portion of the FCAT and problems, peer support, parent support,
had better school attendance. Although the teacher support) were significantly affected by
complete mental health and vulnerable groups group membership, F(12, 876) ⫽ 17.18, p ⬍
had statistically similar mean GPAs and scores .001. After adjustment by covariates, univari-
on the math portion of the FCAT, mean ate analysis of covariance tests for each social
achievement scores for the complete mental functioning variable reached statistical signif-
health group (but not the vulnerable group) icance, indicating that social functioning dif-
significantly exceeded the scores of the two fers among children with different mental
groups with high levels of psychopathology. health profiles. Results of follow-up analyses
Regarding attitudes towards schooling and ed- with Tukey-Kramer tests are shown in Table
ucation that are associated with academic suc- 4, along with means and standard deviations
cess, youth with complete mental health re- for each group on each aspect of social func-
ported higher perceptions of their academic tioning. Social functioning of the youth with
abilities and of the value of schooling, as well complete mental health was superior to that of
more efforts directed towards self-regulation their vulnerable peers on three of four indica-
of academic behaviors. The educational func- tors of social functioning. Regarding interac-
tioning of the symptomatic but content youth tions with peers, youth with complete mental
did not differ from that of their troubled peers health reported fewer social problems (e.g.,
on any indicators of educational functioning. loneliness, difficulty getting along with others,
62
 Dual-Factor Model of Mental Health

Table 5
Mean Levels of Physical Health by Mental Health Group (N ⴝ 341)

Mental Health Group

Complete Symptomatic
Mental Health Vulnerable but Content Troubled
(n ⫽ 194) (n ⫽ 44) (n ⫽ 44) (n ⫽ 59)

Dependent Variable M SD M SD M SD M SD

General health perceptions 4.12a 0.52 3.71b 0.56 3.81b 0.63 3.29c 0.60
Bodily pain 5.14a 0.78 4.97a 0.87 4.95a 1.07 4.08b 1.03
Role limitations: family
activities1 4.53a 0.64 3.98b 0.96 4.04b 0.89 3.42c 0.97
(4.50) (4.02) (4.07) (3.47)

Note. Tukey-Kramer comparisons were employed to analyze group means in cases of significant F tests. Significant
differences between group means are indicated by different letters. Means having the same subscript are not significantly
different. In the cases of dependent variables that were significantly related to covariates, adjusted means are presented
in parentheses.
1
Dependent variables significantly related to the covariate of socioeconomic status.

preference for younger friends) and greater lems. The self-perceived physical health of the
social support from classmates and parents. symptomatic but content youth was superior to
The perceived social functioning of the symp- that of their troubled peers on all indicators.
tomatic but content youth was superior to that
of their troubled peers on all four indicators. Discussion

Physical health. A third between-sub- Results of the current study supported a

jects multivariate analysis of covariance tested four-factor model of mental health, as shown
the main effect of mental health group on in Table 1. Complete mental health was
physical health after adjusting for the two co- present in 57% of a typical sample of middle
variates. With use of Wilks’s criterion, the school students. The 13% of students identi-
combined dependent variables (general health, fied as vulnerable do not exhibit clinical levels
bodily pain, limitations in family activities) of mental illness, but report a relatively poor
were significantly affected by group member- quality of life. Vulnerable adolescents are pos-
ship, F(9, 811) ⫽ 17.18, p ⬍ .001. After sibly in need of assistance but are usually
adjustment by covariates, univariate tests for excluded from study and services because of
each physical health variable reached statisti- psychology’s traditional focus on psychopa-
cal significance, indicating that physical health thology. Thirteen percent of students were
differs among children with different mental symptomatic but content; although identified
health profiles. Results of follow-up analyses as at least minimally distressed, they do not
with the Tukey-Kramer test are shown in Ta- appear to suffer to the same extent because of
ble 5, along with means and standard devia- their positive appraisals of life and relatively
tions for each group on each aspect of physical frequent positive emotions. The troubled sub-
health. Youth with complete mental health group includes the 17% of youth who likely
reported better general health (e.g., sick less receive the bulk of psychological attention
frequently, better feelings about their overall because they manifest clinical levels of mental
health) and fewer limitations in their family health problems and report the quality of their
activities because of health or behavior prob- lives is poor. The current study extends a
63
School Psychology Review, 2008, Volume 37, No. 1

four-group classification system of children’s the absence of mental illness is not sufficient
mental health (cf. Greenspoon & Saklofske, to guarantee optimal academic achievement.
2001) to an additional developmental period— Students with high psychopathology were at
early adolescence. risk for inferior academic functioning (e.g.,
The current study illustrated that posi- lower FCAT scores and school grades) regard-
tive and negative indicators of mental health less of their SWB, suggesting that average to
are not the opposite ends of the same contin- high SWB alone is also not a sufficient con-
uum. Specifically, not all students who dis- dition for academic success. Regarding the
played high psychopathology (30% of the clinical significance of differences between
sample) also reported low SWB; instead, al- groups, the average FCAT reading score in
most half of the high-psychopathology group both groups of students with low SWB was
(13% of total sample) reported average to high below 3 (the level required for grade promo-
levels of SWB. These findings are consistent tion). Average FCAT reading scores were
with a separate study of a clinical sample of above 3 in both groups with average to high
126 children and adolescents referred for out- SWB, even those students who had clinical
patient psychiatric treatment (Bastiaansen et levels of psychopathology.
al., 2005). At 1-year follow-up, 11% of youth Regarding interpersonal functioning, the
still manifested clinical levels of psychopa- current study also confirmed that symptomatic
thology but experienced improvements in sub- but content youth perceive more positive in-
jective quality of life, demonstrating that qual- terpersonal relationships with peers (Green-
ity of life can be positive in the face of per-
spoon & Saklofske, 2001) as well as perceive
sistent mental health problems. Also notable,
more social support from important adults in
not all children in the current study without
their lives than troubled youth. Consistent
symptoms of mental illness reported desirable
with previous research with adolescents dem-
levels of wellness; indeed, 13% of the sample
onstrating that low parental support co-occurs
was vulnerable (i.e., low life satisfaction, in-
with increased symptoms of internalizing and
frequent positive emotions, frequent negative
externalizing psychopathology (Bean, Barber,
emotions).
& Crane, 2006) and diminished life satisfac-
In addition to supporting the mere exis-
tence of a dual-factor model of mental health, tion (Suldo & Huebner, 2006), the troubled
findings in the current study illustrated the youth in the current study perceived the lowest
utility of such a model by identifying differ- levels of social support from parents. On the
ences in functioning between each mental majority of social functioning indicators, stu-
health group. On the majority of educational dents with average to high SWB experienced
functioning indicators, students with complete better social relationships than their peers with
mental health were more academically suc- comparable levels of psychopathology, which
cessful than their vulnerable peers. Vulnerable highlights the additive information SWB
youth have diminished academic self-concept, scores provide in predicting social function-
view school as less important for long-term ing. Teachers were the only social support
goals, and have reduced motivation to self- source on which students with complete men-
regulate behaviors necessary for learning rel- tal health and their vulnerable peers did not
ative to youth with complete mental health. differ. Suldo and Huebner (2006) found that
Vulnerable youth also perform worse on an middle and high school students with very low
objective measure of reading achievement (the and average levels of life satisfaction per-
FCAT) and are absent from school more often ceived similar levels of teacher support, which
than those students with complete mental suggests that relationships with teachers are
health. Although longitudinal research is not as highly associated with students’ SWB
needed, these findings suggest that fostering as are relationships with other important
SWB in all children may be essential to attain adults, such as parents (Suldo & Huebner,
maximum positive academic functioning, as 2004b).
64
 Dual-Factor Model of Mental Health

By examining students’ physical health, tion, and overall health, as well as more social
the current study also extended the range of problems and limitations in family activities
outcomes associated with a dual-factor model because of health problems, when compared
of mental health in youth. On the majority of to their peers with complete mental health.
physical health variables, students with aver- Advances in the assessment of children’s
age to high SWB were healthier than their SWB have resulted in the availability of brief,
peers with comparable levels of psychopathol- validated measures of life satisfaction (the
ogy but lower SWB, which underscores the most stable indicator of SWB) appropriate for
additive information SWB scores provide in use with children and adolescents. In the cur-
predicting students’ physical functioning. Spe- rent study, the seven-item SLSS functioned as
cifically, vulnerable youth reported lower av- a sensitive and reliable measure of life satis-
erage levels of general health and more limi- faction in middle school students. School psy-
tations in family activities because of health or chologists and health care providers should
behavior problems than youth with complete consider including the SLSS (available for
mental health. Relative to their troubled peers free from the author) in routine assessments of
with comparable levels of high psychopathol- children (e.g., comprehensive psychological
ogy but lower SWB, symptomatic but content examinations, health and wellness checks, and
youth reported better general health, experi- so on). Huebner and colleagues (2007) pro-
enced less bodily pain, and experienced fewer vide a comprehensive review of the SLSS and
limitations in their family activities. Whereas other life satisfaction measures for youth and
previous research with pediatric samples an example of how to incorporate information
found that children with a severe health prob- from life satisfaction scales in psychoeduca-
lem (cancer) had levels of SWB comparable to tional evaluations.
healthy controls (McKnight, 2004), the current The use of both SWB and psychopathol-
study suggests that, in a general sample of ogy assessments may also more accurately
youth, dimensions of current health are related identify students at each level of a school’s
to SWB. The associations between SWB and multitiered model of service delivery. All stu-
physical health in the nonclinical sample sup- dents would likely benefit from the provision
port a link between positive indicators of men- of school-wide programming (i.e., Tier 1)
tal health and physical functioning in youth, such as promoting positive student–teacher re-
which augments the literature that identifies lationships or rewarding supportive peer inter-
poor physical health outcomes for youth with actions (Sprague & Horner, 2006). A brief
mental disorders (Bardone et al., 1998). measure of SWB such as the SLSS could be
used to screen for vulnerable youth. These
Implications for Practice youth constitute an at-risk group likely to ben-
efit from Tier 2, small-group strategies with
The current study suggests that examin- increased focus on promoting well-being.
ing students’ psychopathology in isolation Tier 2 interventions for symptomatic but con-
may lead to an over- or underestimation of tent youth would address symptoms of psy-
their functioning in important areas of life. chopathology while continuing to promote
The additive information provided by assess- SWB. Finally, interventions with the greatest
ments of SWB supports the utility of assessing intensity would be provided to troubled stu-
positive indicators of self-perceived wellness. dents; intervention plans for these students
SWB assessments should supplement, not re- would be individualized and include strategies
place, traditional negative indicators; the im- to reduce mental illness while at the same time
portance of monitoring psychopathology was working to increase well-being.
supported by findings that demonstrated that The results of the study underscore the
symptomatic but content youth had inferior superior functioning of adolescents with aver-
school grades, math achievement, perceived age to high SWB; an absence of mental health
academic abilities, motivation or self-regula- problems may not be sufficient to produce
65
School Psychology Review, 2008, Volume 37, No. 1

optimal academic achievement, social rela- peer, and parent appraisals of student interper-
tionships, nor physical health. If mental health sonal functioning, as well as augment the
professionals and educators continue to focus scant literature on the relationship between
exclusively on identifying and treating mental physical health and mental health in youth by
illness, optimal functioning may be less likely examining objective indicators of physical
for the majority of students. Instead, monitor- health (e.g., frequency of visits to physician,
ing and attempting to increase SWB in all body mass index levels, heart rates), to reduce
youth may yield broader effects. This notion any bias associated with student self-report.
has implications for how mental health profes- Given that the current study underscored the
sionals screen and identify children in need of salience of high SWB to students’ functioning,
assistance, how government funding agencies well-designed intervention studies are needed
(e.g., National Institute of Mental Health) se- to determine how mental health professionals
lect which intervention studies should be can improve children and adolescents’ SWB.
funded (i.e., the current focus on ameliorating Long-term follow-up studies should determine
mental illness), and how educators conceptu- the effects of such interventions on students’
alize which children are ready to perform best functioning at school, interpersonal relation-
at school. ships, and physical health.
Despite the limitations enumerated, the
Limitations and Suggestions for Future current study advances the literature by pro-
Research viding the first known test of a dual-factor
model of mental health in adolescents. Results
The current study is limited to a conve- illustrate the notion that subjective well-being
nience sample of students from one commu- and mental illness are not opposite ends of the
nity who attend a single school. The active same continuum. Instead, approximately 25%
consent procedures contributed to a response of adolescents report quality of life that is
rate of only 40% and some demographic counterintuitive to what may be expected
groups were underrepresented, which may based solely on knowledge of their psychopa-
have affected the results in unexpected ways. thology. Identifying and understanding these
For instance, students from low SES families subgroups of vulnerable and symptomatic but
were somewhat underrepresented in the sam- content children is important because of the
ple, and the low SES students at the school differences in educational, social, and physical
who did participate were particularly likely to health adjustment they experience relative to
be in the troubled group. Additional research their peers with comparable levels of mental
with representative samples of economically illness. Results of the current study suggest
diverse youth (as well as students from differ- that the simple absence of clinical levels of
ent communities) is needed to test generaliz- problems is not sufficient to guarantee the best
ability of the current findings. Additional re- adjustment in areas of life central to healthy
search is also needed with high school stu- child development.
dents, to determine if a dual-factor model of
mental health exists in older adolescents. An- References
other limitation of the current study is the
cross-sectional design; variables referred to as Achenbach, T. M., & Rescorla, L. A. (2001). Manual for
the ASEBA school-age forms and profiles. Burlington:
“outcomes” could actually either precede or University of Vermont, Research Center for Children,
result from adolescents’ mental health status. Youth, and Families.
A longitudinal design that tracks student char- Bardone, A. M., Moffitt, T. E., Caspi, A., Dickson, N.,
Stanton, W. R., & Silva, P. A. (1998). Adult physical
acteristics over time would help clarify the health outcomes of adolescent girls with conduct dis-
development of SWB. The current study is order, depression, and anxiety. Journal of the Ameri-
also limited by its reliance on self-report data can Academy of Child & Adolescent Psychiatry, 37,
594 – 601.
to assess all social and physical health out- Bastiaansen, D., Koot, H. M., & Ferdinand, R. F. (2005).
comes. Future research should include teacher, Psychopathology in children: Improvement of quality

66
 Dual-Factor Model of Mental Health

of life without psychiatric symptom reduction? Euro- Malecki, C. K., Demaray, M. K., & Elliot, S. N. (2000). A
pean Child & Adolescent Psychiatry, 14, 354 –370. working manual on the development of the Child and
Bean, R. A., Barber, B. K., & Crane, D. R. (2006). Adolescent Social Support Scale. DeKalb: Northern
Parental support, behavioral control, and psychological Illinois University.
control among African American youth. Journal of Martin, K. M., & Huebner, E. S. (2007). Peer victimiza-
Family Issues, 27, 1335–1355. tion and prosocial experiences and emotional well-
Cowen, E. L. (1994). The enhancement of psychological being of middle school students. Psychology in the
wellness: Challenges and opportunities. American Schools, 44, 199 –208.
Journal of Community Psychology, 22, 149 –179. McCoach, D. B., & Siegle, D. (2001). A comparison of
Diener, E. (2000). Subjective well-being: The science of high achievers’ and low achievers’ attitudes, percep-
happiness and a proposal for a national index. Ameri- tions, and motivations. Academic Exchange Quar-
can Psychologist, 55, 34 – 43. terly, 5, 71–76.
Diener, E., Lucas, R. E., & Oishi, S. (2002). Subjective McCoach, D. B., & Siegle, D. (2003). The School Atti-
well-being: The science of happiness and life satisfac- tude Assessment Survey—Revised: A new instrument
tion. In C. R. Snyder & S. Lopez (Eds.), Handbook of to identify academically able students who under-
positive psychology (pp. 463– 473). London: Oxford achieve. Educational and Psychological Measure-
University Press. ment, 63, 414 – 429.
Evans, D. L., Foa, E. B., Gur, R. E., Hendin, R., O’Brien, McKnight, C. G. (2004). Subjective well-being in pediat-
C. P., Seligman, M. E. P., et al. (Eds.). (2005). Treating ric oncology patients. Unpublished doctoral disserta-
and preventing adolescent mental health disorders: tion, University of South Carolina, Columbia.
What we know and what we don’t know: A research Michalos, A. C., & Zumbo, B. D. (2002). Healthy days,
agenda for improving the mental health of our youth. health satisfaction, and satisfaction with the overall qual-
New York: Oxford University Press. ity of life. Social Indicators Research, 59, 321–338.
Florida Department of Education. (2005). Florida Com- Park, N. (2004). The role of subjective well-being in
prehensive Assessment Test® (FCAT): Assessment and positive youth development. Annals of the American
School Performance. Retrieved February 15, 2007, Academy of Political & Social Science, 591, 25–39.
from http://www.firn.edu/doe/sas/fcat/fcatscor.htm Pine, D. S., Cohen, P., Brook, J., & Coplan, J. D. (1997).
Gilman, R., & Huebner, E. S. (1997). Children’s reports Psychiatric symptoms in adolescence as predictors of
obesity in early adulthood: A longitudinal study. Amer-
of their life satisfaction: Convergence across raters,
ican Journal of Public Health, 87, 1303–1310.
time, and response formats. School Psychology Inter-
Rohde, P., Lewinsohn, P. M., & Seeley, J. R. (1994). Are
national, 18, 229 –243.
adolescents changed by an episode of major depres-
Greenspoon, P. J., & Saklofske, D. H. (2001). Toward an
sion? Journal of the American Academy of Child and
integration of subjective well-being and psychopathol-
Adolescent Psychiatry, 33, 1289 –1298.
ogy. Social Indicators Research, 54, 81–108.
Ryff, C. D., & Singer, B. (1998). The contours of positive
Huebner, E. S. (1991). Initial development of the Stu-
human health. Psychological Inquiry, 9, 1–28.
dents’ Life Satisfaction Scale. School Psychology In- Schmidt, L. J., Garratt, A. M., & Fitzpatrick, R. (2002).
ternational, 12, 231–240. Child/parent-assessed population health outcome mea-
Huebner, E. S., Gilman, R., & Suldo, S. M. (2007). sures: A structured review. Child: Care, Health, &
Assessing perceived quality of life in children and Development, 28, 227–237.
youth. In S. R. Smith & L. Handler (Eds.), Clinical Seligman, M. E. P., & Csikszentmihalyi, M. (2000). Pos-
assessment of children and adolescents: A practitio- itive psychology: An introduction. American Psychol-
ner’s guide (pp. 347–363). Mahwah, NJ: Erlbaum. ogist, 55, 5–14.
Kasser, T., & Sheldon, K. M. (2002). What makes for a Seligson, J. L., Huebner, E. S., & Valois, R. F. (2005). An
merry Christmas? Journal of Happiness Studies, 3, investigation of a brief life satisfaction scale with ele-
313–329. mentary school children. Social Indicators Re-
Keyes, C. L. M. (2003). Complete mental health: An search, 73, 355–374.
agenda for the 21st century. In C. L. M. Keyes & J. Sheldon, K. M., Kasser, T., Houser-Marko, L., Jones, T.,
Haidt (Eds.), Flourishing: Positive psychology and the & Turban, D. (2005). Doing one’s duty: Chronological
life well-lived (pp. 293–312). Washington, DC: Amer- age, felt autonomy, and subjective well-being. Euro-
ican Psychological Association. pean Journal of Personality, 19, 97–115.
Keyes, C. L. M. (2006). Mental health in adolescence: Is Snyder, C. R., Lopez, S. J., Edwards, L. M., Pedrotti, J. T.,
America’s youth flourishing? American Journal of Or- Prosser, E. C., Walton, S. L., et al. (2003). Measuring and
thopsychiatry, 76, 395– 402. labeling the positive and the negative. In S. J. Lopez &
Landgraf, J. M., Abetz, L., & Ware, J. E., Jr. (1999). Child C. R. Snyder (Eds.), Positive psychological assessment: A
Health Questionnaire (CHQ): A user’s manual. Bos- handbook of models and measures. (pp. 21–39). Wash-
ton, MA: HealthAct. ington, DC: American Psychological Association.
Laurent, J., Catanzaro, J., Joiner, T. E., Rudolph, K., Sprague, J. R., & Horner, R. H. (2006). Schoolwide pos-
Potter, K. I., Lambert, S., et al. (1999). A measure of itive behavioral supports. In S. R. Jimerson & M.
positive and negative affect for children: Scale devel- Furlong (Eds.), Handbook of school violence and
opment and preliminary validation. Psychological As- school safety: From research to practice (pp. 413–
sessment, 11, 326 –338. 327). Mahwah, NJ: Lawrence Erlbaum Associates.
Malecki, C. K., Demaray, M. K. (2002). Measuring per- Suldo, S. M., & Huebner, E. S. (2004a). Does life satis-
ceived social support: Development of the Child and faction moderate the effects of stressful life events on
Adolescent Social Support Scale. Psychology in the psychopathological behavior during adolescence?
Schools, 39, 1–18. School Psychology Quarterly, 19, 93–105.

67
School Psychology Review, 2008, Volume 37, No. 1

Suldo, S. M., & Huebner, E. S. (2004b). The role of life adolescents: A school-based population study of the
satisfaction in the relationship between authoritative self-report Child Health Questionnaire. Journal of Ad-
parenting dimensions and adolescent problem behav- olescent Health, 29, 140 –149.
ior. Social Indicators Research, 66, 165–195. Wilkinson, R. B., & Walford, W. A. (1998). The mea-
Suldo, S. M., & Huebner, E. S. (2006). Is extremely high surement of adolescent psychological health: One or
life satisfaction during adolescence advantageous? So- two dimensions? Journal of Youth and Adoles-
cial Indicators Research, 78, 179 –203. cence, 27, 443– 455.
Suldo, S. M., Riley, K., & Shaffer, E. J. (2006). Academic Zullig, K. J., Valois, R. F., Huebner, E. S., & Drane, J. W.
correlates of children and adolescents’ life satisfaction. (2005). Adolescent health-related quality of life and
School Psychology International, 27, 567–582. perceived satisfaction with life. Quality of Life Re-
Suldo, S. M., Shaffer, E. J., & Shaunessy, E. (in press). An search, 14, 1573–1584.
independent investigation of the validity of the School
Attitudes Assessment Survey—Revised. Journal of
Psychoeducational Assessment. Date Received: March 20, 2007
Waters, E. B., Salmon, L. A., Wake, M., Wright, M., & Date Accepted: November 5, 2007
Hesketh, K. D. (2001). The health and well-being of Action Editor: Shane Jimerson 䡲

Shannon Suldo, PhD, is an assistant professor in the School Psychology Program at the
University of South Florida (USF). Her current research interests include subjective
well-being during youth and the social– emotional functioning of adolescents in college
preparatory programs. Her applied interests pertain to school-based mental health ser-
vices, including strengths-based assessment and intervention, as well as universal strat-
egies and targeted interventions to improve students’ life satisfaction. She received her
PhD in School Psychology from the University of South Carolina in 2004.

Emily Shaffer, MA, is a doctoral candidate in the School Psychology Program at USF.
Her research interests are focused on the relationship between children’s physical and
mental well-being. Her current applied work at the USF Children’s Medical Services
Child Development Clinic includes provision of assessment, intervention, and consulta-
tion services for children with emotional and physical health needs and their families.

68