Psychological Empowerment Among Urban Youth

Abstract

Although there are an increasing number of youth development programs that aim to empower young people, there is a dearth of psychometrically sound measures that can be used to assess flexible youth-led organizing and participatory research approaches that tackle a wide range of social and community problems. This study developed and tested measures of psychological empowerment (PE) and self-efficacy for research and action among a sample of 439 ethnically diverse adolescents primarily recruited from public high schools in an urban center. Items for the PE measure were generated through an iterative combination of conducting formative research with our target population while also drawing on existing theory and measures in the field. We conducted a confirmatory factor analysis of the PE measure, testing four factors: adolescents' motivation to influence their school and community settings; participatory behavior; sociopolitical skills; and perceived control. Psychometric analyses for the PE scales and their correlation with adolescents' report of self-esteem, academic achievement, caring relationships with adults at school, and social support from peers are reported; the implications of the present study for research and practice in the youth development and adolescent psychology field are considered.

Full text

http://heb.sagepub.com/
Health Education & Behavior
http://heb.sagepub.com/content/38/4/348
The online version of this article can be found at:
DOI: 10.1177/1090198110373734
2011 38: 348 originally published online 23 May 2011Health Educ Behav
Emily J. Ozer and Marieka Schotland
Psychosocial Functioning
Psychological Empowerment Among Urban Youth : Measure Development and Relationship to
Published by:
http://www.sagepublications.com
On behalf of:
Society for Public Health Education
can be found at:Health Education & BehaviorAdditional services and information for
http://heb.sagepub.com/cgi/alertsEmail Alerts:
http://heb.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
http://heb.sagepub.com/content/38/4/348.refs.htmlCitations:
at UNIV CALIFORNIA BERKELEY LIB on April 10, 2012heb.sagepub.comDownloaded from

Health Education & Behavior
38(4) 348 –356
© 2011 by SOPHE
Reprints and permission:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1090198110373734
http://heb.sagepub.com
Psychological Empowerment Among
Urban Youth: Measure Development and
Relationship to Psychosocial Functioning
Emily J. Ozer, PhD1 and Marieka Schotland, PhD1,2
Abstract
Although there are an increasing number of youth development programs that aim to empower young people, there is
a dearth of psychometrically sound measures that can be used to assess flexible youth-led organizing and participatory
research approaches that tackle a wide range of social and community problems. This study developed and tested measures
of psychological empowerment (PE) and self-efficacy for research and action among a sample of 439 ethnically diverse ado-
lescents primarily recruited from public high schools in an urban center. Items for the PE measure were generated through
an iterative combination of conducting formative research with our target population while also drawing on existing theory
and measures in the field. We conducted a confirmatory factor analysis of the PE measure, testing four factors: adolescents’
motivation to influence their school and community settings; participatory behavior; sociopolitical skills; and perceived control.
Psychometric analyses for the PE scales and their correlation with adolescents’ report of self-esteem, academic achievement,
caring relationships with adults at school, and social support from peers are reported; the implications of the present study
for research and practice in the youth development and adolescent psychology field are considered.
Keywords
psychological empowerment, urban adolescents, measurement
Introduction
In the past decade, the youth development field has paid
increasing attention to “youth empowerment” as the intended
outcome of a growing number of interventions. Diverse pro-
grams now engage youth—particularly youth of color—in a
range of processes, including participatory research and orga-
nizing that are designed to increase youth “voice” and engage-
ment in improving schools and communities and to promote
civic participation (Cammarota & Fine, 2007; Cargo, Grams,
Ottoson, Ward, & Green, 2003; Nieto, 1996; Shor, 1996;
Wallerstein, 1999; Wilson et al., 2006). Although the term
empowerment has been used in a broad fashion to describe
diverse processes and outcomes in multiple fields of practice,
earlier conceptual and empirical work in the field of community
psychology is helpful in clarifying the parameters of the con-
struct at the individual and collective levels. At its most general,
empowerment refers to the process by which people, organiza-
tions, and communities gain mastery over issues of concern
to them (Rappaport, 1987). The consequences or effects of
empowerment on those engaged in these processes are con-
ceptualized as “empowered outcomes” (Zimmerman, 1995).
With respect to psychological effects, empowered outcomes
are expected to reflect key intrapersonal, interpersonal, and
behavioral dimensions including the perception of control
in relevant domains; motivation to control; decision-making
and problem-solving skills; critical understanding of the
sociopolitical environment; and participatory behaviors (Zim-
merman, 1995).
Although there has been rapid growth in the practice of
empowerment-oriented programs for youth, there has been
relatively little specification and operationalization of success-
ful outcomes for participants. This methodological gap under-
mines the rigor of evaluation efforts and the capacity of the
field to strengthen practice and provide evidence of impact.
As discussed in depth by Zimmerman (1995), the empower-
ment construct is challenging to operationalize, partly because
what people need to learn and do in order to gain mastery in
one set of life circumstances and settings versus another can
diverge tremendously. Thus, the task of assessing empowered
outcomes must be specific enough to capture the relevant
dimensions while also capturing core dimensions of mastery
1University of California at Berkeley School of Public Health, Berkeley,
CA, USA
2New York University, Department of Psychology, New York, NY, USA
Corresponding Author:
Emily J. Ozer, 50 University Hall, Berkeley, CA 94720-7360
Email: eozer@berkeley.edu
at UNIV CALIFORNIA BERKELEY LIB on April 10, 2012heb.sagepub.comDownloaded from

Ozer and Schotland 349
and control, resource mobilization, and sociopolitical con-
text and participation. Furthermore, valid assessment of
empowerment-related domains in young people must also
be developmentally appropriate and take into account the
limited power afforded to minors in the legal, political, and
educational systems.
Drawing on theory and methods developed for the study of
psychological empowerment among adults (Rappaport, Swift,
& Hess, 1984; Zimmerman, 2000; Zimmerman & Zahniser,
1991), two recent studies developed quantitative measures
to assess relevant domains of empowerment for youth who
were engaged in antitobacco advocacy efforts (Holden, Evans,
Hinnant, & Messeri, 2005; Marr-Lyon, Young, & Quintero,
2008). Holden and colleagues developed a measure designed
to tap attitudes, beliefs, and skills specifically relevant to
tobacco control and tested the psychometric properties of this
measure among a convenience sample of more than 3,500 youth
who were actively involved in anti-tobacco advocacy groups.
Their measure assessed domain-specific efficacy (i.e., confi-
dence in working effectively against the tobacco industry and
convincing family and friends not to smoke), perceived socio-
political control, participatory competence (i.e., working with
other group members and influencing decisions), knowledge
of resources regarding tobacco issues, assertiveness regarding
tobacco issues, and self-reported advocacy efforts to convince
others to be more concerned about tobacco use. In cross-sectional
analyses, they found that intensity of participation in advocacy
efforts explained variation in dimensions of psychological
empowerment. In another study, Marr-Lyon et al. examined
a group of 112 youth who were already involved in antitobacco
advocacy, assessing specific aspects of behavioral participation
(e.g., giving a presentation about tobacco use, attending a youth
conference), efficacy related to tobacco advocacy, external
organizational involvement, and satisfaction with the work
accomplished.
Unlike the antitobacco advocacy programs studied in prior
research, many youth development programs that use a par-
ticipatory research or youth organizing approach address a
diverse set of social and health issues. In fact, a defining feature
of this kind of youth-led change effort is that the young people
themselves identify the focus of their project and which issues
they want to research and change in their schools, after-schools,
or community settings. Thus, youth at different sites who are
part of the same overall initiative may work on different issues.
This youth-driven, flexible model of youth development raises
major challenges for evaluation of the impact of programs on
youth participants in that it requires measures that can appro-
priately assess relevant domains that are not issue-specific.
Taking a step back from youth development practice to
consider the study of adolescent development more broadly,
it is notable that although there is little research on psychologi-
cal empowerment as a construct per se, several components
of psychological empowerment have been studied in relation
to adolescent psychosocial outcomes. For example, perceived
control was found to be consistently and moderately associated
with self-esteem in a study of adolescents from the United
States and 13 European countries (Grob, Little, Wanner,
Wearing, & Euronet, 1996). Positive control has also been
shown to be inversely related to substance use among an urban
sample of adolescents (Wills, 1994). Furthermore, multiple
studies have demonstrated that participation in organized
extracurricular activities has been linked to higher self-esteem
and academic achievement over time (Fredricks & Eccles,
2008; Mahoney, Cairns, & Farmer, 2003).
In our view, further comprehensive study of the relationship
of adolescents’ psychological empowerment to “traditional”
psychological and educational indicators of adolescent function-
ing such as self-concept, connection to school, and academic
performance is needed. Such research can provide a better inte-
gration of systematic empirical research on adolescent develop-
ment with the practice of interventions intended to promote
adolescents’ participation and power in the settings and systems
in which they develop. To this end, there is a need for measures
of youth psychological empowerment that are not explicitly
linked to any one kind of intervention but rather can be used to
operationalize and investigate the value of this construct more
generally for the study of child and adolescent development.
Focus of Present Study
The first goal of the present study conducted with ethnically
diverse urban adolescents was to contribute to research and
practice in the youth and adolescent development fields by
developing a self-report survey measure of psychological
empowerment that addresses this key gap. Based on the existing
literature (Zimmerman, 1995), we conceptualized our measure
of youth psychological empowerment as consisting of four main
dimensions or factors: sociopolitical skills, motivation to influ-
ence their schools and communities, participatory behavior, and
perceived control. Second, we explored the relationship among
psychological empowerment and other commonly used indica-
tors of adolescent mental health such as self-esteem, academic
performance, and adolescents’ relationships with adults and
peers in their school settings. In this part of the investigation,
we sought to compare our measures of psychological empower-
ment with related but distinct established measures of adolescent
functioning to help establish criterion validity. We expected that
adolescents with higher levels of psychological empowerment
might also demonstrate good functioning in other domains but
that psychological empowerment could be differentiated from
global measures of well-being such as self-esteem and measures
of relationship quality with peers and teachers. Third, beyond
capturing more general dimensions of psychological empower-
ment relevant for urban adolescents, we also sought to develop
a survey measure to assess skills and efficacy in domains such
as communication, advocacy, and research that are often targeted
in flexible youth development approaches such as youth orga-
nizing and participatory action research.
at UNIV CALIFORNIA BERKELEY LIB on April 10, 2012heb.sagepub.comDownloaded from

350 Health Education & Behavior 38(4)
Method
Design and Selection
Sample. The present measurement study was conducted
in the context of a longitudinal experimental study to evaluate
the implementation and impact of a participatory action
research program for youth across multiple urban sites. As
described further below, the majority of the adolescents in this
study were drawn from those who participated in the baseline
assessment of the intervention study, before the experimental
group began the participatory research program.
The sample for this study included 439 high school students
from four schools and several summer programs in a West
Coast urban community. The average age of the adolescent
participants was 16 years (range 13 to 19 years). The sample
was 61% female and 29% Chinese, 28% Latino, 8% African
American, 7% European American, 5% Filipino, and 23%
Other (including Native Americans, Japanese, Korean, Pacific
Islander, Vietnamese, declined to state). Sample recruitment
was intended to maximize the diversity of the adolescents
with respect to ethnicity and academic achievement in order
to develop measures that would be as relevant and applicable
as possible across adolescent groups. The students were recruited
in two ways: First, 391 students were recruited from high schools
who were participating in an ongoing research study conducted
by the first author on youth-led participatory research. The
present study utilizes the baseline survey data for these stu-
dents recruited from elective classes (computer, peer mentoring,
music) at the four schools. One third of these students assessed
at baseline were randomized to participate in a youth-led
research intervention study but none of the students had already
participated at the time of their baseline assessment. The study
schools ranged in size and aggregate academic performance:
Two were large schools ranked in the top three in the district,
whereas the other two were small schools ranked in the bottom
three in the district. The schools also varied with respect to their
geographic location and demographic makeup of the student
body: Two schools had a majority of Chinese American students;
the third had a majority of Latino students; and the fourth had
a majority of African American students. Furthermore, in order
to expand the sample to include adolescents who were not
attending the study schools, we recruited an additional 48 ado-
lescents from summer programs run by community-based
organizations in low-income neighborhoods in the same city.
The ethnic diversity of our sample paralleled that of the overall
school district, in which the high school population was 36%
Chinese, 21% Latino, 12% African American, 9% White, 6%
Filipino, and 17% Other.
Procedure. After receiving approval from the university and
school district institutional review boards, research team mem-
bers visited elective classes at each school to describe the study
and distribute parental consent/student assent forms. Incentives
of water bottles or key chains were used to reward the return
of the consent forms regardless of whether permission was
granted or denied. Consistent with school district policy, only
youth for whom there were signed consent forms from a parent
participated in the study. One to two graduate-level members
of the research team administered the 45-minute survey (about
130 items) and assisted as needed with questions. The admin-
istration procedure was similar for surveys conducted during
the summer programs.
Measures
Psychological empowerment. The measure of psychological
empowerment that was developed and tested in this study ini-
tially consisted of 61 items covering four core conceptual areas:
Sociopolitical skills, motivation to influence, participatory
behavior, and perceived control. Informed by the theoretical
frameworks discussed earlier, these items were generated
through an iterative combination of conducting formative
research with our target population and drawing on existing theory
and measures relevant to these constructs, such as Zimmerman’s
sociopolitical control and leadership competence scales
(Zimmerman & Zahniser, 1991) and perceived autonomy in
school (Developmental Studies Center, 2000). Consistent with
prior theory and research in psychological empowerment, we
had expected that domain specificity might be salient; thus,
we specified the settings (i.e., school, community) for many
of the skill and control-related items.
Our formative research primarily consisted of group inter-
views with prior cohorts of youth from our study population
enrolled in peer mentoring electives at their public high schools
(some of whom were trained in participatory research). In the
development of this measure, we conducted 17 group interviews
across five school sites to help generate any additional items
that were salient to the teenagers’ experience of psychological
empowerment; we also asked the eight classes of students who
took the initial pilot version of the survey to provide feedback
just after survey administration regarding the clarity and mean-
ingfulness of the items. The main suggestion from students was
to reduce the length of the survey, prompting us to reduce our
item pool as much as possible. To provide additional practitioner
perspectives on our measure, we also met with the staff of two
community-based organizations with expertise in implementing
youth empowerment-oriented programs to assess if our items
covered the relevant dimensions and hear any suggestions.
The initial measure tested here consisted of 36 original items
generated as part of the present study and 25 items derived from
existing measures in the field. Sample items include, “I have
spoken with other students about issues that I want to improve
at the school” and “I can usually figure out how to get an adult
to see my point of view, even if they don’t agree with me.”
Response scales were Likert-type 4-point scales from strongly
disagree to strongly agree. This response scale was selected
to eliminate the option of a middle or neutral choice that we
have found in prior research to be a common default choice
among this population.
at UNIV CALIFORNIA BERKELEY LIB on April 10, 2012heb.sagepub.comDownloaded from

Ozer and Schotland 351
Research and action self-efficacy. The research and action
self-efficacy measure consists of 16 items (α = 0.96) that were
generated in this study via formative research in order to tap
students’ perceptions of their own confidence in skills relevant
to research and advocacy. These efficacy items were not spe-
cific to any one issue (e.g., tobacco, school reform) but instead
were intended to be relevant across different issues addressed
by youth research or advocacy (e.g., “I can respectfully discuss
pros and cons of an issue I care about with adults at my school”)
and were rated on a 4-point Likert-type scale (see Table 1 for
full list of items).
Self-esteem. The Self-Esteem Questionnaire (SEQ) (Dubois
et al., 1996) assesses global feelings of self-worth and consists
of eight items (α = .75), such as “I am happy with myself as
a person”; all items are rated on a 4-point Likert-type response
scale. The SEQ was developed for use with middle and high
school students and has demonstrated excellent psychometric
properties.
Academic achievement. For each student whom we recruited
from their schools, we obtained their grade point averages
(GPAs) from school district records and their scores on state
standardized tests in reading and math. GPA is reported on a
5-point scale (0-4); this sample had an average GPA in the
B– to C+ range.
School climate—caring relationships. The three-item Caring
Relationships measure (α = .88) from the California Healthy
Kids Resilience Module (Constantine & Benard, 2001) was
used to identify the availability of teachers or adults who serve
in caring roles for students. Items ask about the presence of
an adult who “cares about me,” “notices when I am not there,”
and “who listens to me when I have something to say.”
Responses are on a 4-point scale.
Sense of community—social support. The four-item Social
Support subscale from the Sense of Community Measure
(Developmental Studies Center, 2000; α = .87) was used to
tap the available social support that students have from other
students, for example, the extent to which students “are willing
to go out of their way to help” and “look out for each other.”
Responses are on a 4-point scale.
Analyses
Several analyses were conducted to refine the original set of
61 psychological empowerment items. After initial review
of the psychometric properties of each item, 8 of the original
items were excluded because of low response rates and an
additional eight of the original items were excluded because
of low response variability. The remaining 45 items were then
tested for reliability across the four projected scales. Items that
were not theoretically central and that did not correlate highly
with other subscale items in the reliability analysis were
removed from the subscales (17 items). We then tested our
hypothesized four-factor model using a confirmatory factor
analysis (CFA). EQS, version 6.1 (Bentler, 2004), was used to
model and estimate all parameters.
Results
Table 3 provides the mean and standard deviation for all major
study variables. Study participants reported midrange levels
of the psychological empowerment dimensions across all four
subscales, ranging from 2.0 to 2.6.
Confirmatory Factor Analysis
Our hypothesized four-factor model was tested using EQS 6.1,
and each of the factors was correlated. Factor loadings were
allowed to vary freely. Factor variances were constrained to one
(1). The model was tested and based on the sample covariance
matrix estimates using maximum likelihood estimation (ML).
ML was used because it is robust against violations of normality
assumptions (Hoyle, 1995) and the majority of the variables
here exhibit some degree of both skewness and kurtosis. Mardia’s
coefficients for each of the models illustrate that they are kurtotic
(Kline, 1998) suggesting that these data are nonnormal. The ML
estimation methods used are robust to skew (absolute value < 3)
and kurtosis (absolute value < 10) (Kline, 1998). Consequently,
corrected statistics are presented and discussed.
Table 1. Research and Action Self-Efficacy Items
1. I can explain guidelines for conducting good research to
another person.
2. I can teach other students to develop their own research
projects on issues of concern for them.
3. I can work well with a group of other students to organize a
program or event at my school.
4. I can make a real difference in improving my school.
5. I can make a real difference in improving my city.
6. I can write a letter to the editor of a newspaper about an
issue I care about.
7. I can make a good presentation to students at my school on
an issue I care about.
8. I can make a good presentation to my teachers and principal
on an issue that I care about.
9. I can start an organizing effort among my peers on an issue
that I care about.
10. I can work well with a group of other students to make a
difference at my school.
11. I can make a presentation to the school board or board of
supervisors on an issue that I care about.
12. I can develop a research tool that gathers useful information.
13. I can respectfully discuss pros and cons of an issue I care
about with youth my age.
14. I can use research results to come up with realistic
recommendations.
15. I can respectfully discuss pros and cons of an issue I care
about with adults at my school.
16. I know how to break down important issues facing youth to
figure out what we can work on.
at UNIV CALIFORNIA BERKELEY LIB on April 10, 2012heb.sagepub.comDownloaded from

352 Health Education & Behavior 38(4)
Goodness-of-fit indices. The overall model fit for the CFA
was assessed using several indices suggested by Hu and
Bentler (1995). The nonnormed fit index (NNFI) and the
comparative fit index (CFI) are both incremental fit indices
that compare the target model to the null model, which speci-
fies no relations among variables where only variances are
estimated. A CFI and NNFI greater than .90 suggest a good
fit. The root mean square error of approximation (RMSEA)
is an index of absolute fit that examines how well the present
model approximates the data; values less than .05 suggest
a “good” fit and less than .08 suggest an “acceptable” fit.
The overall chi square (χ
2
) is also reported although it is only
used to compare the fit of contrasting models. It is the statisti-
cal test of the lack of fit based on overidentifying restrictions
on the target model. This index should not be statistically
significant, although this is a function of sample size and will
usually be rejected with large samples. After these four fit
indices were examined, correlations among subscales and
with the external criteria were analyzed. Our four-factor
model showed adequate fit based on the four fit indices,
χ2(344) = 1006.65, p < .001; NNFI = .82, CFI = .84, and
RMSEA = 0.07.3
Post hoc modification. Inspection of the factor loadings for
our hypothesized four-factor model suggested the removal
of two items that loaded low (i.e., between 0 and .17) on
specified factors. These two items were deleted from the
model to provide a more parsimonious measure. Consistent
with prior research and theory suggesting that context may
be salient in the assessment of psychological empowerment
(Zimmerman, 1995), we had specified the city or school con-
text for multiple items in an effort to assess any domain-
dependent aspects of our psychological empowerment
dimensions. Based on this expectation that items tapping
specific contexts might be correlated with each other, we
conducted post hoc LaGrange Multiplier (LM) tests to exam-
ine if any context-specific correlations were indicated between
items (i.e., correlated error terms). The results of our LM tests
suggested that model fit would be improved with the addition
of four additional parameters to account for context-specific
correlations that tapped the city context; for example, the
items “If issues come up that affect youth in my city, we do
something about it” and “Youth have a say in what happens
in this city” demonstrated correlated errors despite loading
on different factors. These four correlations were all signifi-
cant (p < .05), with r values ranging from .18 to .33.
The final four-factor model with these theoretically con-
sistent modifications showed improved fit based on the
four fit indices, χ2(342) = 926.71, p < .001; NNFI = .84,
CFI = .86, and RMSEA = .06). The final model included
the four hypothesized scales of general sociopolitical skills
(eight items, α = .81), motivation to influence (four items,
α = .80), participatory behavior (eight items, α = .83), and
perceived control (six items, α = .80). (See Table 2 for factor
loadings.)
Correlations With Other Measures
of Adolescent Functioning
Pearson intercorrelations for each subscale and the external
criteria of self-esteem, efficacy in participatory research and
advocacy activities (PAR self-efficacy), student GPA, and
two student-reported school climate variables are presented
in Table 4. As shown, all four psychological empowerment
subscales were positively correlated with each other, with
correlations ranging from .59 to .66. Each of the four subscales
was also positively and moderately correlated with research
and action self-efficacy, self-esteem, caring relationships with
adults at school, and social support from students at school;
correlations ranged from .32 to .72. For the external criterion
of GPA, correlation effect sizes were low (range: –.13 to .07).
Discussion
The present study extends our understanding and assessment
of the construct of psychological empowerment among ado-
lescents in several key ways. First, using formative research
to adapt and supplement existing measures in the field (Devel-
opmental Studies Center, 2000; Zimmerman & Zahniser,
1991), we developed and tested the psychometric properties
of a self-report survey of psychological empowerment among
a primarily school-based sample of ethnically diverse urban
adolescents. We conducted a confirmatory factor analysis
of psychological empowerment that demonstrated a good fit
for a model consisting of four factors: adolescents’ motivation
to influence their school and community settings; participatory
behavior; general sociopolitical skills; and perceived control
in their schools. These subscales demonstrate good internal
consistency and address the need for psychometrically sound
measures to assess these dimensions of psychological empow-
erment among adolescents. Nineteen of the 26 items across
these four factors were new items developed as part of the
present study. Although Zimmerman’s earlier theoretical work
had highlighted motivation to control as a key dimension of
psychological empowerment, to our knowledge this study
provides the first operationalization of this construct among
young people.
Prior theory and research laid the groundwork for the assess-
ment of dimensions of psychological empowerment among
youth involved in health and political advocacy (Holden,
Crankshaw, Nimsch, Hinnant, & Hund, 2004; Holden et al.,
2005); one key way in which the present study advances the
field is in its study of adolescent psychological empowerment
among youth who had not already participated in activism nor
already selected themselves to serve in an activist role. Rather,
our sample consisted of diverse adolescents—primarily Asian
American and Latino—from different types of elective classes
in a set of public high schools that represented a wide range
of academic and geographic diversity in an urban center. An
additional subsample of students from summer programs was
at UNIV CALIFORNIA BERKELEY LIB on April 10, 2012heb.sagepub.comDownloaded from

Ozer and Schotland 353
also included in the study to further diversify the sample. The
heterogeneity of our sample represents an important strength
with respect to the development of measures intended to be
broadly applicable to a wide range of urban adolescents.
Another contribution of this study is its comparison of the
psychological empowerment subscales developed here to more
traditional indicators of adolescents’ psychosocial functioning.
We found that all four psychological empowerment subscales
were moderately and positively correlated (in the .3 to .4 range)
with self-esteem, suggesting that psychological empowerment
is consistent with but does not strongly overlap with this more
global indicator of psychological well-being. Our results fur-
ther showed that adolescents who reported higher levels of
psychological empowerment were also likely to report more
caring relationships with adults at school and greater perceived
social support from peers at school (correlations in the .4 to
.6 range). This pattern of positive associations makes concep-
tual sense in that all of these dimensions represent various
forms of adolescents’ prosocial engagement in key develop-
mental settings of school and community. Interestingly, only
Table 2. Standardized Factor Loadings for Four Factor Model
Item Factor Loadings
Sociopolitical skills—general 1 2 3 4
I feel like I have a pretty good understanding of the important political issues
which confront our society.a
.42
I am often a leader in groups.a.34
I can usually figure out how to get an adult to see my point of view, even if they
don’t agree with me.b
.33
If I want to improve a problem at my school, I know how to gather useful data
about the issue.b
.74
I know how school rules and policies are made at my school.b.57
If I want to improve a problem in my city, I know how to gather useful data about
the issue.b
.68
If I want to improve a problem in my city, I can work effectively with other
students on this issue.b
.61
I know how city rules and policies are made.b.63
Motivation to influence
It is important for youth to try to improve our city even if we can’t always make
the changes we want.b
.73
I want to have as much say as possible in making decisions in my city.b.72
I want to have as much say as possible in making decisions in my school.b.69
Students should work to improve our school even if we can’t always make the
changes we want.b
.68
Participatory behavior
I have led a group of young people working on an issue we care about.b.59
I have made a presentation to a group of people I don’t know.b.58
I have spoken with adults in my school about issues that I want to improve at
the school.b
.74
I have interviewed an adult to learn their perspectives about an issue.b.45
I have spoken with other students about issues that I want to improve at the school.b.72
If issues come up that affect students at my school, we do something about it.b.69
If issues come up that affect youth in my city, we do something about it.b.54
I have spoken with other youth about issues that I want to improve in the city.b.57
Perceived control
There is a student council here that gets to decide on some really important things.c.69
There are plenty of ways for students like me to have a say in what our school does.b.67
Students have a say in what happens at this school.b.64
Students at this school get to help plan special activities and events.c.62
There are plenty of ways for young people like me to have a say in what our city
government does.b
.62
Youth have a say in what happens in this city.b.53
a. Items from original Zimmerman, 1991 measure.
b. Items generated by study authors.
c. Items from DSC, 2000, school autonomy measure.
at UNIV CALIFORNIA BERKELEY LIB on April 10, 2012heb.sagepub.comDownloaded from

354 Health Education & Behavior 38(4)
one of the psychological empowerment subscales was signifi-
cantly (and weakly) associated with adolescents’ academic
achievement when measured by their overall GPA, suggesting
that academic achievement is a distinctive domain from psy-
chological empowerment for these urban youth. A different
pattern was found for standardized test scores, however, in
which adolescents with higher scores in English reported mod-
estly lower levels of sociopolitical skills and perceived control
at school; those with higher scores in Math reported modestly
lower sociopolitical skills, perceived control, and participatory
behavior (all significant correlations in the .1 to .3 range). There
was no linear relationship found between adolescents’ standard-
ized test scores and their report of their motivation to influence
their school and community settings. Because the larger schools
in our sample had higher aggregate standardized achievement
scores, it may be that students in the larger schools had higher
achievement scores but reported lower levels of psychological
empowerment and participatory behavior because of being in
very large schools in which there were fewer opportunities
for influence and participation. Being in a larger school set-
ting, however, did not affect their motivation to exert an impact
on their school or community.
Several methodological limitations of the study should be
considered. First, our results here should be replicated with
an independent sample for more robust confirmation of the
factor structure of the psychological empowerment measure.
Second, we note that several of the original items tested in
this study were eliminated because of low response variability;
it is possible that these items were too socially desirable, an
issue to consider in assessing self-report of psychological
empowerment. Furthermore, because all of the psychosocial
dimensions studied here (excluding academic performance)
were assessed via adolescents’ self-report, it is possible that
shared method variance may have increased the associations
among these variables. Our results that demonstrated major
variation among the strength of the correlations found among
these self-reported psychosocial variables, however, suggest
that there are other sources of substantive variability at play.
In light of the highly perceptual nature of psychological
empowerment and the other psychosocial dimensions studied
here, adolescents themselves would be considered to be the
most valid reporters of our variables of interest. A final meth-
odological limitation of this initial examination of the relation-
ship among psychological empowerment and other psychosocial
and academic indicators is that the cross-sectional design of
the present analysis cannot provide evidence regarding the
causal direction of these relationships. Future research using
a longitudinal design should be used to investigate multiple
key questions including how psychological empowerment var-
ies over time, and the extent to which changes in psychological
empowerment may precede or follow changes in other aspects
of adolescent functioning.
A final area where this research should be fruitfully
extended concerns the ecological framing of the psychological
empowerment of young people within schools as organiza-
tional settings (Hughey, Peterson, Lowe, & Oprescu, 2008;
Peterson & Zimmerman, 2004). The current measure assesses
young people’s sense of perceived control in school as well
as their motivation to influence decisions within their schools.
We also found that students with higher levels of psychologi-
cal empowerment report a stronger sense of social support
from other students within their school. We did not, however,
explicitly consider how students’ psychological empowerment
relates to overall levels of organizational empowerment or
a broader sense of community in their schools or after-school
programs. Much prior theory and research on U.S. public
schools emphasizes the limited power of students over their
own learning as well as over decisions and policies that shape
their school experiences (Sarason, 1996). This problem is even
more striking in secondary schools, when adolescents with a
growing capacity and desire for autonomy enter “mismatched”
school environments, with few opportunities for them to
exercise control (Eccles, Midgley, Wigfield, & Buchanan,
1993; Midgley & Feldlaufer, 1987). Systematic measurement
of the construct of psychological empowerment is thus par-
ticularly important for the study of adolescent development
and its relationship to school settings.
Implications for Practitioners
The scales developed in the present study are well suited
for application across a broad range of youth development
programs—including youth-led research or organizing inter-
ventions that involve the youth choosing their own issues to
address—because they assess the kinds of knowledge, skills,
and actions typically involved in the process of youth empow-
erment but are not linked to any one specific issue or goal. It is
Table 3. Means and Standard Deviations for Empowerment
Subscales and Other Key Variables
Scale (1-4 response) Mean
Standard
Deviation n
Sociopolitical skills–generala2.31 .56 435
Motivation to influencea2.64 .69 432
Participatory behaviora2.02 .64 436
Perceived controla2.45 .62 401
PAR self-efficacya2.15 .76 434
Self-esteema2.91 .54 438
GPA (0-4 scale) 2.96 .79 332
Caring relationshipsa2.71 .81 437
Social supporta2.11 .73 438
English language test (0-600 scale) 374.5 69.7 302
Math test (0-600 scale) 341.1 89.5 298
Note: GPA = grade point average.
a. Response scale: 1-4.
at UNIV CALIFORNIA BERKELEY LIB on April 10, 2012heb.sagepub.comDownloaded from

Ozer and Schotland 355
possible that practitioners will benefit from supplementing
the existing measures in order to assess any specific areas of
skills, self-efficacy, or group efficacy in the domains most
relevant to their own programs. For example, an organization
focused on environmental justice efforts would likely want to
assess specific areas of research and advocacy concerning the
regular activities of young people engaged in these kinds of
efforts. Evaluators of community- rather than school-based
efforts would likely need to adapt the measures to focus on
the community organization rather than schools per se. This
measure of psychological empowerment can easily be used
in conjunction with qualitative assessments to enable a mul-
timethod assessment of the construct; for example, interview-
ing and observations can help to provide further evidence of
the specific ways in which young people demonstrate socio-
political or civic engagement skills. In sum, these measures
contribute to the study of empowerment-oriented interventions
for youth, an area of much programmatic growth in which
more systematic evaluation is needed.
Acknowledgments
The authors express appreciation to Marc Zimmerman, Thomas Cook,
Meredith Minkler, Lawrence Green, Edward Seidman, and Dawn With-
erspoon for their consultation; Elizabeth Hubbard, Brian Stanley, Gary
Cruz, Adee Horn, Morgan Wallace, LaWanda Mohammed, Lauren
Moret, Christopher Pepper, Luis Mashek, Jodi Cohen, Minh Luc, Robert
Clothier, Dina Wright, Othello Jackson, Youth in Focus, the SF Unified
School District and participating school sites, and the SF Boys and Girls
Club for their collaboration; and Laura Douglas, Maggie Wanis, Miranda
Ritterman, Eddy Jara, Linda Chhoa, Teresa Igaz, Sami Newlan, Rachana
Modi, Issac Acosta, Eric Koo, Becky Lee, Chris Wu, Kathryn Clark,
Stacy Dao, Carlos Molina III, Christina Law, Gilbert Parada, and Heather
Hochrein for their research assistance.
Declaration of Conflicting Interests
The authors declared no conflicts of interests with respect to the
authorship and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for
the research and/or authorship of this article:
This research was supported by a William T. Grant Scholars’
Award to the first author, and by funding from the Centers for Disease
Control and Prevention Center as part of the Center of Excellence
for Youth Violence Prevention (Center for Culture, Immigration and
Youth Violence Prevention at UC-Berkeley).
References
Bentler, P. M. (2004). EQS 6 Structural Equation Program Manual.
Encino, CA: Multivariate Software.
Cammarota, J., & Fine, M. (Eds.). (2007). Revolutionizing education:
Youth participatory action research. New York, NY: Routledge.
Cargo, M., Grams, G., Ottoson, J., Ward, P., & Green, L. (2003).
Empowerment as fostering positive youth development and
citizenship. American Journal of Health Behavior, 27(Suppl. 1),
S66-S79.
Constantine, N. A., & Bernard, B. (2001). California Healthy Kids
Survey Resilience Assessment Module (Technical Report).
Berkeley, CA: Public Health Institute.
Developmental Studies Center. (2000). Middle School Student
Questionnaire Measures. Oakland, CA: Author.
DuBois, D. L., Felner, R. D., Brand, S., Phillips, R. S. C., & Lease,
A. M. (1996). Early adolescent self-esteem: A developmen-
tal ecological framework and assessment strategy. Journal of
Research on Adolescence, 6(4), 543-579.
Eccles, J. S., Midgley, C., Wigfield, A., & Buchanan, C. M. (1993).
Development during adolescence: The impact of stage-environment
fit on young adolescents’ experiences in schools and in families.
American Psychologist, 48, 90-101.
Fredricks, J. A., & Eccles, J. S. (2008). Participation in extracurricu-
lar activities in the middle school years: Are there developmental
benefits for African American and European American youth?
Journal of Youth and Adolescence, 37, 1029-1043.
Grob, A., Little, D. L., Wanner, B., Wearing, A. J., & Euronet.
(1996). Adolescents’ well-being and perceived control across 14
Table 4. Correlation Statistics for Empowerment Subscales and Other Key Variables
Factor/Scale 1 2 3 4 5 6 7 8 9 10 11
Sociopolitical skills (1) 1
Motivation to influence (2) .60** 1
Participatory behavior (3) .64** .62** 1
Perceived control (4) .66** .64** .59** 1
PAR self-efficacy (5) .72** .67** .72** .61** 1
Self-esteem (6) .36** .34** .36** .32** .42** 1
GPA (7) –.13* .07 –.05 .04 .01 –.12* 1
Caring relationships (8) .54** .57** .52** .58** .55** .32** .09 1
Social support (9) .51** .45** .52** .51** .48** .17** .10 .47** 1
English language test (10) –.22** .02 –.04 –.12** –.02 –.22* .55** .03 .09 1
Math test (11) –.29** –.08 –.18* –.20* –.17** –.25** .46** –.12** .04 .70** 1
*p > .05. **p > .01.
at UNIV CALIFORNIA BERKELEY LIB on April 10, 2012heb.sagepub.comDownloaded from

356 Health Education & Behavior 38(4)
sociocultural contexts. Journal of Personality and Social Psychol-
ogy, 71, 785-795.
Holden, D., Evans, W. D., Hinnant, L., & Messeri, P. (2005).
Modeling psychological empowerment among youth involved
in local tobacco control efforts. Health Education & Behavior, 32,
264-278.
Holden, D. J., Crankshaw, E., Nimsch, C., Hinnant, L. W., & Hund, L.
(2004). Quantifying the impact of participation in local tobacco con-
trol groups on the psychological empowerment of involved youth.
Health Education & Behavior, 31, 615-628.
Hoyle, R. H. (1995). Structural equation modeling. Thousand Oaks,
CA: Sage.
Hu, L.-T., & Bentler, P. (1995). Evaluating model fit. In R. H. Hoyle
(Ed.), Structural equation modeling: Concepts, issues, and
applications (pp. 76-99). London, UK: Sage.
Hughey, J., Peterson, N. A., Lowe, J. B., & Oprescu, F. (2008).
Empowerment and sense of community: Clarifying their relation-
ship in community organizations. Health Education & Behavior,
35, 651-663.
Kline, R. B. (1998). Principles and practice of structural equation
modeling. New York, NY: Guilford.
Mahoney, J. L., Cairns, B. D., & Farmer, T. W. (2003). Promoting
interpersonal competence and educational success through extra-
curricular activity participation. Journal of Educational Psychology,
95, 409-418.
Marr-Lyon, L., Young, K., & Quintero, G. (2008). An evaluation of
youth empowerment tobacco prevention programs in the south-
west. Journal of Drug Education, 38, 39-53.
Midgley, C., & Feldlaufer, H. (1987). Students’ and teachers’ decision-
making fit before and after the transition to junior high school.
Journal of Early Adolescence, 7, 225-241.
Nieto, S. (1996). Lessons from students on creating a chance to
dream. Harvard Educational Review, 64, 392-426.
Peterson, N. A., & Zimmerman, M. A. (2004). Beyond the individual:
Toward a nomological network of organizational empowerment.
American Journal of Community Psychology, 34, 129-145.
Rappaport, J. (1987). Terms of empowerment/exemplars of pre-
vention: Toward a theory for community psychology. American
Journal of Community Psychology, 15(2), 121-148.
Rappaport, J., Swift, C., & Hess, R. (1984). Studies in empower-
ment: Steps toward understanding and action. New York, NY:
Haworth.
Sarason, S. B. (1996). Revisiting “The Culture of the School and the
Problem of Change.” New York, NY: Teachers College Press.
Shor, I. (1996). When students have the power: Negotiating authority
in critical pedagogy. Chicago, IL: University of Chicago Press.
Wallerstein, N. (1999). Power between evaluator and community:
Research relationships within New Mexico’s healthier communi-
ties. Social Science & Medicine, 49, 39-53.
Wills, T. A. (1994). Self-esteem and perceived control in adolescent
substance use: Comparative tests in concurrent and prospective
analyses. Psychology of Addictive Behavior, 8, 223-234.
Wilson, N., Minkler, M., Dasho, S., Carrillo, R., Wallerstein, N., &
Garcia, D. (2006). Training students as partners in community
based participatory research: The Youth Empowerment Strate-
gies (YES!) Project. Journal of Community Practice, 14, 1-2.
Zimmerman, M., & Zahniser, J. (1991). Refinements of sphere-specific
measures of perceived control: Development of a sociopolitical
control scale. Journal of Community Psychology, 19, 189-204.
Zimmerman, M. A. (1995). Psychological empowerment: Issues and
illustrations. American Journal of Community Psychology, 23,
581-599.
Zimmerman, M. A. (2000). Empowerment theory: Psychological,
organizational, and community levels of analysis. In J. Rappaport
& E. Seidman (Eds.), Handbook of community psychology
(pp. 43-64). New York, NY: Kluwer Academic/Plenum.
at UNIV CALIFORNIA BERKELEY LIB on April 10, 2012heb.sagepub.comDownloaded from