Journal of Early Adolescence
© The Author(s) 2015
Reprints and permissions:
Self-Esteem and Future
Danielle M. Jackman1 and David MacPhee1
This study’s purpose was to examine the relations among future orientation,
self-esteem, and later adolescent risk behaviors, and to compare two
mediational models involving self-esteem versus future orientation as
mediators. An ethnically diverse sample of 12- to 14-year-olds (N = 862,
54% female, 53% ethnic minority) was assessed longitudinally. Correlations
supported the hypotheses that future orientation and self-esteem are (a)
positively correlated with each other (r > .40) and (b) negatively related
to various risky behaviors that were composited into a measure of risk
orientation. Controlling for social desirability, self-esteem partially mediated
the relation between future orientation and later risk orientation, and future
orientation partially mediated the relation between self-esteem and risk
orientation. The results shed light on how the concept of possible selves
might factor into adolescents’ risk avoidance. In particular, both future
orientation and healthy self-esteem may serve as protective factors against
engaging in risky behavior.
adolescence, future orientation, risk behaviors, self-esteem
1Colorado State University, Fort Collins, USA
David MacPhee, Department of Human Development & Family Studies, Colorado State
University, Fort Collins, CO 80523, USA.
602756JEAXXX10.1177/0272431615602756Journal of Early AdolescenceJackman and MacPhee
2 Journal of Early Adolescence
It is well established that adolescents are at the apex of engaging in risky
behaviors (Gullone & Moore, 2000; Steinberg, 2004). Such risky behaviors
include unsafe sex, delinquency, and impulsivity (Robbins & Bryan, 2004;
Rooke & Hine, 2011), many of which co-occur (Coleman & Cater, 2005).
Researchers, interventionists, and policy makers are especially interested in
protective factors that mitigate adolescent risk taking because such insights
contribute to effective prevention efforts (Biglan, Brennan, Foster, & Holder,
2004). In the current study, we focus on two theoretically important protec-
tive factors that seldom have been studied in tandem and over time: adoles-
cents’ future orientation and self-esteem.
Problem behavior theory (Jessor et al., 2003) asserts that youth who have
positive expectations for success or an optimistic outlook about their future
are less likely to engage in problem behaviors. Other theorists (Nurmi, 1991)
and a handful of studies (e.g., Chen & Vazsonyi, 2011, 2013; Robbins &
Bryan, 2004) concur that adolescents who are high in future orientation are at
lower risk for delinquency, drug use, or school problems. Similarly, studies of
self-esteem have also shown a complex but mostly inverse relation with ado-
lescent risk taking (e.g., Salazar et al., 2005). The underlying developmental
process contributing to both future orientation and self-esteem is identity
Identity development is a key developmental process during the transition
from childhood to adolescence. Identity development facilitates self-
evaluation and positive psychological well-being (Crocetti, Rubini, Luyckx,
& Meeus, 2008; Luyckx, Goossens, Soenens, Beyers, & Vansteenkiste, 2005)
as well as social relationships (Zimmer-Gembeck & Petherick, 2006). Having
a clear sense of self is inversely related to depression and anxiety as well as
externalizing behavior problems (Schwartz, 2007). These individuals tend to
have positive thoughts about their future (Nurmi, 1991; Seginer, 2009) as
well as exude more confidence (Harter, 1999; Schwartz, Mason, Pantin, &
Szapocznik, 2008; Steinberg et al., 2009). Theoretical and empirical work
also finds that adolescents who think critically about their future selves as
well as have a positive sense of self are more inclined to explore various
aspects of their identity and fare better than those who do not (Harter, 1999;
Hoyle & Sherrill, 2006).
While neo-Eriksonian theorists were developing complementary identity
models (Crocetti et al., 2008; Luyckx et al., 2005; Schwartz, 2007), a focus
of empirical research has been on the timing of identity development.
Previous research focused primarily on identity development among adoles-
cents (Kroger, Martinussen, & Marcia, 2010). However, recent longitudinal
studies have highlighted a progression in identity development, beginning in
early adolescence with diffusion, which extends into emerging adulthood
Jackman and MacPhee 3
(Meeus, 2011). In addition to this normative progression as well as relative
stability of identity statuses across time, individual differences in identity
development are associated with indicators of adolescent adjustment; for
example, diffusion more often is correlated with higher levels of behavioral
and adjustment problems (Meeus, 2011).
In addition, studies have shown self-esteem to be positively related to
future orientation (Nurmi & Pulliainen, 1991). As adolescents develop their
future orientation in the self-system, attitudes and beliefs about the self
become central to behavioral choices that they make (Harter, 1999; Marcia,
1980). However, there has been a dearth of longitudinal evidence related to
whether future orientation and self-esteem are potential mediators with
regard to adolescent risk engagement. In the sections that follow, future ori-
entation is examined as a possible explanation for the relation between self-
esteem and adolescent risk. This putative mediational pathway will then be
compared with the potential role of self-esteem as a mediator of the associa-
tion between future orientation and adolescent risky behavior.
Future Orientation and Adolescents’ Risky
Broadly speaking, one’s future orientation is consciously constructed within a
social context and represents images of how the self is projected in the future
(Seginer, 2009). These images of the self involve expectations about and actions
related to the future (Nurmi, 1991). This aspect of self-perception is related to
other constructs such as future time orientation or temporal perspective (i.e., the
importance attached to the past, present, and future; Gjesme, 1979), temporal
extension (i.e., the time horizon an individual projects into life in the future;
Lessing, 1972), and possible selves (i.e., elements of the self-concept that repre-
sent individuals’ anxieties and fears, goals and motives; Markus & Nurius,
1986; Oyserman, Bybee, & Terry, 2006). The constituent processes of future
orientation may involve understanding the level of control over future endeav-
ors, how optimistic or pessimistic are future thoughts, and thinking about inter-
ests and goals for the future (motivation), mapping out ways to execute these
interests and goals (planning), and assessing a realistic time frame for executing
these interests and goals (evaluating; Nurmi, 1991; Trommsdorff & Lamm,
1980). Nurmi (1991) has argued that adolescents, as part of their identity devel-
opment, are especially likely to orient to life trajectories and engage in specific
actions that are informed by their future orientation.
Growing empirical evidence links future orientation to behavioral outcomes in
adolescence. Engagement in risk behaviors may presume that future consequences
are not taken into consideration or are discounted to satisfy immediate needs and
4 Journal of Early Adolescence
desires (Gottfredson & Hirschi, 1990), and previous studies have shown strong,
inverse relations between positive future orientation and engagement in risky
behaviors (e.g., Chen & Vazsonyi, 2011; Harris, Duncan, & Boisjoly, 2002). For
example, inner-city African American youth who have a bleak view of their
future are more likely to engage in risky sexual behaviors such as unprotected
sex and multiple sex partners (Bolland, 2003). This relation is heightened for
younger adolescents compared with older adolescents (Steinberg et al., 2009).
The neuroscience literature indicates that delayed development of the pre-
frontal cortex—which is responsible for self-evaluation, planning, and self-
regulation—in part accounts for adolescents’ difficulties with controlling
impulses and delaying rewards (Steinberg et al., 2009). Thus, an association
between low levels of future orientation and negative behavioral outcomes is
more often observed among early adolescents compared with late adoles-
cents or emerging adults (Steinberg et al., 2009). Steinberg and colleagues
(2009) assessed a diverse sample of adolescents and found that early adoles-
cents (10-15 years old) more often possess poorer future orientation and self-
regulatory skills compared with older adolescents. Similarly, a recent
cross-sectional study found that the ability to delay gratification and avoid
risky behaviors was weakest among early compared with late adolescents
(Romer, Duckworth, Sznitman, & Park, 2010).
Conversely, a diverse sample of adjudicated early and middle adolescents
with a more positive future orientation was found to be less impulsive and to
engage in fewer risky behaviors (Robbins & Bryan, 2004). Even though there
are age differences in future orientation percepts and risk behaviors, these stud-
ies highlight a negative relation between optimistic thoughts about the future
and engaging in risky behaviors. This suggests that a positive view of the future
may function as a protective factor in adolescence, regardless of age.
Although research generally finds an inverse relation between future ori-
entation and risky behaviors, there are several important limitations with the
extant literature. First, there is some evidence that future orientation fails to
distinguish those who engage in more risky behaviors from those who do not
(Trommsdorff & Lamm, 1980). An early study that assessed future orienta-
tion among self-described delinquents versus nondelinquents showed no dif-
ference between the groups in their opinions about the future (Losel, 1975).
A second limitation noted by several researchers (e.g., Chen & Vazsonyi,
2011) is that few longitudinal studies have been conducted on the relation
between future orientation and risky behaviors. Considering that future orien-
tation deals with thoughts about the future, it would be informative to con-
duct longitudinal studies to assess how well risk taking is predicted by
long-term goals and sense of self. Finally, most research in this area has
focused on a single domain of problem behavior; the current study included
multiple risk-taking attitudes and behaviors as outcomes.
Jackman and MacPhee 5
Future orientation also is related to affective components of individuals’
perceptions of themselves (Nurmi, 1991; Pulkkinen & Rönkä, 1994). For
instance, poor self-perceptions in conjunction with low perception of control
about the future have been shown to increase pessimistic views of ability and
identity, which in turn can increase the likelihood of risk engagement (Harter,
1999; Wills, 1994). Thus, one would expect future orientation and self-esteem
to covary in relation to adolescents’ risk taking.
Self-Esteem and Risky Behaviors
Self-esteem is an important heuristic concept for the study of personality and
behavior. It is an assessment of one’s worth that is a component of self-
schema (Harter, 1999; Marsh, 1986; Rosenberg, Schooler, Schoenbach, &
Rosenberg, 1995) and can be described in two ways: global and domain-
specific (situational; McGee & Williams, 2000; Rosenberg et al., 1995).
Global self-esteem deals with one’s assessment of self-worth in its totality
whereas domain-specific self-esteem applies more to the assessment of self-
worth in a certain context (Harter, 1999). A person who displays high self-
esteem tends to exude more confidence and also values the self much more
than someone with low self-esteem (Rosenberg et al., 1995). However,
diverse theories have stated that early adolescents’ self-esteem tends to be
fragile given that it is more susceptible to social comparison processes during
this developmental stage (Dunkel, 2000; Harter & Whitesell, 2003).
In later adolescence, self-esteem tends to gradually increase and becomes
more positive as freedom, personal authority, and role-taking ability increase,
and more opportunities to behave in socially appropriate ways are available
(Harter & Whitesell, 2003). However, this is not always the case: Some stud-
ies have questioned the assumption that high self-esteem inversely correlates
with engagement in risks (Baumeister, Heatherton, & Tice, 1993; Baumeister,
Smart, & Boden, 1996). That is, although studies have shown high self-
esteem to be related to low levels of risk engagement (e.g., Donnellan,
Trzesniewski, Robins, Moffitt, & Caspi, 2005; McGee & Williams, 2000),
other authors assert that threatened egotism results in inflated views of the
self, which can then lead to risky behaviors (Crocker & Park, 2004). This
argument is based on the assumption that people sometimes engage in risky
behaviors to increase or maintain self-esteem, or to reduce the threat of hav-
ing low self-esteem (see Baumeister et al., 1993; Crocker & Park, 2004).
Although the current study does not assess threat to adolescents’ egos, if there
is a positive relation between self-esteem and risky behavioral engagement,
one might assume that there is a possibility of threatened egotism.
Extant research provides a conflicting picture that calls into question the
magnitude of influence self-esteem has on adolescents’ engagement in risky
6 Journal of Early Adolescence
behavior (Donnellan et al., 2005; McGee & Williams, 2000; Wills, 1994).
Even when self-esteem is correlated with adolescents’ risky behavior, the
effect sizes are small (e.g., Salazar et al., 2005). One possible reason for the
weak or contradictory findings is that most studies are cross-sectional
(Donnellan et al., 2005; McGee & Williams, 2000; Rosenberg, Schooler, &
Schoenbach, 1989), making it difficult to parse the direction of effects
between self-esteem and adolescent problem behaviors.
The few longitudinal studies that have been conducted support the conclu-
sion that self-esteem predicts risky behavior. For example, Donnellan et al.
(2005) conducted three cross-sectional and longitudinal studies with ethni-
cally diverse adolescents that assessed the relation between global self-
esteem and externalizing behaviors. Overall, low self-esteem was strongly
related to externalizing behaviors. In another longitudinal study of eighth
graders that assessed self-attitudes, including self-esteem and perceived con-
trol, Wills (1994) found an inverse relation between positive self-attitudes
and substance use among a group of diverse early adolescents. Similar results
were also obtained by McGee and Williams (2000), who evaluated the asso-
ciations among global and domain-specific self-esteem and a multitude of
health-compromising behaviors including early sexual activity, substance use,
and suicidal ideation. The results showed an inverse relation between global
self-esteem and health-compromising behaviors. Academic self-esteem, how-
ever, did not significantly relate to any of the health-compromising behaviors.
Thus, longitudinal studies, in contrast to those that are cross-sectional, more
often find an inverse relation between adolescents’ self-esteem and risk
Self-Esteem and Future Orientation as Mediating
Despite the fact that theoretical perspectives suggest that self-esteem is related
to future orientation (Nurmi, 1991; Steinberg, 2004; Steinberg et al., 2009),
there is a paucity of empirical research that focuses specifically on this asso-
ciation. For instance, according to self-enhancement theory, individuals seek
to maximize their self-esteem by viewing themselves in favorable ways,
increasing or maintaining their feelings of self-worth, increasing activities that
improve their competence, and remaining consistently positive about them-
selves (Epstein, 1973). An inference from this theory is that self-enhancement
is promoted when individuals think positively about their future.
Even though there is a dearth of studies assessing the direct link between
future orientation and self-esteem, a study by McFarlin and Blascovich
(1981) found that emerging adults with high self-esteem remained consistent
Jackman and MacPhee 7
in having hopeful expectations about the future even when presented with an
unexpected outcome. Conversely, those with low self-esteem desired opti-
mistic future outcomes but discounted them as improbable. In a later study,
Epel, Bandura, and Zimbardo (1999) documented a strong link between
future perspectives and self-efficacy. Such studies suggest that future orienta-
tion and self-esteem are correlated, although it is not yet clear whether posi-
tive views of the self contribute to future orientation. Also, it is possible that
future orientation directly accounts for the relation between self-esteem and
adolescents’ engagement in risky behaviors.
Intervention studies for adolescents, especially programs targeted at pre-
vention of risky sex or promotion of health or behavioral adjustment, have
incorporated self-perceptual concepts such as self-esteem (DuBois et al.,
2002; Jemmott, Jemmott, Spears, Hewitt, & Cruz-Collins, 1992; Salazar
et al., 2005). The presumption for this focus on self-esteem is that it is caus-
ally related to, or mediates, healthy adolescent behaviors. Outcomes of the
few related intervention studies tend to support this presumption. For
instance, DuBois et al. (2002) found that changes in self-esteem mediated the
effects of social support on both emotional and behavioral adjustment of
early adolescents. In a later study, Salazar et al. (2005) assessed African
American female adolescents’ self-esteem and sexual risk taking. No associa-
tion was found between self-esteem and sexually transmitted diseases (STDs)
or pregnancy, but significant relations were observed between the mediators
(domain-specific self-efficacy) and self-esteem: Those who were higher in
self-esteem were more likely to be high in self-efficacy.
Although self-esteem was not considered a mediating variable, Salazar
et al. (2005) suggested that self-esteem be included as a mediator in future
studies of adolescent health behavior, given that it was strongly related to
other self-perceptions (i.e., situational self-efficacy) that were correlated with
risk taking. Therefore, self-esteem may explain adolescents’ risky behaviors
above and beyond future orientation. However, may it also be that future
orientation serves as a mediator between self-esteem and adolescents’ risky
behaviors? Even though there are few studies focused on the mediating
effects of self-esteem on future orientation and risky behavioral engagement,
one can extrapolate from the above studies that future orientation could medi-
ate the association between self-esteem and risky behaviors.
The Current Study
This study focused on two potential developmental cascades: whether future
orientation mediates the association between self-esteem and adolescents’
risky behaviors, or whether self-esteem accounts for the relation between
8 Journal of Early Adolescence
future orientation and adolescents’ risky behaviors. Three hypotheses related
to bivariate associations among the study variables were tested:
Hypothesis 1: There is an inverse relation between future orientation and
adolescents’ risky behaviors.
Hypothesis 2: There is an inverse relation between self-esteem and ado-
lescents’ risky behaviors.
Hypothesis 3: A more optimistic sense of future among young adoles-
cents is associated with higher self-esteem.
Two mediational models were also compared. Given the limited literature,
exploratory hypotheses were tested:
Hypothesis 4: Higher levels of self-esteem mediate the inverse relation
between future orientation and later engagement in risk orientation.
Hypothesis 5: High levels of future orientation mediate the inverse rela-
tion between self-esteem and adolescents’ risk orientation.
Participants (N = 887) included youth ages 12 to 14 (M = 13.23) who were
recruited, along with at least one family member, into an intervention
designed to delay sexual debut. Families were recruited from seven sites: two
ethnically diverse cities and two rural counties in Colorado, Miami (FL),
rural Maryland, and the Eastern Shore of Virginia. Recruitment was by means
of flyers sent home from middle schools; presentations at family nights and
youth organizations; referrals from teachers, community service providers,
and churches; and word of mouth. Families were randomized into the inter-
vention and control groups after the baseline survey was administered. This
study focuses on the pooled sample given that no intervention effects were
detected on any of the outcome measures.
Demographic data collected at baseline showed that in the initial sample,
46.7% self-identified as non-Hispanic White, 21.3% as Hispanic, 26.4% as
Black, and 5.6% other ethnic groups. Approximately, 27% of families lived
below the poverty line. The parents’ average age was 40 years, and typically,
they were high school educated (M = 13 years). In terms of family structure,
35.8% were single or divorced and 48.4% were married or remarried, with
the remainder cohabiting or separated. Site-specific demographic informa-
tion is available in an online supplement.
Jackman and MacPhee 9
Three validity questions were included at the conclusion of the surveys,
each of which was rated from 1 (all) to 4 (hardly any): “I understood the
questions in this survey,” “I answered the questions carefully,” and “I
answered the questions honestly.” Twenty-five youth were omitted from the
sample because their average on the three validity questions exceeded 3.00
(some), which resulted in a combined sample of N = 862 youth in this study
Future orientation. The eight-item Goals and Aspirations scale from the Healthy
Kids Resilience Assessment (HKRA; Constantine, Benard, & Diaz, 1999) was
used to assess teens’ future orientation. This scale included items pertaining to
how optimistic or pessimistic teens viewed their future as well as their views on
the likelihood that they will become parents before age 19, graduate from high
school and college, get married, and have a good job. A sample item included
“I have goals and plans for the future,” which was rated from 1 (not at all true)
to 4 (very much true). Internal reliability for the original scale was sound, α =
.84 (Constantine et al., 1999), but in the current sample, the reliability was
marginal, α = .64. Data from a large, multiethnic sample of youth in elementary
and middle school found a three-item version of the scale to be correlated with
risk taking (inversely) as well as school achievement (Hanson & Kim, 2007).
However, this study of the HKRA also found the internal consistency of the
Goals and Aspirations scale to be marginal.
Self-esteem. The Rosenberg Self-Esteem Scale (RSES; Rosenberg, 1965) is a
10-item measure of general self-worth and self-acceptance. Sample items
included “I feel that I have a number of good qualities” and “I certainly feel
useless at times,” which are rated from 1 (strongly disagree) to 4 (strongly
agree). Multiple studies of adolescents, including those in middle school
(Choi, Meininger, & Roberts, 2006; Hagborg, 1993), found the RSES to have
high internal consistency, α = .88-.92 (α = .76 in the current sample), and test-
retest reliability, r = .82-.88 (see Fischer & Corcoran, 1994). The RSES’s
validity has been established in terms of correlations with depression, anxiety,
and peer group reputation as well as other self-esteem measures such as the
Harter Self-Perception Profile (Hagborg, 1993) and general self-regard
(Fischer & Corcoran, 1994). Some studies find that minority adolescents
report lower RSES scores than do non-Hispanic White teen (Choi et al., 2006).
Perceptions of risk. It was measured using a subset of the 22-item Adolescent
Risk Questionnaire (ARQ; Gullone, Moore, Moss, & Boyd, 2000). This scale
10 Journal of Early Adolescence
was used to assess the decision-making component to engage in risk taking
such as thrill seeking (driving 20 mph over the speed limit), rebellious behav-
iors (e.g., underage drinking), and reckless behaviors (e.g., having unpro-
tected sex). These items were rated from 0 (not at all risky) to 4 (extremely
risky) in response to questions such as “How risky is it to . . . smoke mari-
juana once or twice a month?” Thus, the higher the score, the less likely the
youth would be to engage in the risky behavior. The ARQ was developed and
validated with large, representative samples of 11- to 18-year-olds. Both test-
retest (.44-.80) and alpha reliabilities (α = .88, in this sample) are adequate.
The ARQ was validated with confirmatory factor analysis and inverse cor-
relations between perceived risk and actual risk-taking behaviors.
Impulsivity. Impulsivity/sensation seeking was measured using an 11-item
scale (Cahalan & Room, 1974) that included items such as “I often act on the
spur of the moment without stopping to think” and “I get a real kick out of
doing things that are a little dangerous,” which are rated from 1 (not at all) to
4 (quite a lot). The scale has a high internal consistency of α = .86 (Schafer,
Blanchard, & Fals-Stewart, 1994; α = .84, in the current sample) and is highly
correlated with risky behavior such as substance use (Trocki, Drabble, &
Sexual attitudes. Sexual attitudes were measured using items from the Sexual
Risk Behavior Beliefs and Self-Efficacy Scales (SRBBS; Robinson, Shaver,
& Wrightsman, 1991). Sample items include “People should not have sex
before marriage” and “It’s okay for two people to have sex before marriage if
they are in love” (reverse scored), with items rated from 1 (strongly disagree)
to 6 (strongly agree). Thus, teens with high scores are less likely to endorse
risky sexual behavior. Construct validity of the SRBBS was documented with
a large, multiethnic sample of teens through confirmatory factor analysis,
comparisons showing that virgins and younger teens (age 14) have more con-
servative social norms and attitudes toward sexual intercourse than nonvirgin
and older teens, and correlations with sexual self-efficacy (Basen-Engquist
et al., 1999). Internal consistency was high in the current sample, α = .86.
Risky sexual behavior. Four items from the Scale of Sexual Risk-Taking (Met-
zler, Noell, & Biglan, 1992) were used to assess whether teen participants
had ever had intercourse or had in the last 6 months and whether they engaged
in the following risky sexual behaviors: multiple partners, casual partners, or
failure to use condoms or contraceptives. Items on this scale, which are
weighted by risk level, are interrelated consistently across samples (α = .75-
.90), and scale scores are correlated with measures of peer deviance as well
Jackman and MacPhee 11
as other problem behaviors (Capaldi & Patterson, 1989). This scale has been
used with ethnically diverse samples of teens as young as age 13 (Luster &
Small, 1994; Metzler et al., 1992). A log transformation was performed
because the scale score was highly positively skewed; for example, 13.2% of
the teens reported being sexually active.
Peer delinquency. Participants completed a measure of the frequency with
which their friends engaged in deviant behaviors (Capaldi & Patterson,
1989). A sample item includes “During the last year, how many of your
friends have ruined or damaged something on purpose that wasn’t theirs?”
On this 12-item scale (α = .86), respondents rated the extent to which the
teen’s peer group engaged in activities such as drinking, using drugs, skip-
ping class, delinquency, and sex. The mean score of 2 in this sample indicates
that 25% of their friends engage in such activities. In a longitudinal study of
youth from ages 8 to 16, high scale scores were related to measures of delin-
quent and conduct behaviors as well as family conflict and parents’ antisocial
behavior (Fergusson & Horwood, 1999). An abbreviated version of this scale
has been used in a longitudinal study of ethnically diverse middle schoolers
and found to have acceptable reliability as well as factorial and concurrent
validity (Dishion, Kim, Stormshak, & O’Neill, 2014).
Social desirability. The 13-item short form of the Marlowe-Crowne Social
Desirability Scale (MCSDS), developed by Reynolds (1982), was included
because some adolescents’ responses to questions about sensitive topics such
as sex and peer deviance may be affected by biased self-presentation. This
short form has adequate reliability (α > .76) and construct validity (Robinson
et al., 1991). The MCSDS is correlated with various personality traits indica-
tive of high psychological adjustment, such as emotional intelligence, consci-
entiousness, and self-esteem (see Mesmer-Magnus, Viswesvaran, Deshpande,
& Joseph, 2006), which suggests that the MCSDS may be inversely related to
measures of risk taking in the current study.
Surveys were administered by trained data collectors who were not involved
with providing intervention services. All surveys were given orally to ensure
that participants understood the questions. In order to maintain confidentiality
and to reduce social desirability, privacy screens were provided, and partici-
pants placed their completed surveys in sealable envelopes before turning them
in. Participation was voluntary, and University Institutional Review Board pro-
cedures were followed related to informed consent and confidentiality.
12 Journal of Early Adolescence
Preliminary Analyses and Plan of Analysis
Preliminary analyses were performed on the various risky behaviors in order
to test for multicollinearity. As shown in Table 1, the different measures of
risky attitudes and behavior were significantly intercorrelated at both Time 1
and Time 2, with the exception of the relation between impulsivity and per-
ceptions of risk. Therefore, an exploratory factor analysis was conducted. All
five risk indicators loaded on a single factor, which was labeled Risk
Orientation. Accordingly, perceptions of risk and sexual attitudes were
reverse scored; all risk variables were then centered and combined into a
single indicator of risk orientation at Time 1 (α = .85) and at Time 2 (α = .90).
Prior to testing the hypotheses, potential intervention effects on self-
esteem, future orientation, and risky behaviors were examined using repeated-
measures ANOVA. No intervention effects were detected nor were there site
or intervention differences in the pattern of correlations among the study
variables, so a pooled sample was used for all analyses.
Analyses were conducted to determine whether findings might be moder-
ated by ethnicity. First, Cronbach’s alphas were computed separately for the
three major ethnic groups in the sample—Black, Hispanic, and non-Hispanic
White. The reliability coefficients did not differ across groups by more than
±.05 from the alphas for the group as a whole. Second, repeated-measures
MANOVAs were performed to test for any moderating effects of gender and
ethnicity; Box’s M test was requested to assess violations of the assumption
of homogeneity of variance. With ethnicity as the moderator, neither the two-
way or three-way interaction involving time (pretest/posttest) or measure was
significant, F < 2.17, and Box’s M test also was nonsignificant, F = 1.63, p =
.062. With gender as the moderator, neither the two-way or three-way inter-
action involving time or measure was significant, F < .25, and Box’s M test
also was nonsignificant, F = 1.46. Therefore, gender and ethnicity were not
included as covariates.
In order to test whether attrition was systematic, participants who were
missing the Risk Orientation variable at Time 2 were compared, on 26 demo-
graphic and Time 1 outcome measures, with participants who remained in the
study. The univariate analyses revealed one significant difference at p = .04—
teens who completed the posttest were 2 months younger than those who did
not—or what would be expected by chance alone. Therefore, a propensity
score to correct for selective attrition was not included as a covariate.
Next, missing data were imputed for youth who did not complete follow-
up measures. We used a multiple imputation strategy recommended by
Schafer and Graham (2002). The fully conditional Markov chain Monte
Carlo (MCMC) method was used to impute missing data for normally
Table 1. Correlations Among Predictors of Adolescent Risky Behaviors.
Variables 1 2 3 4 5 6 7
Time 1 Time 2
1. Future orientation .57** .42** .17** −.19** .32** −.10* −.10* 5.13 (.79) 5.19 (.72)
2. Self-esteem .46** .56** .08* −.11* .13** −.04 −.17** 4.48 (1.01) 4.59 (.93)
3. Perceptions of risk .20** .11* .45** −.22** .31** −.10** −.12** 3.23 (.82) 3.22 (.80)
4. Peer delinquency −.27** −.14** −.20** .62** −.44** .14** .37** 1.94 (.85) 1.93 (.82)
5. Sexual attitudes .35** .15** .34** −.47** .62** −.18** −.36** 4.41 (1.13) 4.47 (1.11)
6. Risky sexual behavior −.10** −.13** −.09* .19** −.21** .57** .10** 1.81 (1.75) 2.29 (1.84)
7. Impulsivity −.14** −.21** −.10** .37** −.38** .08* .62** 3.85 (1.10) 3.76 (1.05)
Note. Correlations among variables at baseline are above the diagonal; correlations among variables at the 6-month follow-up are below the diagonal.
Correlations along the diagonal (in bold) are stability coefficients.
*p < .05. **p < .01.
14 Journal of Early Adolescence
distributed dependent variables. It assumes an iterative approach that fits a
single variable using all other variables in the model as predictors and then
imputes missing data for the single variable being fit. The method continues
for each variable in the model to the maximum number of iterations speci-
fied, which was 20 in the present study.
The baseline measure of future orientation and the 6-month follow-up
measures of self-esteem and risk orientation were used to test the first media-
tional model. For the second mediational model, the baseline measure of self-
esteem was used to predict the 6-month follow-up measures of future
orientation and risk orientation. Social desirability was included as a covari-
ate because MCSDS scores were significantly (p < .05) correlated with self-
esteem at Time 1, r(860) = .23, and Time 2, r(860) = .19, as well as with Risk
Orientation at Time 1, r(860) = −.36, and Time 2, r(860) = −.30. As described
below, the Hayes (2013) MEDIATE macro was used to test the hypothesized
developmental cascades with all variables, including the covariate, included
in the model. As well, longitudinal analyses provide less biased tests of medi-
ation than do cross-sectional data (Maxwell & Cole, 2007). Given the impor-
tance of temporal precedence in testing mediation (MacKinnon, Fairchild, &
Fritz, 2007), the Time 2 variable (either self-esteem or future orientation,
depending upon the model tested) was used as the mediator.
Correlations Among Predictors, Mediators, and Risk Orientation
Correlations were computed among the predictor, mediator, and outcome
variables to determine whether there is an association that could possibly be
mediated. If there is no significant correlation between the predictor and out-
come variables, then the primary condition for mediation is not met (Baron &
Kenny, 1986). Furthermore, the first three hypotheses were tested with cor-
Regarding the first hypothesis, a significant relation was obtained between
future orientation and the two attitudinal measures, perceptions of risk and
sexual attitudes that endorse abstinence (see Table 1). These associations
were observed both at baseline and at the 6-month follow-up, and were small
to medium effect sizes (Cohen, 1988). In addition, future orientation was
significantly associated with peer delinquency at baseline and the 6-month
follow-up, with medium effect sizes for both. The relations between future
orientation and impulsivity as well as with risky sexual behavior were signifi-
cant but small at baseline and follow-up. Future orientation was significantly
correlated with the composite variable, Risk Orientation, r(860) = −.29,
Jackman and MacPhee 15
p < .001 at Time 1, and r(860) = −.35, p < .001 at Time 2. Thus, the results
support the hypothesis of an inverse relation between future orientation and
adolescents’ risky behaviors.
The second hypothesis is that self-esteem is a protective factor in relation
to risky behavior. The correlational results are similar to those involving
future orientation as a predictor (see Table 1), although the effect sizes in
relation to sexual attitudes were small, whereas they were medium effect
sizes for future orientation, and they were larger in relation to impulsivity as
compared with the correlations between future orientation and impulsivity.
Self-esteem was significantly related to Risk Orientation, r(860) = −.18, p <
.001 at Time 1, and r(860) = −.23, p < .001 at Time 2. Thus, the findings sup-
port Hypothesis 2.
A second criterion for mediation is that the putative mediating variable is
significantly correlated with both the predictor variable and the outcome
variables. Thus, correlations were computed between the baseline predictors
and the Time 2 mediators and outcome variables. First, it should be noted that
future orientation and self-esteem were stable between the two time points
(see Table 1), as was the composite Risk Orientation variable, r(860) = .65,
p < .0001. Second, related to Hypothesis 3, the correlation between future
orientation at Time 1 and self-esteem at Time 2 was significant, with a
medium effect size, r(860) = .30, p < .0001. As well, the correlation between
self-esteem at Time 1 and future orientation at Time 2 was significant,
r(860) = .26, p < .0001. The cross-time correlations as well as the within-time
correlations between future orientation and self-esteem (see Table 1), thus,
support Hypothesis 3. Both future orientation and self-esteem at Time 1 were
significantly correlated with the Time 2 risky behavior composite, r(860) =
−.23 and −.10, respectively, p < .01, meeting the second condition for
Mediation by Self-Esteem Versus Future Orientation
The mediational models represent half-longitudinal designs and thus require
controlling for initial estimates of the proposed mediators and dependent
variable (Cole & Maxwell, 2003). Although in testing a half-longitudinal
model stationarity is assumed and cannot be assessed, the extant literature
suggests the potential error possible from making such an assumption is far
less than failing to account for initial estimates of the mediator and dependent
variable (Cole & Maxwell, 2003). In addition, because the mediational mod-
els are half-longitudinal in design, we may only directly assess the signifi-
cance of the indirect path and, therefore, may only assess partial mediation.
16 Journal of Early Adolescence
The Hayes (2013) MEDIATE macro for IBM SPSS Statistics version 22
was used to estimate the mediational effects of two competing mediational
models: (a) causal self-esteem on risky behavior through future orientation
while controlling for social desirability and initial estimates of future orienta-
tion and risky behavior and (b) causal future orientation on risky behavior
through self-esteem while controlling for social desirability and initial esti-
mates of self-esteem and risky behavior. The MEDIATE macro estimates
Monte Carlo 95% confidence intervals (CIs) for each mediational effect,
where the significance of the indirect path is indicated when the confidence
interval does not contain zero (p > .05; Hayes & Preacher, 2014). We imple-
mented 10,000 samples to assess the confidence intervals and significance of
the partial mediation effects. The extant literature suggests that these pro-
cesses increase the precision of estimates of mediating effects and makes no
additional assumptions about the normality of the sample distribution as is
the case with the more traditional Sobel test (Hayes, 2013; MacKinnon,
Lockwood, & Williams, 2004; Williams & MacKinnon, 2008).
Both mediational models were supported by the analyses. That is, future
orientation partially mediated the relation between self-esteem and risky
behavior after controlling for social desirability and initial estimates of future
orientation and risky behaviors, b = −.032, SE(mc) = .041, 95% CI = [.054,
−.011]. An ordinary least squares (OLS) multiple regression found that the
set of predictors, covariates, and mediator explained Radj
2472=. of the vari-
ance in Time 2 risk orientation, F(5, 857) = 155.30, p < .001. As well, self-
esteem partially mediated the relation between future orientation and risky
behavior after controlling for social desirability and Time 1 estimates of self-
esteem and risky behaviors, b = −.028, SE(mc) = .037, 95% CI = [−.048,
−.008]. All variables in the model explained Radj
2471=. of the variance in
Time 2 risk orientation, F(5, 857) = 154.48, p < .001. In tandem, these find-
ings provide initial support for developmental cascades in which future ori-
entation influences risky behavior partially through its impact on youth
self-esteem, and self-esteem contributes to later risk orientation partially
through its influence on future orientation.
The purpose of this study was to assess the relations among future orienta-
tion, self-esteem, and adolescents’ risky behaviors. Specifically, we tested
two competing models of how the self-system of adolescents may be related
to risky behavior: whether self-esteem mediates the relation between future
orientation and risk orientation, or whether future orientation—a form of pos-
sible selves—mediates the association between self-esteem and later risk
Jackman and MacPhee 17
orientation. Both future orientation and self-esteem were significantly,
inversely related to risk orientation, thus supporting the first two hypotheses.
As found in prior research (Bolland, 2003; Chen & Vazsonyi, 2011; Gardner
& Steinberg, 2005), adolescents who engaged in various risky behaviors,
such as being impulsive and affiliating with delinquent peers, tended to have
more negative thoughts about their future and also had lower levels of self-
esteem. In addition, the results supported both hypotheses related to media-
tion: Both future orientation and self-esteem at Time 2 partially mediated the
relation between the Time 1 predictor and later risk orientation, with social
desirability and cross-time autocorrelations taken into account.
Future Orientation and Risk Orientation in Early Adolescence
What is meant by risk orientation? This study used multiple indices of young
adolescents’ attitudes toward risk and sexual activity as well as engagement
in various risky behaviors such as having unprotected sex, deviant peer affili-
ation, and impulsive behavior. Consistent with previous research on comor-
bidity of adolescent behavior problems (Coleman & Cater, 2005), the various
measures of adolescents’ attitudes toward and engagement in risky behavior
were significantly interrelated and formed a single dimension of risk orienta-
tion. Furthermore, future orientation was significantly correlated with each
risk indicator; self-esteem also predicted later sexual attitudes and impulsiv-
ity as well as the risk composite, although it was not consistently related to
the other measures of risk. These findings indicate that adolescents’ engage-
ment in risk behaviors co-occur, as has been found for 30% or more of ado-
lescents (Dryfoos, 1997), and that these problem behaviors may share similar
risk and protective factors, or equifinality (Dryfoos, 1997). Of practical
import, one implication of these findings is that youth development efforts
that target one or several protective factors, such as future orientation
(Bolland, 2003; Robbins & Bryan, 2004), may have beneficial effects on a
variety of adolescent risk-taking behaviors, as will be discussed in a later
Although previous research on the relation between future orientation and
risky behaviors is limited, most studies have observed that future orientation
may act as a protective factor in relation to adolescent risk taking (e.g., Chen
& Vazsonyi, 2011; Steinberg et al., 2009; Wills, 1994). The inverse relation
between future orientation and peer delinquency is in contrast to Losel’s
(1975) earlier study in which there was no difference in future orientation
between delinquent and nondelinquent teens. One possible explanation for
this finding is that the age of the participants in the current study ranged from
12 to 14 whereas in those in Losel’s study were between the ages of 16 and
18 Journal of Early Adolescence
19. The older teens in Losel’s study may have thought more about their future
selves and what they want from life, which relates to lower levels of risk
engagement (Steinberg et al., 2009), even if they did label themselves as
Findings from the present study support the proposition that possible
selves activate self-regulation when teens strive to avoid negative future
selves and, instead, pursue positive future selves (Oyserman & James, 2009).
Considering that possible selves—one’s goals, motives, anxieties, and
fears—provide a platform for understanding the development of one’s future
in relation to engagement in risks, it is an appropriate organizing construct in
understanding the current study’s findings (Markus & Nurius, 1986). The
adolescents in this study who had more negative perceptions and attitudes
toward risk engagement also were more likely to avoid risky behavior and
had higher levels of future orientation, associations that have been suggested
in previous work on possible selves and positive youth development (e.g.,
Connell, Spencer, & Aber, 1994; Hoyle & Sherrill, 2006; Markus & Nurius,
In addition, social influences may be formative to one’s possible selves
(Hoyle & Sherrill 2006; Markus & Nurius, 1986). Heavy reliance on social
cues and influences has been noted as a possible limitation on what adoles-
cents might actually achieve or how they might behave, in that those whose
social influence is negative tend to have negative future selves and may
engage in more risky behaviors (Oyserman et al., 2006). As well, individual
differences in adolescents’ susceptibility to social influence may obscure the
relation between future orientation and adolescent risk taking, as was found
in this study: Social desirability correlated significantly with both the predic-
tor and criterion variables. In order to avoid confounding results, social desir-
ability was, therefore, covaried such that the results reflect more accurately
the unique contributions of future orientation and self-esteem to risky behav-
Self-Esteem and Risk Orientation
The results also are in accord with the literature showing self-esteem to be
inversely related to adolescents’ risky behaviors (e.g., Donnellan et al., 2005;
McGee & Williams, 2000). By using longitudinal data, the findings were able
to capture the magnitude of self-esteem in relation to risk behaviors across and
within time. Interestingly, self-esteem was not significantly correlated with
risky sexual engagement at baseline or across time. This null finding could be
due to the fact that the measure of self-esteem reflects global perceptions of
the self whereas literature shows that using situation-specific self-efficacy
Jackman and MacPhee 19
might be a more pertinent protective factor related to risky sexual behaviors
(McGee & Williams, 2000). In addition, due to the small number of 12- to
14-year-olds who were sexually active, there was not sufficient variability in
the measure, even with scores transformed, to detect a meaningful relation to
As hypothesized, self-esteem was significantly related to future orientation.
Even though there is limited literature on this association, several studies have
shown future orientation and self-esteem to be highly correlated with each other
(e.g., McGee, Williams, & Nada-Raja, 2001; Trzesniewski et al., 2006). A strong,
positive sense of self-worth may aid adolescents’ identity development (Harter,
1999; Marcia, 1980). Self-theories assert that adolescents who think critically
about their future selves as well as have a positive sense of self are more inclined
to explore various aspects of their identity and also fare better than those who do
not (Harter, 1999; Hoyle & Sherrill, 2006). The correlations between future ori-
entation and self-esteem within each measurement occasion as well as across
time suggest that these two processes may develop in tandem, although this pos-
tulate has not yet been examined. In addition, the relation between these two
processes, as well as their relation to risky behavior, lends credence to the focus
of positive youth development efforts on promoting positive futures and compe-
tent selves (see Lerner, Almerigi, Theokas, & Lerner, 2005).
The hypothesis that self-esteem mediates the relation between future orien-
tation and risk orientation was supported as was the hypothesis that future ori-
entation mediates the association between self-esteem and risk orientation.
Given the limited literature related to which mediational model best predicts
later risk engagement, the results provide a useful starting point for a discussion
of protective mechanisms related to adolescents’ risky behavior. For instance,
interventions to prevent adolescent risk taking not uncommonly include atten-
tion to positive self-perceptions (Crenshaw & Lee, 2010), based on the assump-
tion that higher self-esteem or self-worth empowers youth to engage in healthy
behavior and to resist peer pressure (Bámaca & Umaña-Taylor, 2006). Indeed,
findings from the current study build on the limited literature, which docu-
ments contemporaneous inverse relations between self-esteem and adolescents’
risk-taking behavior (Donnellan et al., 2005; Wills, 1994), by demonstrating
that self-esteem predicts later risk orientation and also mediates the association
between future orientation and risk taking, net other predictors and covariates.
The finding that future orientation mediated the relation between self-
esteem and risk orientation is consistent with previous studies in that more
positive thoughts of one’s future are associated with controlling and regulating
one’s behaviors (Harris et al., 2002; Pulkkinen & Rönkä, 1994; Steinberg
et al., 2009), including delay of gratification (Trommsdorff, Lamm, &
Schmidt, 1979). Future orientation serves as a protective factor against
20 Journal of Early Adolescence
engaging in risky behavior, in that youth with a clear sense of self or purpose
are less likely to risk their long-term goals by engaging in delinquent activity
or unprotected sex (Chen & Vazsonyi, 2011, 2013; Nurmi, 1991; Steinberg,
2004; Steinberg et al., 2009). Also, the finding that both future orientation
and self-esteem were significant mediators is an important contribution to the
literature because it suggests that they may be intertwined protective factors
in relation to adolescent risk behaviors.
Although self-concept frameworks and theories are useful tools to guide
prevention efforts such as positive youth development programs (Jemmott
et al., 1992), an intervention program to reduce risky behaviors should miti-
gate risk factors as well as promote protective factors. The results of the cur-
rent study, as well as recent long-term longitudinal studies (Chen & Vazsonyi,
2011, 2013), suggest that interventions to promote adolescents’ more opti-
mistic perceptions of their futures could improve their mental health (Reivich,
Gillham, Chaplin, & Seligman, 2013) and reduce their engagement in risky
behavior. As well, other aspects of the self-system may be implicated in the
prevention of adolescent risk taking, particularly self-efficacy (Wills, 2004).
Limitations and Future Research
Although these results were informative and supported the hypotheses, it is
important to note the study’s limitations. First, the measure used to assess
future orientation had marginal reliability, in part because an 8-item scale was
selected from a broader 56-item measure of resilience (Constantine et al.,
1999). Also, future orientation is a multifaceted construct that includes moti-
vation, emotion, and cognition; not all of these facets were included in the
measure used. A more complete understanding of the mechanisms involved
in future orientation’s contribution to risk orientation would have been pos-
sible if other pertinent factors had been included such as selection of peers
who promote achievement of goals, types of adult support recruited in the
service of goals (such as parents, mentors, teachers), planfulness, and conse-
quential thinking. Therefore, even though future orientation was strongly
correlated with risk orientation, whether consequential thinking, sense of
purpose, or another aspect of future orientation accounts for this association
remains to be ascertained.
In addition, the self-esteem measure used was an assessment of global
self-esteem. However, as other studies have shown, domain-specific mea-
sures of self-esteem sometimes better assess the relation between self-esteem
and specific risk behaviors (see McGee & Williams, 2000). Therefore, in
future research, domain-specific self-esteem measures could be used that are
aligned with the specific outcomes being assessed, in conjunction with global
Jackman and MacPhee 21
measures of self-esteem, in order to tease apart the contributors to engage-
ment in risky behaviors.
Another limitation of this study is that only two time points, separated by
6 months, were included, which does not permit as powerful a test of media-
tion as a longitudinal study with multiple data points, and also does not allow
one to assess violations of stationarity (see Maxwell & Cole, 2007). Causal
pathways among self-esteem, future orientation, and adolescents’ risk orien-
tation would be better assessed with long-term longitudinal studies using
growth curve modeling (see, for example, Chen & Vazsonyi, 2011).
The results of this study illuminate the need for more studies to include
future orientation and self-esteem in their intervention and health promotion
efforts. This is especially important when developing a program grounded in
the positive youth developmental framework (Lerner et al., 2005). This frame-
work is comprised of five core developmental components: care, competence,
compassion, confidence, and connection. The “competence” domain sub-
sumes self-esteem and future orientation and, in turn, is presumed to influence
as well as be sensitive to the other facets of youth development (Lerner et al.,
2005). As a result, if future orientation and self-esteem are included in such
programs, further assessments of causality can be made, by determining
whether changes in either predictor lead to changes in risky behavior. One
effective approach is the possible selves intervention (Oyserman et al., 2006),
which showed that long-term changes in academic skills and school miscon-
duct were mediated by changes in future orientation and planning skills.
The overarching goal of this study was to better understand the underlying
mechanisms and processes involved in adolescents’ risky behavioral engage-
ment. Results from this study extend the existing literature on the relations of
self-esteem and future orientation to adolescents’ risky behaviors and attitudes.
The effects of mediation also provide a basis for follow-up research. The find-
ings show that future orientation and self-esteem have unique prospective asso-
ciations with risky behaviors and attitudes, which is necessary but not sufficient
when attempting to ascertain causal pathways (Nurmi, 1991; Salazar et al.,
2004; Wills, 2004). By examining some of the putative mechanisms involved
in adolescent risk taking, insights have been gained into how possible selves
may contribute to avoidance of behaviors that may imperil adolescents’ futures.
This study was completed in partial fulfillment of the requirements for Danielle
Jackman’s master’s degree, Department of Human Development and Family Studies,
Colorado State University. Aimée K. Walker provided valuable assistance with the
analyses of mediation. An earlier version of this study was presented at the 2013 bien-
nial meeting of the Society for Research in Child Development, Seattle.
22 Journal of Early Adolescence
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.
The author(s) disclosed receipt of the following financial support for the research, author-
ship, and/or publication of this article: This research was supported by Grants
APHPA002066 and APHPA006045 to Jan Miller-Heyl and David MacPhee from the
Office of Adolescent Pregnancy Programs, Office of Population Affairs, U.S. Department
of Health and Human Services.
Bámaca, M. Y., & Umaña-Taylor, A. J. (2006). Testing a model of resistance to peer
pressure among Mexican-origin adolescents. Journal of Youth and Adolescence,
35, 631-645. doi:10.1007/s10964-006-9055-4
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinc-
tion in social psychological research: Conceptual, strategic, and statistical
considerations. Journal of Personality and Social Psychology, 51, 1173-1182.
Basen-Engquist, K. K., Mâsse, L. C., Coyle, K. K., Kirby, D. D., Parcel, G. S.,
Banspach, S. S., & Nodora, J. J. (1999). Validity of scales measuring the psy-
chosocial determinants of HIV/STD-related risk behavior in adolescents. Health
Education Research, 14, 25-38. doi:10.1093/her/14.1.25
Baumeister, R. F., Heatherton, T. F., & Tice, D. M. (1993). When ego threats lead
to self-regulation failure: Negative consequences of high self-esteem. Journal of
Personality and Social Psychology, 64, 141-156. doi:10.1177/0146167206289408
Baumeister, R. F., Smart, L., & Boden, J. M. (1996). Relation of threatened egotism
to violence and aggression: The dark side of high self-esteem. Psychological
Review, 103, 5-33.
Biglan, A., Brennan, P. A., Foster, S. L., & Holder, H. D. (2004). Helping adolescents
at risk: Prevention of multiple problem behaviors. New York, NY: Guilford.
Bolland, J. M. (2003). Hopelessness and risk behaviors among adolescents living in
high-poverty, inner city neighborhoods. Journal of Adolescence, 26, 145-158.
Cahalan, D., & Room, R. (1974). Problem drinking among American men (Monograph
No. 7). New Brunswick, NJ: Rutgers Center of Alcohol Studies.
Capaldi, D. M., & Patterson, G. (1989). Psychometric properties of fourteen latent
constructs from the Oregon Youth Study. New York, NY: Springer-Verlag.
Chen, P., & Vazsonyi, A. T. (2011). Future orientation, impulsivity, and problem
behaviors: A longitudinal moderation model. Developmental Psychology, 47,
Chen, P., & Vazsonyi, A. T. (2013). Future orientation, school contexts, and prob-
lem behaviors: A multilevel study. Journal of Youth and Adolescence, 42, 67-81.
Jackman and MacPhee 23
Choi, H., Meininger, J. C., & Roberts, R. E. (2006). Ethnic differences in adolescents’
mental distress, social stress, and resources. Adolescence, 41, 263-283.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.).
Hillsdale, NJ: Lawrence Erlbaum.
Cole, D. A., & Maxwell, S. E. (2003). Testing mediational models with longitudinal
data: Questions and tips in the use of structural equation modeling. Journal of
Abnormal Psychology, 112, 558-577. doi:10.1037/0021-843X.112.4.558
Coleman, L. M., & Cater, S. M. (2005). A qualitative study of the relationship between
alcohol consumption and risky sex in adolescents. Archives of Sexual Behavior,
34, 649-661. doi:10.1007/s10508-005-7917-6
Connell, J. P., Spencer, M. B., & Aber, J. L. (1994). Educational risk and resilience
in African American youth: Context, self, actions, and outcomes in school. Child
Development, 65, 493-506. doi:10.0009-3920/94/6502-0016
Constantine, N., Benard, B., & Diaz, M. (1999, June). Measuring protective factors
and resilience traits in youth: The healthy kids resilience assessment. Paper pre-
sented at the Seventh Annual Meeting of the Society for Prevention Research,
New Orleans, LA.
Crenshaw, D. A., & Lee, J. (2010). At-risk adolescents: Self-esteem, youth violence,
and teen pregnancy. In M. H. Guindon (Ed.), Self-esteem across the lifespan:
Issues and interventions (pp. 143-157). New York, NY: Routledge.
Crocetti, E., Rubini, M., Luyckx, K., & Meeus, W. (2008). Identity formation in early
and middle adolescents from various ethnic groups: From three dimensions to
five statuses. Journal of Youth Adolescence, 37, 983-996. doi:10.1007/s10964-
Crocker, J., & Park, L. E. (2004). The costly pursuit of self-esteem. Psychological
Bulletin, 130, 392-414. doi:10.1037/0033-2909.130.3.392
Dishion, T. J., Kim, H., Stormshak, E. A., & O’Neill, M. (2014). A brief measure of
peer affiliation and social acceptance (PASA): Validity in an ethnically diverse
sample of early adolescents. Journal of Clinical Child & Adolescent Psychology,
43, 601-612. doi:10.1080/15374416.2013.876641
Donnellan, B. M., Trzesniewski, K. H., Robins, R. W., Moffitt, T. E., & Caspi,
A. (2005). Low self-esteem is related to aggression, antisocial behavior,
and delinquency. Psychological Science, 16, 328-335. doi:10.1111/j.0956-
Dryfoos, J. G. (1997). The prevalence of problem behaviors: Implications for pro-
grams. In R. P. Weissberg, T. P. Gullotta, R. L. Hampton, B. A. Ryan, & G. R.
Adams (Eds.), Healthy children 2010: Enhancing children’s wellness (pp. 17-
46). Thousand Oaks, CA: SAGE.
DuBois, D. L., Burk-Braxton, C., Swenson, L. P., Tevendale, H. D., Lockerd, E. M.,
& Moran, B. L. (2002). Getting by with a little help from self and others: Self-
esteem and social support as resources during early adolescence. Developmental
Psychology, 38, 822-839. doi:10.1037//0012-16184.108.40.2062
Dunkel, C. S. (2000). Possible selves as a mechanism for identity exploration. Journal
of Adolescence, 23, 519-529. doi:10.1006/jado.2000.0340
24 Journal of Early Adolescence
Epel, E. S., Bandura, A., & Zimbardo, P. G. (1999). The influence of self-efficacy
and time perspective on coping with homelessness. Journal of Applied Social
Psychology, 29, 575-596.
Epstein, S. (1973). The self-concept revisited: Or a theory of a theory. American
Psychologist, 28, 404-416.
Fergusson, D. M., & Horwood, L. (1999). Prospective childhood predictors of deviant
peer affiliations in adolescence. Journal of Child Psychology and Psychiatry, 40,
Fischer, J., & Corcoran, K. (1994). Measures for clinical practices: A sourcebook
(2nd ed., Vol. 1: Couples, families, and children). New York, NY: Free Press.
Gardner, M., & Steinberg, L. (2005). Peer influence on risk taking, risk preference,
and risky decision making in adolescence and adulthood: An experimental
study. Journal of Developmental Psychology, 41, 625-635. doi:10.1037/0012-
Gjesme, T. (1979). Future time orientation as a function of achievement motives, abil-
ity, delay of gratification, and sex. Journal of Psychology, 101, 173-188. doi:10.
Gottfredson, M. R., & Hirschi, T. (1990). A general theory of crime. Stanford, CA:
Gullone, E., Moore, S., Moss, S., & Boyd, C. (2000). The Adolescent Risk-Taking
Questionnaire: Development and psychometric evaluation. Journal of Adolescent
Research, 15, 231-250. doi:10.1177/0743558400152003
Hagborg, W. J. (1993). The Rosenberg Self-Esteem Scale and Harter’s Self-
Perception Profile for Adolescents: A concurrent validity study. Psychology
in the Schools, 30, 132-136. doi:10.1002/1520-6807(199304)30:2<132::AID-
Hanson, T. L., & Kim, J. (2007). Measuring resilience and youth development: The
psychometric properties of the Healthy Kids Survey. Retrieved from http://www.
Harris, K. M., Duncan, G. J., & Boisjoly, J. (2002). Evaluating the role of “nothing to
lose” attitudes on risky behavior in adolescence. Social Forces, 80, 1005-1039.
Harter, S. (1999). The construction of the self: A developmental perspective. New
York, NY: Guilford.
Harter, S., & Whitesell, N. R. (2003). Beyond the debate: Why some adolescents report
stable self-worth over time and situation, whereas other report changes in self-
worth. Journal of Personality, 71, 1027-1052. doi:10.1111/1467-6494.7106006
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process
analysis: A regression-based approach. New York, NY: Guilford.
Hayes, A. F., & Preacher, K. J. (2014). Statistical mediation analysis with a multicat-
egorical independent variable. British Journal of Mathematical and Statistical
Psychology, 67, 451-470. doi:10.1111/bmsp.12028
Hoyle, R. H., & Sherrill, M. R. (2006). Future orientation in the self-system: Possible
selves, self-regulation, and behavior. Journal of Personality, 74, 1673-7696.
Jackman and MacPhee 25
Jemmott, J. B., Jemmott, L. S., Spears, H., Hewitt, N., & Cruz-Collins, M. (1992).
Self-efficacy, hedonistic experiences, and condom use intentions among inner-
city Black adolescent women: A social cognitive approach to AIDS risk behavior.
Journal of Adolescent Health, 13, 512-519. doi:10.1016/1054-139X(92).0016-5
Jessor, R., Turbin, M. S., Costa, F. M., Dong, Q., Zhang, H., & Wang, C. (2003).
Adolescent problem behavior in China and the United States: A cross-national
study of psychosocial protective factors. Journal of Research on Adolescence,
13, 329-360. doi:10.1111/1532-7795.1303004
Kroger, J., Martinussen, M., & Marcia, J. (2010). Identity status change during ado-
lescence and young adulthood: A meta-analysis. Journal of Adolescence, 33,
Lerner, R. M., Almerigi, J. B., Theokas, C., & Lerner, J. (2005). Positive youth devel-
opment: A view of the issues. The Journal of Early Adolescence, 25, 10-16.
Lessing, E. (1972). Extension of personal future time perspective, age, and satis-
faction of children and adolescents. Developmental Psychology, 6, 457-468.
Losel, F. (1975). Handlungskontrolle und Jugenddelinquenz [Action-control and
youth delinquency]. Stuttgart, Germany: Enke.
Luster, T., & Small, S. A. (1994). Factors associated with sexual risk-taking
behaviors among adolescents. Journal of Marriage and Family, 56, 622-632.
Luyckx, K., Goossens, L., Soenens, B., Beyers, W., & Vansteenkiste, M. (2005).
Identity statuses based on 4 rather than 2 identity dimensions: Extending and
refining Marcia’s paradigm. Journal of Youth and Adolescence, 34, 605-618.
MacKinnon, D. P., Fairchild, A. J., & Fritz, M. S. (2007). Mediation analysis. Annual
Review of Psychology, 58, 593-614.
MacKinnon, D. P., Lockwood, C. M., & Williams, J. (2004). Confidence limits for the
indirect effect: Distribution of the product and resampling methods. Multivariate
Behavioral Research, 39, 99-128. doi:10.1207/s15327906mbr3901_4
Marcia, J. E. (1980). Identity in adolescence. In J. Adelson (Ed.), Handbook of ado-
lescent psychology (pp. 159-187). New York, NY: Wiley.
Markus, H., & Nurius, P. (1986). Possible selves. American Psychologist, 41, 954-969.
Marsh, H. W. (1986). Global self-esteem: Its relation to specific facets of self-concept
and their importance. Journal of Personality and Social Psychology, 51, 1224-
Maxwell, S. E., & Cole, D. A. (2007). Bias in cross-sectional analyses of longitudinal
mediation. Psychological Methods, 12, 23-44. doi:10.1037/1082-989X.12.1.23
McFarlin, D. B., & Blascovich, J. (1981). Effects of self-esteem and performance
feedback on future affective preferences and cognitive expectations. Journal of
Personality and Social Psychology, 3, 521-531. doi:10.1037/0022-35220.127.116.111
McGee, R., & Williams, S. (2000). Does low self-esteem predict health compro-
mising behaviors among adolescents? Journal of Adolescence, 23, 569-582.
26 Journal of Early Adolescence
McGee, R., Williams, S., & Nada-Raja, S. (2001). Low self-esteem and hopelessness
in childhood and suicidal ideation in early adulthood. Journal of Abnormal Child
Psychology, 29, 281-291.
Meeus, W. (2011). The study of adolescent identity formation 2000-2010: A review
of longitudinal research. Journal of Research on Adolescence, 21, 75-94.
Mesmer-Magnus, J., Viswesvaran, C., Deshpande, S., & Joseph, J. (2006). Social
desirability: The role of over-claiming, self-esteem, and emotional intelligence.
Psychology Science, 48, 336-356.
Metzler, C. W., Noell, J., & Biglan, A. (1992). The validation of a construct of high-
risk sexual behavior in heterosexual adolescents. Journal of Adolescent Research,
7, 233-249. doi:10.1177/074355489272007
Nurmi, J. E. (1991). How do adolescents see their future? A review of the devel-
opment of future orientation and planning. Development Review, 11, 1-59.
Nurmi, J. E., & Pulliainen, H. (1991). The changing parent-child relationship,
self-esteem, and intelligence as determinants of orientation to the future dur-
ing early adolescence. Journal of Adolescence, 14, 35-51. doi:10.1016/
Oyserman, D., Bybee, D., & Terry, K. (2006). Possible selves and academic out-
comes: How and when possible selves impel action. Journal of Personality and
Social Psychology, 91, 188-204. doi:10.1037/0022-3518.104.22.168
Oyserman, D., & James, L. (2009). Possible selves: From content to process. In K. D.
Markman, W. M. P. Klein, & J. A. Suhr (Eds.), The handbook of imagination and
mental stimulation (pp. 373-394). New York, NY: Psychology Press.
Pulkkinen, L., & Rönkä, A. (1994). Personal control over development, identity
formation, and future orientation as components of life orientation: A develop-
mental approach. Developmental Psychology, 30, 260-271. doi:10.1037/0012-
Reivich, K., Gillham, J. E., Chaplin, T. M., & Seligman, M. P. (2013). From helpless-
ness to optimism: The role of resilience in treating and preventing depression in
youth. In S. Goldstein & R. B. Brooks (Eds.), Handbook of resilience in children
(2nd ed., pp. 201-214). New York, NY: Springer Science + Business Media.
Reynolds, W. M. (1982). Development of reliable and valid short forms of the
Marlowe-Crowne Social Desirability Scale. Journal of Clinical Psychology, 38,
Robbins, R., & Bryan, A. (2004). Relationships between future orientation, impulsive
sensation seeking, and risk behavior among adjudicated adolescents. Journal of
Adolescent Research, 19, 428-445. doi:10.1177/0743558403258860
Robinson, J. P., Shaver, P. R., & Wrightsman, L. S. (Eds.). (1991). Measures of per-
sonality and social psychological attitudes. San Diego, CA: Academic Press.
Romer, D., Duckworth, A. L., Sznitman, S., & Park, S. (2010). Can adolescents learn
self-control? Delay of gratification in the development of control over risk taking.
Prevention Science, 11, 319-330. doi:10.1007/s11121-010-0171-8
Jackman and MacPhee 27
Rooke, S., & Hine, D. W. (2011). A dual-process account of adolescent and adult
binge drinking. Journal of Addictive Behaviors, 36, 341-346. doi:10.1016/j.add-
Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton
Rosenberg, M., Schooler, C., & Schoenbach, C. (1989). Self-esteem and adoles-
cent problems: Modeling reciprocal effects. American Sociological Review, 54,
Rosenberg, M., Schooler, C., Schoenbach, C., & Rosenberg, F. (1995). Global
self-esteem and specific self-esteem: Different concepts, different outcomes.
American Sociological Review, 60, 141-156.
Salazar, L. F., Crosby, R. A., DiClemente, R. J., Wingood, G. M., Lescano, C. M.,
Brown, L. K., . . . Davies, S. (2005). Self-esteem and theoretical mediators of
safer sex among African American female adolescents: Implications for sex-
ual risk reduction interventions. Health Education & Behavior, 32, 413-427.
Salazar, L. F., DiClemente, R. J., Wingood, G. M., Crosby, R. A., Harrington, K.,
Davies, S., . . . Oh, K. M. (2004). Self-concept and adolescents’ refusal of
unprotected sex: A test of mediating mechanism among African American girls.
Prevention Science, 5, 137-149.
Schafer, J. L., Blanchard, L., & Fals-Stewart, W. (1994). Drug use and risky sex-
ual behaviors. Psychology of Addictive Behaviors, 8, 3-7. doi:10.1037/0893-
Schafer, J. L., & Graham, J. W. (2002). Missing data: Our view of the state of the art.
Psychological Methods, 7, 147-177. doi:10.1037/1082-989X.7.2.147
Schwartz, S. J. (2007). The structure of identity consolidation: Multiple correlated
constructs of one superordinate construct? Identity: An International Journal of
Theory and Research, 7, 27-49.
Schwartz, S. J., Mason, C. A., Pantin, H., & Szapocznik, J. (2008). Effects of fam-
ily functioning and identity confusion on substance use and sexual behavior
in Hispanic immigrant early adolescents. Identity: An International Journal of
Theory and Research, 8, 107-124.
Seginer, R. (2009). Future orientation: Developmental and ecological perspectives.
New York, NY: Springer.
Steinberg, L. (2004). Risk taking in adolescence: What changes and why? Annals of
the New York Academy of Sciences, 1021, 51-58. doi:10.1196/annals.1308.005
Steinberg, L., O’Brien, L., Cauffman, E., Graham, S., Wooland, J., & Banich, M.
(2009). Age differences in future orientation and delay discounting. Child
Development, 80, 28-44. doi:10.1111/j.1467-8624.2008.01244.x
Trocki, K. F., Drabble, L. A., & Midanik, L. T. (2009). Tobacco, marijuana, and
sensation seeking: Comparisons across gay, lesbian, bisexual, and heterosexual
group. Psychology of Addictive Behaviors, 23, 620-631. doi:10.1037/a0017334
Trommsdorff, G., & Lamm, H. (1980). Future orientation of institutionalized and
noninstitutionalized delinquents and nondelinquents. European Journal of Social
Psychology, 10, 247-278.
28 Journal of Early Adolescence
Trommsdorff, G., Lamm, H., & Schmidt, R. W. (1979). A longitudinal study of ado-
lescent’s future orientation (time perspective). Journal of Youth and Adolescence,
Trzesniewski, K. H., Donnellan, M. B., Moffitt, T. E., Robbins, R. W., Poulton, R., &
Caspi, A. (2006). Low self-esteem during adolescence predicts poor health, crim-
inal behavior, and limited economic prospects during adulthood. Developmental
Psychology, 42, 381-390. doi:10.1037/0012-1622.214.171.1241
Williams, J., & MacKinnon, D. P. (2008). Resampling and distribution of the prod-
uct methods for testing indirect effects in complex models. Structural Equation
Modeling, 15, 23-51. doi:10.1080/10705510701758166
Wills, T. (1994). Self-esteem and perceived control in adolescent substance use:
Comparative tests in concurrent and prospective analyses. Psychology of
Addictive Behaviors, 8, 223-234. doi:10.1037/0893-164X.8.4.223
Zimmer-Gembeck, M., & Petherick, J. (2006). Intimacy dating goals, and relation-
ship satisfaction during adolescence and emerging adulthood: Identity formation,
age, and sex are moderators. International Journal of Behavioral Development,
Danielle M. Jackman is a graduate of the Applied Developmental Science doctoral
program at Colorado State University and is currently a program evaluator for Aurora
Mental Health Services. Prior to her attendance at Colorado State University, she
attended Barry University where she obtained her bachelor’s degree in May 2009. Her
research interests include adolescent risky behaviors (delinquency, substance use),
positive youth developmental (PYD) framework, and adolescent self-perceptions.
David MacPhee is a professor in human development and family studies at Colorado
State University. He received his PhD in developmental psychology from the
University of North Carolina–Chapel Hill. His research and teaching focus on parent-
ing, risk and resilience (especially related to school readiness and youth develop-
ment), effective family-strengthening programs, and diversity issues.