University of Southern California Dissertations and Theses

Emotional intelligence and smoking risk factors in early adolescents

(USC Thesis Other)

Emotional intelligence and smoking risk factors in early adolescents

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content EMOTIONAL INTELLIGENCE AND SMOKING RISK FACTORS
IN EARLY ADOLESCENTS
Copyright 2002
by
Dennis Ryan Trinidad
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Preventive Medicine / Health Behavior Research)
August 2002
Dennis Ryan Trinidad
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 3094375
Copyright 2002 by
Trinidad, Dennis Ryan
All rights reserved.
®
UMI
UMI Microform 3094375
Copyright 2003 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
The Graduate School
University Park
LOS ANGELES, CALIFORNIA 900894695
This dissertation, w ritten b y
P e w u Tftt^tDAQ _______________
U nder th e direction o f hJ.L. D issertation
Com m ittee, an d approved b y a ll its m em bers,
has been p resen ted to and accepted b y The
Graduate School, in p a rtia l fulfillm en t o f
requirem ents fo r th e degree o f
DOCTOR OF PHILOSOPHY
£ * .....I T 1 *
‘ ‘ Dean o f G raduate S tu d ies
D ate
August 6 .^ .2002
DISSER TA TION COMMITTEE
C hairperson
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DEDICATION
To This Great Nation: A Wonderful Land Of Opportunity
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGMENTS
I would like to thank members of my guidance committee: Margy Gatz for
keeping me straight in thinking through my conclusions, for her positive feedback,
and for putting me at ease at the start of my dissertation; Jennifer Unger for being
my “academic big sister,” sharing knowledge, being generous with co-authorships,
being a sounding board, and teaching me all sorts of stuff; Chih-Ping Chou for
always answering my questions, being patient in waiting for the light to come on in
my head, providing guidance and willingness to share his statistical knowledge;
Stan Azen for making statistics fun, guiding me in what classes to take, and for
providing one of the many soundtracks for my late night studies; and especially
Andy Johnson for being my mentor, for believing in me, opening doors for me,
constantly challenging me, teaching me the value of optimism, resilience and
enthusiasm; and for being one of the best friends I’ve known in my short life.
I would also like to thank Drs. Peter Salovey, David Caruso, and John Mayer
for allowing the use of the MEIS and for sharing their ideas and research; Dr. Ping
Sim for taking the time to share his statistical and SAS expertise; Dr. Francisco
Buchting for his assistance with the TRDRP grant, as well as Dr. Phil Gardiner and
the rest of the TRDRP staff; Dr. Paula Palmer for the lessons she taught me about
the IRB; Gaylene Gunning for helping me recruit schools; Steven Cen for his data
management skills; Dr. Peggy Gallaher for her methodological and statistical
advice; Jolanda Lisath, Susan Cooper, Karyn Evaro, Bruce Missagia, Gabby Torres,
Kari-Lyn Kobayakawa, Annette Stoneking, Albert Rivera, Michelle Orsillo, Mary
iii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Trujillo, Patti Gutierrez, and the rest of the IPR/Biometry staff for teaching me
about the other sides of academia; Drs. Luanne Rohrbach and Tess Cruz for lending
their ears and providing suggestions; Dr. Steve Sussman for always giving me
something to think about; Mamy Barovich for everything; Carol Gordon for looking
out for me; and my fellow students, especially Anamara Ritt-Olson, Jenny Zogg,
Cheryl Nordstrom, Silvana Skara, Lorena Teran, Nikki Shipley, Darleen Schuster,
Made’ Wenten, Terry Huang, Sondos Islam, Chaoyang Li, Dongyun Yang, Cindy
Zheng, Cher Bingman, Joel Milam, Michelle Weiner and Beth Hoffman for their
friendship.
I also want to thank all of my parents, siblings and cousins for their support,
even though they might not totally understand what I do. I am especially grateful
for the encouragement and unconditional support provided by my wonderful wife,
Geraldine. Her own hard work and perseverance, sense of humor, and beautiful
smile help me keep things in perspective. She is, and always will be, my
inspiration.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
Dedication ii
Acknowledgements iii
List of Tables vi
List of Figures vii
Abstract viii
Chapter 1: The association between emotional intelligence and adolescent 1
tobacco use: Theoretical conceptualization and implications
Chapter 2: The protective association of emotional intelligence with 20
psychosocial smoking risk factors for adolescents
Chapter 3: Emotional intelligence and smoking risk factors in adolescents: 52
Interactions on smoking intentions
Chapter 4: Emotional intelligence and ethnicity: 85
Interactions on smoking intentions in early adolescents
Chapter 5: Emotional intelligence and acculturation to the US: 111
Interactions on the perceived social consequences
of smoking in early adolescents
Chapter 6: Summary and Conclusion 142
Master Reference List 145
v
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF TABLES
Chapter 1
Table 1: Models of Emotional Intelligence
Chapter 2
Table 2: Age & Gender Characteristics
Table 3: Mean Social Consequences and Refusal Self-Efficacy Scores
by El Quartile
Table 4. Odds Ratios for Smoking Intentions by El Quartile
Chapter 3
Table 5: Age & Gender Characteristics
Table 6. Summary of separate logistic regression models examining
interactions between smoking risk factors and El
on smoking intentions
Chapter 4
Table 7: Demographic Characteristics
Table 8: Emotional Intelligence and Smoking Intention
across Ethnicity and Gender
Table 9: Ethnic differences in El on Intentions to Smoke in the Next Year
Chapter 5
Table 10: Demographic Characteristics
Table 11: Mean scores for Emotional Intelligence, AHIMSA US Orientation,
and Perceived Social Consequences of Smoking across
Ethnicity and Gender
Table 12: US Acculturation differences in Emotional Intelligence on
Perceptions of Negative Social Consequences of Smoking
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF FIGURES
Chapter 2
Figure 1: Perceived Negative Social Consequences by El Quartile
Figure 2: Refusal Self-Efficacy by El Quartile
Figure 3: Odds Ratios for Intentions to Smoke by El Quartile
Chapter 3
Figure 4: Interaction between emotional intelligence and refusal
self-efficacy on smoking intentions
Figure 5: Interaction between emotional intelligence and ever trying
cigarettes on smoking intentions
Figure 6: Interaction between emotional intelligence and hostility
on smoking intentions
Chapter 5
Figure 7: Odds Ratio of Perceiving Social Consequences of Smoking
by Emotional Intelligence Level by US Acculturation
Chapter 6
Figure 8: Summary of Relationship Between El and Smoking Risk Factors
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABSTRACT
Background. Adolescent tobacco use continues to be a major public health concern.
Adolescent smoking prevalence has leveled off after an increase in the early 1990s,
but is still higher than at any period of time in the 1980s. New avenues must be
explored in order to further increase the effectiveness of today’s tobacco use
prevention programs. Recent interest in the concept of emotional intelligence (El)
has increased due to the popular media claiming it to be the most important
predictor of life success, with some suggesting that El could account for up to 80%
of that variance. Though much scientific and popular literature views such claims
as overly optimistic, this exciting field of research continues to grow. Exploration
of the relationship between El and specific health behaviors, such as adolescent
smoking, is of particular interest. An understanding of this relationship may help in
designing improved targeted smoking prevention programs for adolescents.
Methods. The data are from the baseline portion of a longitudinal school-based
experimental trial of smoking prevention strategies in a multicultural, urban
population of adolescents in Southern California. The purpose of the baseline
survey was to assess tobacco use and related psychosocial and cultural variables
before the implementation of culturally relevant smoking prevention programs. A
measure of El was administered to a subset of students from the control condition of
the main prevention trial. Analyses were performed on this subset of students.
viii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Subjects were 416 actively consented sixth-grade adolescents from public middle
schools in the greater Los Angeles area. This sample population had a mean age
11.3 years and was 53% female. The ethnic distribution was 32% Latino, 29%
Asian/Pacific Islander, 13% White, 19% Multiethnic, and 6% Other (African
American students were included in this category due to a very small number in our
sample).
Results. High El was related to an increased perception of the negative social
consequences of smoking, higher refusal self-efficacy toward potential cigarette
offers, and lower smoking intentions. Those with low El were more likely to intend
to smoke if they had low refusal skills or were more hostile, while those with high
El were more likely to intend to smoke if they have previously experimented with
cigarettes. The association between emotional intelligence and smoking intentions
did not statistically significantly vary across culture/ethnicity, though there was a
trend towards significance. The trend suggested that El might have been more
protective against smoking intentions in the next year for White adolescents
compared to A/PI and H/L adolescents. Finally, as El increased so did perceptions
of the social consequences of smoking for those who were acculturated to the US
culture.
Conclusion. There is now mounting evidence for emotional intelligence as a
protective factor against adolescent smoking and smoking risk factors. Much
ix
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
research on adolescent smoking prevention programs has identified risk factors,
many of which cannot be modified such as ethnicity. However, inclusion of an EI-
enhancing component in smoking prevention programs may be beneficial.
Improving El skills such as identifying emotions within oneself and in others,
understanding how emotions come about, and managing emotions can augment
existing prevention efforts. With the recent plateau in the success of adolescent
smoking prevention programs, identification of such a novel, protective factor as El
brings hope for further reducing adolescent tobacco use.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 1:
The association between emotional intelligence and adolescent tobacco use:
Theoretical conceptualization and implications
Adolescent Smoking Today
Adolescent tobacco use continues to be a major public health concern.
Adolescent smoking prevalence has leveled off after an increase in the early 1990s,
but is still higher than at any period of time in the 19S0s (Johnston, O'Malley &
Bachman, 1999). New avenues must be explored in order to further increase the
effectiveness of today’s tobacco use prevention programs (Lee, Gilpin & Pierce,
1993, Sussman, Lichtman, Ritt, & Pallonen, 1999).
It is intriguing that adolescent cigarette smoking continues to be a major
1
public health concern despite several psychosocial risk factors being fairly well
established and targeted for intervention (Lee, Gilpin & Pierce, 1993; Welte,
Barnes, Hoffman & Dintcheff, 1999). Recent research literature indicates those
who perceive few negative social consequences associated with smoking are more
likely to smoke in the future (Dalton, Sargent, Beach, Bernhardt & Stevens, 1999;
Distefan, Gilpin, Choi & Pierce, 1998), as are those with low refusal self-efficacy
for cigarette offers (Botvin, Griffin, Diaz, Miller, Ifill-Williams, 1999; Epstein,
Williams, Botvin, Diaz, & Ifill-Williams, 1999). Similarly, smoking intentions, a
component of smoking susceptibility, or an absence of absolute commitment to not
smoke, is a major factor related to future adolescent smoking (Pierce, Choi, Gilpin,
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Farkas, & Merritt, 1996; Unger, Johnson, Stoddard, Nezami & Chou, 1997). So, for
example, if a student perceives few negative consequences associated with smoking,
is less able to refuse tobacco offers, and/or is not absolutely committed to abstaining
from smoking, his or her risk for smoking increases.
To date, the most effective tobacco use prevention programs have focused
on combating social influences, including the aforementioned perceived social
consequences such as perceived peer approval and peer tobacco offers, by
restructuring adolescents’ perceptions regarding cigarette smoking. (Hansen &
Graham, 1991; MacKinnon et al., 1991; Noland, Kryscio, Riggs, Linville, Ford, &
Tucker, 1998). However, given that our current knowledge appears to have reached
a threshold of effectiveness in preventing adolescent smoking, new theories that
may potentially increase an existing program’s effectiveness need to be proposed
and, eventually, evaluated.
Emotional Intelligence
Over the past decade, particularly the latter half, the attention paid to
emotional intelligence (El) has burgeoned. Emotional intelligence has been said to
account for human success more than traditional predictors such as IQ (Goleman,
1995). Emotional intelligence has also been hypothesized to be a key element in
combating health-compromising behaviors, including tobacco use (Goleman, 1995).
Necessarily, any discussion of El should clearly define what emotion is and
what intelligence is. Emotions are processes that arouse, sustain and direct activity
2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and are primary motivating forces (Leeper, 1948). Emotions typically serve to
signal and respond to changes in an individual’s environment (Mayer, Salovey, &
Caruso, 2000). Emotions can also change cognitions, making them negative when
one is sad and positive when one is happy (Mayer, et al., 2000).
Intelligence implies a mental ability associated with cognitive operations
(Mayer, et al., 2000). Sternberg (1997) pointed out that a clear mark of intelligence
is high-level mental ability, such as abstract reasoning. Based on their research on
intelligence Mayer and Salovey (1997) argued that what could be considered
intelligence seemed to be divided into three sub-groups including verbal-
propositional intelligence, spatial-performance intelligence, and social intelligence.
Verbal-propositional intelligence includes logical thinking and measures of
vocabulary, verbal fluency, and the ability to perceive similarities. Spatial-
performance intelligence includes abilities of assembling objects and recognizing
and constructing designs and patterns. Social intelligence involves people’s skills in
relating to one another.
In their conception of intelligence, Mayer and Salovey replaced social
intelligence with emotional intelligence because social intelligence was too closely
correlated with both verbal-propositional and spatial-performance intelligence that it
was virtually indistinguishable from them. In short (but discussed later in greater
detail), social intelligence is a set of social skills involving the modification of one’s
behavior to best suit a given situation and emotional intelligence is a mental ability
that includes emotional perception, relativity, and management. Mayer and Salovey
3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
argued that so much of what is considered intelligence operates in the social domain
that it is difficult for social intelligence to have much discriminant validity (Mayer
& Salovey, 1993). In addition to emotional intelligence being more distinct from
the other two intelligences than social intelligence, Salovey and Mayer also believed
that emotional intelligence was “close enough to the concept of an intelligence to
belong to the triad (of intelligences)” (Mayer & Salovey, 1997, p.8).
A Brief History of Emotional Intelligence
How did the concept of El develop? Prior to work by Salovey and Mayer
(1990) research on emotion and on intelligence resided in different domains, with
intelligence research being conducted in the cognitive realm and emotional research
on the affective realm. It was in this 1990 work that Salovey and Mayer initially
defined emotional intelligence as the ability to understand feelings within oneself
and in others and to use these feelings in guiding thought and action (Salovey &
Mayer, 1990). In other words, their definition suggested that emotion and
intelligence worked together in that feelings could be used to guide and/or assist
thinking.
It was this definition that led to much excitement and discussion over EL
Each new discussion of El, however, seemed to use different definitions and made
different claims about El’s importance (Mayer & Salovey, 1997). Goleman
furthered this excitement and increased El’s popularity by claiming in his 1995
book, Emotional Intelligence: Why It Can Matter More Than IQ, that El was
potentially responsible for up to the 80% of unexplained variance in life success
4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
relative to traditional IQ, which accounts for approximately 20% (Goleman, 1995).
Goleman defined El as “abilities such as being able to motivate oneself and persist
in the face of frustrations; to control impulse and delay gratification; to regulate
one’s moods and keep distress from swamping the ability to think; to empathize and
hope” (p.34). That same year TIME magazine featured on its cover the title,
“What’s Your EQ? It’s not your IQ. It’s not even a number. But EQ may be the best
predictor of success in life, redefining what it means to be smart (TIME, 1995).
Undoubtedly it was these very optimistic claims that contributed to the
popularization of El in the mass media. Mayer and Salovey label these popular
conceptions of El “mixed models” because these new concepts went beyond what
they first defined El to be. These mixed models overlapped considerably with
characteristics of personality and motivation (e.g., Goleman, 1995 and Bar-On,
1997) (Mayer, Salovey & Caruso, 2000). Mayer and Salovey then refined and
specified their definition of El to focus on emotions only and their interactions with
thought and described El in terms of several emotional abilities. Mayer and Salovey
called this an “ability” model and argued for the measurement of El as such (Mayer,
et al., 2000). What follows is a description and critique of these El models.
5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 1 . Models of Emotional Intelligence
MENTAL ABILITY MODEL MIXED MODELS
Mayer & Salovey (1997) Bar-On (1997) Goleman (1995)
Definition Definition Definition
“Emotional intelligence is the set “Emotional intelligence “The abilities called here
of abilities that account for how is... an array of emotional intelligence, which
people’s emotional perception noncognitive capabilities, include self-control, zeal and
and understanding vary in their competencies, and skills persistence, and the ability to
accuracy. More formally, we that influence one’s ability motivate oneself.” (Goldman,
define emotional intelligence as to succeed in coping with 1995a, p. xii). [...and...]
the ability to perceive and environmental demands “There is an old-fashioned
express emotion, assimilate and pressures.” (Bar-On, work for the body of skill that
emotion in thought, understand 1997, p. 14). emotional intelligence
and reason with emotion, and represents: character ”
regulate emotion in the self and
others” (after Mayer & Salovey,
1997).
(Goldman, 1995a, p.28).
Maior Skills & Examples Maior Skills & Examples Maior Skills & Examples
Emotional Perception & Intrapersonal Skills Knowing One’ s Emotions
Expression ■ Emotional self- * Recognizing a feeling
■ Identifying and awareness as it happens
expressing emotions in ■ Assertiveness ■ Monitoring feelings
one’s physical states, ■ Self-Regard from moment to
feelings, and thoughts > Self- moment
Assimilating Emotion in Thought Actualization Management o f Emotions
■ Emotions prioritize ■ Independence ■ Handling feelings so
thinking in productive Interpersonal Skill: they are appropriate
ways ■ Interpersonal ■ Ability to soothe
■ Emotions generated as relationships oneself
aids to judgment and * Social ■ Ability to shake off
memory responsibility rampant anxiety,
Understanding & Analyzing ■ Empathy gloom, or irritability
Emotion Adaptability Scales Motivating Oneself
• Ability to label • Problem solving ■ Marshalling emotions
emotion, including ■ Reality testing in the service of a
complex emotions and H Flexibility goal.
simultaneous feelings Stress -Management ■ Delaying gratification
■ Ability to understand Scales and stifling
relationships associated • Stress tolerance impulsiveness
with shifts of emotion ■ Impulse control ■ Being able to get into
Reflective Regulation o f Emotion General Mood the flow” state
■ Ability to stay open to ■ Happiness Recognizing Emotions in
feelings * Optimism Others
" Ability to monitor and ■ Empathic awareness.
regulate emotions ■ Attunement to what
reflectively to promote other need or want
emotional and Handling Relationships
intellectual growth • Skill in managing
(after Mayer & Salovey, emotions in others
1997, p. 11) * Interacting smoothly
with others.
Adapted from Mayer, Salovey & Caruso, 2000.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Models of Emotional Intelligence
Models of El can be classified into two categories, mental ability models and
mixed models. These models are summarized in Table 1. The mental ability model
emphasizes emotional abilities (Mayer, et al., 2000) and is formally defined as the
ability to accurately perceive, appraise and express emotion; the ability to
access/generate emotions in facilitating thought; the ability to understand emotion
and emotional knowledge; and the ability to regulate emotion in order to promote
emotional and intellectual growth (Mayer & Salovey, 1997). The purpose of this
mental ability model is to hone El to a core conception not mixed with other traits
with which El had gotten associated since its first definition in 1990 by Salovey and
Mayer. Mayer and his colleagues (2000) argue that by utilizing this specific
definition of El it would be possible to analyze the degree to which El abilities
independently contribute to a person’s behavior and general life competence
(Mayer, et al., 2000). However, Mayer and his colleagues (2000) do not claim a
specific amount of life success variance that El accounts for but they do
acknowledge than an increase in even 1% to 5% in such variance accounted for is
important. The first column of Table 1 summarizes how these emotional abilities
are divided into four hierarchical branches, with the most basic branch, Identifying
Emotion, being the accurate perception and expression of emotion. The second
branch, Assimilating Emotion, concerns assimilating emotions in thought (e.g., how
emotions facilitate thought, or how emotions prioritize thinking). The third branch,
Understanding Emotions, describes the ability to understand and analyze emotion
7
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(e.g., labeling emotion and understanding the relationships associated with changes
in emotion). The fourth branch, Managing Emotions, involves the reflective
regulation of emotion and describes the ability to monitor and regulate emotions to
enhance emotional and intellectual growth, such as knowing how to calm oneself
down after being angry.
Mixed models of El can be considered as expanded conceptions of El that
include non-ability traits. They are wider in scope and less specific than the mental
ability model. For example, Bar-On’s (1997) El model combines intrapersonal
skills, interpersonal skills, adaptability, stress management, and general mood.
These five broad areas combine mental abilities, such as emotional self-awareness,
with other characteristics separate from mental abilities including self-regard, social
responsibility, and assertiveness. Bar-On defines emotional intelligence as “an
array of noncognitive capabilities, competencies, and skills that influence one’s
ability to succeed in coping with environmental demands and pressures.” (Bar-On,
1997, p. 14). Bar-On believes that emotional intelligence, along with IQ provides a
more balanced view of one’s intelligence and predicts the potential to succeed. This
model, however, encompasses domains that go beyond the term “emotional
intelligence” (e.g., social responsibility). The second column of Table 1 describes
Bar-On’s model of EL
Goleman’s (1995) mixed model of El is summarized in column 3 of Table 1.
In this model, Goleman identifies five components of El, including knowing one’s
emotions, management of emotions, motivating oneself, recognizing emotions in
8
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
others, and handling relationships. Goleman, by far, makes the most extraordinary
claims about what his model of emotional intelligence predicts, stating that high El
will give “an advantage in any domain in life,” (Goleman, 1995, p. 36) and
potentially accounts for the other 80% of variance not explained by traditional IQ.
However, Goleman himself even states that there is an old-fashioned word for El:
“character” (Goleman, 1995, p. 285). Goleman’s broad definition of El has also
been criticized because what he labels emotional intelligence actually falls under the
realm of social values and social policy (Gardner, 1999), such as working toward a
more smoothly functioning family or society. Mayer et al. (2000) agree with
Gardner (1999) and point out that Goleman’s definition goes beyond the realm of
what would be acceptably defined as an intelligence by including things like
empathy, being considerate, and entering flow states.
These three models are similar in that they each point out that emotional
intelligence contributes to some degree of life success (though these claimed levels
vary greatly) and that each model contains a mental ability component such as the
ability to identify emotions. The main difference between the three models is that
the mental ability model focuses strictly on the interaction of emotion and cognition
while the mixed models include other traits that are beyond the scope of a stringent
definition of emotional intelligence, such as optimism and motivation. Considering
these different conceptions, for scientific research, the best model to employ would
be the mental ability model of Mayer and Salovey (1997). The mental ability model
9
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
is more localized to the common domain between emotion and cognition than the
mixed models.
Is Emotional Intelligence a True Intelligence?
Why should emotional intelligence be considered an intelligence? Early in
their work, Mayer and Salovey (1993) argued that knowing what the self or another
person is feeling may involve considerable thinking and thus be considered an
intelligence (Mayer & Salovey, 1993). This was, however, not empirically
validated until Mayer, Caruso and Salovey (1999) outlined three criteria that make
up traditional standards for an intelligence. First, an intelligence should be able to
be operationalized as a set of abilities. Second these abilities should be
intercorrelated and be related to pre-existing intelligences, while showing unique
variance. Third, these abilities should develop with both age and experience. In
two studies, one with adults and the other with adolescents using Mayer et al.’s
ability measure of emotional intelligence, the Multifactor Emotional Intelligence
Scale (discussed in detail below), Mayer et al. showed that emotional intelligence
met these three criteria for an intelligence. Specifically, the ability tasks that made
up Mayer et al.’s model of El were intercorrelated and formed a single El factor.
These tasks also separated into three factors, Perceiving Emotion, Understanding
Emotion, and Managing Emotion (as opposed to the four factors the authors
hypothesized). El was also moderately correlated with a measure of verbal
intelligence, an accepted form of intelligence. Finally, El level was shown to
increase with age in that adults performed better than adolescents on the MEIS.
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Therefore, classifying El as a traditional intelligence, measured as a mental ability,
provides us with an opportunity to expand our understanding of tobacco use.
Measurement of Emotional Intelligence
Given that the mental ability model of El is the more appropriate model to
use in scientific research, an increasingly accepted measure of El as a mental ability
will now be discussed. The Multifactor Emotional Intelligence Scale (MEIS) is the
first theory-based, comprehensive, competence-based measure of El (Mayer,
Caruso, & Salovey, 1997). Davies, Stankov, & Roberts (1998) point out that self-
report measures of El have been a disappointment in terms of discriminant and
construct validity. The MEIS hopes to improve on those shortcomings. The MEIS
provides an advantage over self-report measures because it gauges an individual’s
performance on a task. Self-report measures are prone to social desirability bias and
an individual’s self-concept. For example, an academically smart student who
doesn’t believe she is smart may report that she thinks she is a below average
student but her grades in school indicate otherwise. In this instance, the grades
received in school would be the competence-based measure of academic smartness.
The MEIS offers this advantage in that it tests an individual’s performance on a set
of mental ability tasks covering the four branches of Mayer & Salovey’s (1997)
model. The first branch, Perceiving Emotion, measures the ability to perceive
emotions in faces, music, designs and stories. The second branch, Assimilating
Emotion, is assessed with synesthesia judgments (e.g., “how hot is anger?”) and
feeling biases, and was expected to measure emotion’s facilitation of cognition.
11
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
However, tests revealed that this branch was a weak factor and Mayer et al. (1999)
recommend that it could be dropped. The third branch, Understanding Emotions,
examines the understanding of emotion, for example how an owner of a dog would
feel if his dog were hit by a car. The fourth branch, Managing Emotions, measures
an individual’s ability to manage emotions in the self and in others by asking the
respondent to rate possible responses to scenarios
In their study of El using the MEIS Mayer and his colleagues (1999)
experimented with three methods of scoring the MEIS to determine correct answers
for the MEIS items (e.g., identifying emotions in faces). Target scoring, the first
method and available for only three of the Perceiving Emotions tasks, asked the
person whose face was depicted on a test item what he or she was feeling. To the
extent a respondent’s answer matched the person’s answer, the answer was scored
as correct. This was repeated for the musical selections and personal stories. A
second method, expert scoring, asked experts on emotion, such as psychotherapists
and emotion researchers, to complete the MEIS and provide correct answers. To the
extent a respondent’s answer matched the experts’ answer, the answer was scored as
correct. The third method, consensus scoring, normed the test based on a large,
heterogeneous sample. To the extent the respondent’s answer matched that of the
group consensus, the answer was scored as correct. Mayer, et al. found that
consensus scoring produced the most reliable measures of emotional intelligence,
and that the various scoring methods (target, expert, and consensus scoring) were
positively correlated, indicating that indeed a set of “correct” answers exist for items
12
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and stimuli such as those contained within the MEIS. Therefore, Mayer et al.
recommend a group consensus-based approach to scoring the MEIS.
Emotional Intelligence and Other Similar Theories
Emotional intelligence can also be considered to be closely related to a
subset of Gardner’s theory of multiple intelligences, the personal intelligences
(Gardner, 1993). Gardner’s multiple intelligences include four categories:
kinesthetic, linguistic, musical, and personal intelligence. Personal intelligence is
further divided into two subcategories: interpersonal and intrapersonal intelligence.
El falls into the category of personal intelligences but it is the latter subcategory
with which emotional intelligence overlaps most. Gardner (1999) defines
intrapersonal intelligence as the “capacity to understand oneself, to have an effective
working model of oneself—including one’s own desires fears, and capacities—and
to use such information effectively in regulating one’s own life” (p.43). Gardner,
however, believes that ability measures of intelligence provide only an indirect
measure of the actual internal processes within the brain. But at this time relatively
little is known about the human brain structure that empirical tests of his theory do
not yet exist.
As mentioned above, Salovey and Mayer (1997) replaced social intelligence
with emotional intelligence in the general triad of intelligences. Emotional
intelligence and social intelligence are closely related and have often been confused
with one another. Thorndike (1920) first defined social intelligence as the ability to
understand and manage people and to act wisely in human relations. Recently, the
13
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
definition of social intelligence was expanded to include an understanding of a
social situation and evaluating social action (Kaukiainen, et al., 1999). Kaukiainen,
et al. asserted that a “socially intelligent individual shows behavioral flexibility and
is able to change his/her behavior depending on the circumstances of the situation”
(p.82). In other words, a socially intelligent individual knows how to assess a
situation and modify his behavior depending on what is needed in that particular
situation. These definitions of social intelligence make it clear that it is distinctly
different from El in that El is a mental ability (e.g., figuring out one’s own and
other’s emotions) as opposed to being more of a set of social skills (e.g., being
sociable).
Is Emotional Intelligence Related to Tobacco Use?
High levels of emotional intelligence have been found to be negatively
related to correlates of smoking. Mayer, Carlsmith and Chabot (1998) found that
higher scores on the MEIS were negatively associated with lower levels of engaging
in violent behavior. An unpublished manuscript by Rubin (1999) found a negative
association between emotional intelligence and peer ratings of one’s aggressiveness
among adolescents. Given these findings, it is conceivable that high emotional
intelligence could also be negatively related to smoking risk factors. High El may
increase one’s resistance to smoking influences by facilitating the identification of
feelings regarding tobacco use and tobacco offers, and the detection of unwanted
peer pressure.
14
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
To date, only one empirical study has been conducted relating emotional
intelligence to tobacco use (Trinidad & Johnson, 2000). In a study of 205
multiethnic adolescents using an adolescent version of the Multifactor Emotional
Intelligence Scale, Trinidad and Johnson produced the first empirical results
showing a significant negative association between emotional intelligence tobacco
and alcohol use. Specifically, emotional intelligence was found to be negatively
correlated with ever trying a cigarette, smoking in the last 30 days, smoking daily,
smoking weekly, drinking alcohol in the last week, and drinking more than 2
alcoholic drinks within 2 hours in the last 30 days. When the effect of academic
inclination was removed from the correlations, all previously statistically significant
correlations between emotional intelligence tobacco and alcohol use remained, with
the exception of number of days smoked in the last 30 days, which approached
significance (r=-0.14; p=0.0514). Regression models controlling for age, gender
and academic inclination revealed that all associations remained significant. These
results provide empirical evidence that emotional intelligence is a protective factor
against tobacco and alcohol use behaviors and show that emotional intelligence may
account for a certain percentage of the variance in these health-compromising
behaviors.
How Might Emotional Intelligence Related to Adolescent Tobacco Use?
Trinidad and Johnson (2000) argue that it is plausible that the adolescents in
their sample with high emotional intelligence possessed a greater mental ability to
identify their feelings, read others well and understand and detect unwanted peer
15
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
pressure (i.e., understand others’ and manage their own emotions). These abilities
may then have led to an increased resistance to tobacco and alcohol use influences,
thus explaining the negative correlations that were found.
Coupled with a social influences-based tobacco use prevention program, those with
high levels of El may be better able and prepared to process and utilize the
information learned in such prevention programs and thus better benefit from the
associated interventions targeting social influences.
The following four chapters describe results from analyses of associations
between El and psychosocial smoking risk factors (chapter 2), interactions between
El and psychosocial risk factors on smoking intentions (chapter 3), El and ethnicity
interactions on smoking intentions (chapter 4), and El and US acculturation
interactions on perceived social consequences of smoking (chapter 5). Data for
these analyses are from 416 actively consented sixth-grade adolescents from public
middle schools in the greater Los Angeles area. These were a subset of students
from the control condition of the baseline portion of a longitudinal school-based
experimental trial of smoking prevention strategies in a multicultural, urban
population of adolescents in Southern California. These analyses hope to provide
further information on the relationship between adolescent smoking and emotional
intelligence.
16
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
Bar-On, R. (1997). The Emotional Quotient Inventory (EQ-i): Technical
Manual. Toronto: Multi-Health Systems.
Bickel, W. K., Odum, A. L., & Madden, G. J. (1999). Impulsivity and
cigarette smoking: delay discounting in current, never, and ex-smokers.
Psvchopharmacology. 146. 447-454.
Chen, K., & Kandel. D. B. (1995). The natural history of drug use from
adolescence to the mid-thirties in a general population sample. American Journal of
Public Health, 85. 41-47.
Choi, W. S., Pierce, J. P., Gilpin, E. A., Farkas, A. J., & Berry, C. C. (1997).
Which adolescent experimenters progress to established smoking in the United
States? American Journal of Preventive Medicine.13. 385-391.
Conrad, K. M., Flay, B. R., & Hill, D. (1992). Why children start smoking
cigarettes: Predictors of onset. British Journal of Addiction, 87, 1711-1724.
Davies, M., Stankov, L., & Roberts, R.D. (1998). Emotional intelligence:
In search of an elusive construct. Journal of Personality and Social Psychology, 75,
989-1015.
Hansen, W. B., & Graham, J. W. (1991). Preventing alcohol, marijuana,
and cigarette use among adolescents: peer pressure resistance training versus
establishing conservative norms. Preventive Medicine, 20. 414-430.
Gardner, H. (1983). Frames of Mind (10th Anniversary Edition). New
York: Basic Books.
Gardner, H. (1999). Intelligence Reframed: Multiple Intelligences for the
21st Century. New York: Basic Books.
Goleman, D. (1995). Emotional Intelligence. New York: Bantam Books.
Griffin, K. W., Botvin, G. J., Doyle, M. M., Diaz, T., & Epstein, J. A.
(1999). A six-year follow-up study of determinants of heavy cigarette smoking
among high-school seniors. Journal of Behavioral Medicine. 22. 271-284.
Jessor, R. (1984). Adolescent development and behavioral health. InJ. D.
Matarazzo, S. M. Weiss, J. A. Herd, N. E. Miller, & S. M. Weiss (Eds.), Behavioral
Health, A Handbook of Health Enhancement and Disease Prevention (pp.69-90).
New York: John Wiley & Sons.
17
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Johnston, L.D., O'Malley, P.M., & Bachman, J.G. (Dec. 1999). Cigarette
smoking among American teens continues gradual decline. University of Michigan
News and Information Services: Ann Arbor, MI.[On-line], Available:
www.monitoringthefuture.org; accessed 05/23/00.
Kaukiainen, A., Bjorkqvist, K., Lagerspetz, K., Osterman, K., Salmivalli, C.,
Rothberg, S., & Ahlbom, A. (1999). The relationships between social intelligence,
empathy, and three types of aggression. Aggressive Behavior, 25. 81-89.
Lee, L. L., Gilpin ,E. A., & Pierce, J. P. (1993). Changes in the patterns of
initiation of cigarette smoking in the United States: 1950, 1965, and 1980. Cancer
Epidemiology, Biomarkers & Prevention, 2, 593-597.
Leeper, R. (1948). A motivational theory of emotions to replace “emotions
as disorganized response.” Psychological Review, 55, 5-21.
Legree, P. J. (1995). Evidence for an oblique social intelligence factor
established with a likert-based testing procedure. Intelligence, 21, 247-266.
MacKinnon, D. P., Johnson, C. A., Pentz, M. A., Dwyer, J. H., Hansen, W.
B., Flay, B. R., & Wang, E. Y. (1991). Mediating mechanisms in a school-based
drug prevention program: first-year effects of the Midwestern Prevention Project.
Health Psychology, 10, 164-172.
Mayer, J.D., Carlsmith, K.M., & Chabot, H.F. (1998). Describing the
person’s external environment: Conceptualizing and measuring the Life Space.
Journal of Research in Personality, 32, 253-296.
Mayer, J.D., Caruso, D.R., & Salovey, P. (1999). Emotional intelligence meets
traditional standards for an intelligence. Intelligence. 27. 267-298.
Mayer, J. D., Caruso, D. R., & Salovey, P. (1997). Multifactor Emotional
Intelligence Scale, Student Version. Durham, New Hampshire: Authors.
Mayer, J.D., & Salovey, P. (1993). The intelligence of emotional
intelligence. Intelligence, 17, 433-442.
Mayer, J.D., & Salovey, P. (1997). What is emotional intelligence? In P.
Salovey & D. Sluyter (Eds). Emotional Development and Emotional Intelligence:
Implications for Educators (pp.3-31). New York: Basic Books.
18
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Mayer, J. D., Salovey, P., Caruso D. (2000). Models of emotional
intelligence. In R.J. Sternberg (Ed.). Handbook of Intelligence. Cambridge, UK:
Cambridge University Press.
Noland, M.P., Kryscio, R.J., Riggs, R.S., Linville, L.H., Ford, V.Y., &
Tucker, T.C. (1998). The effectiveness of a tobacco prevention program with
adolescents living in a tobacco-producing region. American Journal of Public
Health. 88. 1862-1865.
Patton, G. C., Carlin, J. B., Coffey, C., Wolfe, R., Hibbert, M., & Bowes, G.
(1998). The course of early smoking: a population-based cohort study over three
years. Addiction, 93, 1251-1260.
Rubin, M.M. (1999). Emotional intelligence and its role in mitigating
aggression: A correlational study of the relationship between emotional intelligence
and aggression in urban adolescents. Unpublished manuscript, Immaculata College,
Immaculata, PA.
Salovey, P. & Mayer, J.D. (1990). Emotional intelligence. Imagination,
Cognition, and Personality, 9, 185-211.
Sternberg, R. J. (1997). The concept of intelligence and its role in lifelong
learning and success. American Psychologist, 5 2 ,1030-1045.
Sussman, S., Lichtman, K., Ritt, A., & Pallonen, UE. (1999). Effects of
thirty-four adolescent tobacco use cessation and prevention trials on regular users of
tobacco products. Substance Use & Misuse, 34, 1469-1503.
Thorndike, E. (1920). Intelligence and its uses. Harper’s Magazine, 140,
227-235.
TIME (October 2, 1995). [Cover]. New York: Time Warner.
Trinidad, D.R. & Johnson, C.A. (2002). The association between
emotional intelligence and early adolescent tobacco and alcohol use. Personality
and Individual Differences, 32, pp.95-105.
Wills, T. A., Vaccaro, D., & McNamara, G. (1994). Novelty seeking, risk
taking, and related constructs as predictors of adolescent substance use: an
application of Cloninger’s theory. Journal of Substance Abuse, 6, 1-20.
19
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 2:
The protective association of emotional intelligence with
psychosocial smoking risk factors for adolescents
This study was supported by Grant Number 9DT-0174 from the Tobacco
Related Diseases Research Program of the University of California, Office of the
President, and Grant Number 5 P5 CA 84735 from the National Cancer Institute.
All correspondence regarding this manuscript should be directed to Dennis R.
Trinidad, Institute for Health Promotion and Disease Prevention Research and
Department of Preventive Medicine, Keck School of Medicine, University of
Southern California, 1000 S. Fremont Ave., Unit 8, Alhambra, CA 91803. Phone:
626-457-4163. Fax: 626-457-4012. Email:dtrinida@usc.edu.
Abstract
Though previous work has explored the direct association between
emotional intelligence (El) and adolescent smoking, its relation to psychosocial risk
factors for smoking has not yet been explored. Psychosocial risk factors of interest
include negative social consequences of smoking (SC) because it is generally
targeted by social-influences based tobacco prevention programs, and refusal self-
efficacy (RSE) and intentions to smoke as both variables have been shown in past
research to be predictors of future smoking behaviors. El is defined as the ability to:
20
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
accurately perceive, appraise, and express emotion; access and/or generate feelings
in facilitating thought; understand emotion and emotional knowledge; and regulate
emotions. El was assessed with a shortened version of the Multifactor Emotional
Intelligence Scale, Adolescent Version, and was administered to 416 6th graders
(53% girls) from middle schools in the Los Angeles area (mean age=l 1.3 yrs; 32%
Latino, 29% Asian/Pacific Islander, 13% White, 19% Multiethnic, 6% Other).
Regression models controlling for age, gender, ethnicity, acculturation, SES,
grades received in school, perceived social norms of smoking, and perceived peer
attitudes toward smoking indicate that El is protective against SC, RSE and
smoking intentions. Specifically, those scoring in the highest El quartile perceived
greater negative social consequences associated with smoking than those in the
other quartiles (p<0.05). Those in the highest El quartile were also more efficacious
in refusing cigarette offers relative to those in the lowest quartile (p<0.05).
Relative to those in the lowest El quartile those in the 2n d highest El quartile were
less likely to intend to smoke in the next year (OR=0.22; 90% C.I.: 0.06-0.86;
p-0.03).
Though results indicate that high El is a protective factor for these smoking
precursors in adolescents and should be considered when designing prevention
programs, further analyses need to be conducted to better understand the role of El
in the relationship between psychosocial risk factors and smoking in adolescents.
Future adolescent smoking prevention programs may be improved by incorporating
aspects of emotional intelligence.
21
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Introduction
Evidence for emotional intelligence as a correlate of success in various
domains of life is becoming abundant, yet a dearth of literature exists exploring its
relationship with health behaviors such as smoking and its precursors. In this paper,
we propose and test the hypothesis that emotional intelligence (El), or the
interaction of emotions with cognition, may be protective against psychosocial risk
factors for smoking in adolescents. As adolescent smoking prevention programs
evolve, new methods of increasing their effectiveness via the identification and
targeting of novel protective factors becomes important (Lee, Gilpin & Pierce,
1993). Emotional intelligence may be such a factor.
Two differing models conceptualizing El exist: a mental ability model and a
mixed model. The mental ability model is a new construct that focuses primarily on
emotion and its interactions with cognition (Mayer, Caruso & Salovey, 2000). This
model of El is conceptualized as the “ability to perceive accurately, appraise, and
express emotion; the ability to access and/or generate feelings when they facilitate
thought; the ability to understand emotion and emotional knowledge; and the ability
to regulate emotions” (Mayer & Salovey, 1997, p. 10). Mixed models of El differ
from ability models in that mixed models combine both mental abilities with
traditional traits, such as optimism, motivation or mood (Goleman, 1995; Bar-On,
1997). The mental ability model is a more focused conceptualization emphasizing
mental skills, as opposed to variables already known to be associated with smoking,
such as optimism and mood (Carvajal, Wiatrek, Evans, Knee & Nash, 2000; Defino,
22
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Jamner & Whalen, 2001; Pallonen, Prochaska, Velicer, Prokhorov & Smith, 1998).
Therefore, this research will utilize the mental ability model because it will provide
an opportunity to expand our understanding of adolescent smoking and its
precursors.
Past research has found adolescent smoking to be related to violent behavior
and anti-social behavior (DuRant, Treiber, Goodman & Woods, 1996; Patton,
Barnes & Murray, 1997). High El was also found to be negatively related to these
smoking correlates (Mayer, Carlsmith & Chabot, 1998; Rubin, 1999). Mayer and
his colleagues (1998) found that higher El scores were negatively associated with
lower levels of engaging in violent behavior. Rubin’s work (1999) revealed a
negative association between El and peer ratings of one’s aggressiveness among
adolescents. Given these findings, it is conceivable that high El could also be
negatively related to other smoking correlates and risk factors, such as psychosocial
predictors of smoking.
The only scientific research to date exploring El and smoking in adolescents
provides evidence for El as a protective factor against smoking behaviors in a
sample of 7th and 8th grade students (Trinidad & Johnson, 2002). El was inversely
related to smoking behaviors and accounted for eleven percent of the variance after
controlling for potential confounding variables including age, gender and grades
received in school. Though this previous work explored El’s direct association with
adolescent smoking, it did not examine its relation to psychosocial factors that
potentially lead to smoking. It is plausible that adolescents with high El may also
23
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
be at lower risk for psychosocial smoking precursors. Understanding the link
between El and smoking precursors would provide a more thorough understanding
of the smoking uptake process and potentially identify whether El should be
considered in interventions aimed at psychosocial risk factors, smoking behaviors,
or both.
The current research was undertaken to build upon and improve on previous
work in the field of El and health-related behavior, specifically adolescent smoking.
The goal of the current project is to determine whether El may be a protective factor
against certain psychosocial predictors of adolescent smoking. It is intriguing that
adolescent cigarette smoking continues to be a major public health concern despite
several psychosocial risk factors being fairly well established and targeted for
intervention (Lee, Gilpin & Pierce, 1993; Welte, Barnes, Hoffman & Dintcheff,
1999). Recent research literature indicates those who perceive few negative social
consequences associated with smoking are more likely to smoke in the future
(Dalton, Sargent, Beach, Bernhardt & Stevens, 1999; Distefan, Gilpin, Choi &
Pierce, 1998), as are those with low refusal self-efficacy for cigarette offers (Botvin,
Griffin, Diaz, Miller, Ifill-Williams, 1999; Epstein, Williams, Botvin, Diaz, & Ifill-
Williams, 1999). Similarly, smoking intentions, a component of smoking
susceptibility, or an absence of absolute commitment to not smoke, is a major factor
related to future adolescent smoking (Pierce, Choi, Gilpin, Farkas, & Merritt, 1996;
Unger, Johnson, Stoddard, Nezami & Chou, 1997). So, for example, if a student
perceives few negative consequences associated with smoking, is less able to refuse
24
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
tobacco offers, and/or is not absolutely committed to abstaining from smoking, his
or her risk for smoking increases. Emotional intelligence may be a novel variable
that is protective against these psychosocial risk factors, and thus indirectly reduces
smoking (we present our rationale below). Factors indirectly linked to adolescent
smoking need to be considered in order to augment existing prevention knowledge
and efforts.
Increasing perceptions of negative social consequences associated with
smoking are generally targeted by social-influences based tobacco prevention
programs (Dalton, et al., 1999; Distefan, et al., 1998). Those with high El may
more accurately perceive greater negative social consequences associated with
smoking. They may view smoking as a characteristic of more delinquent behavior
as opposed to being a more grown up behavior. Adolescents with high El may
perceive that smokers do not have more friends, or the friends that smokers have are
not the friends that they would want for themselves. These students may also more
likely associate smoking with negative emotional states/consequences that result
from its social costs.
High refusal self-efficacy is associated with a decrease in smoking risk
(Botvin, et al., 1999; Epstein, et al., 1999). Emotional intelligence maybe related to
high refusal self-efficacy for cigarette offers. Those with high El may be better able
to refuse cigarette offers by understanding more clearly their own feelings about
smoking. This greater understanding may contribute to feeling more confident in
25
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
articulating a potential refusal response that does not hurt one’s social position or
the feelings of the refused person.
Recent research on the risk associated with tobacco use experimentation has
clearly shown that those who are susceptible to smoking are at great risk of
initiating smoking (Choi, Peirce, Gilpin, Farkas & Berry, 1997; Pierce, et al., 1996;
Unger, et al., 1997). Those with high El may be more comfortable about making
absolute, committed decisions not to smoke in the future because they may be able
to better identify, understand and manage unwanted peer pressures to smoke and
thus be less susceptible to smoking. Their ability to manage emotions means they
may be able to employ a wider array of coping strategies to deal with situations that
may increase smoking risk. For example, they may be able deal with the inner
conflict that could result from the decision to not smoke in a party where several
popular peers are smoking, or they may be more comfortable and able to talk about
their feelings regarding the situation to friends. This ability to manage emotions
resulting from explicit or implicit peer pressures to smoke may decrease one’s
smoking intentions.
Since El research is still in its infancy, currently very little published
literature exists regarding its relationship with, and its influence on, adolescent
smoking risk factors. It is our hope that this work will positively add to the body of
knowledge about the risk factors for adolescent smoking and whether El is
associated with these factors. This knowledge will provide insight in developing
more effective adolescent tobacco use prevention programs.
26
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Methods
The data described in this article are from the baseline portion of a
longitudinal school-based experimental trial of smoking prevention strategies in a
multicultural, urban population of adolescents in Southern California. The purpose
of the baseline survey was to assess tobacco use and related psychosocial and
cultural variables before the implementation of culturally relevant smoking
prevention programs. A measure of emotional intelligence was administered to a
subset of students from the control condition of the main prevention trial. Analyses
were performed on this subset of students.
Subjects
Subjects were 416 actively consented 6th grade adolescents from public
middle schools in the greater Los Angeles area. This sample population had a mean
age 11.3 years and was 53% male. Table 2 summarizes the age and gender
characteristics of the study participants. The ethnic distribution was 32% Latino,
29% Asian/Pacific Islander, 13% White, 19% Multiethnic, and 6% Other. Students
were considered Latino if they answered "yes" to the question, "Are you
Hispanic/Latino?" Similar coding schemes were utilized for Asians and Whites.
Due to the small number of African Americans and other ethnicities, students who
were not Latino, Asian or White were considered Other. Students who were more
than one ethnicity, including Other, were coded as Multiethnic.
27
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 2. Age & Gender Characteristics
Frequency (%)
Gender
Male 194 (46.8)
Female 221 (53.2)
Age
10 3 (0.7)
11 300 (72.5)
12 110 (26.6)
13 1 (0.2)
n=416 but had 1 missing gender and 2 missing age
Procedure
To assess demographics and tobacco related behaviors and attitudes,
students completed the baseline questionnaire, a 160-item paper-and-pencil survey,
in their classrooms during a single class period (45-50 minutes). Trained data
collectors, who were not previously acquainted with the students, distributed the
surveys. The surveys were identified only by a code number, not with the students’
names or any other identifying information. Because the students all were attending
English-language schools in which their classes were conducted only in English
(California state law prohibits bilingual education in public schools), a basic level of
English-language proficiency was assumed and the surveys were provided only in
English. However, students were encouraged to ask the data collectors to clarify the
meanings of any unfamiliar words.
Emotional intelligence was measured during another classroom session
several weeks after the baseline survey. Students completed the emotional
intelligence scale individually with the lead author reading each item, including
28
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
examples, out loud to the students. This allowed for the maintenance of uniformity
in survey administration and the minimization of differences in participation and
reading skills. Students were informed that the emotional intelligence survey was
part of the larger study, which began at the first data collection session, examining
culture and health.
Analyses described in this study represent the sample of students present at
both data collection sessions. Active parental consent was obtained several days
prior to the first data collection session. Student assent was obtained at each data
collection session.
Measures
Emotional Intelligence.
The Multifactor Emotional Intelligence Scale, Adolescent Version (MEIS;
Mayer, Salovey & Caruso, 1997) was employed to assess the El of the participants.
A thorough description of this test and how it is scored can be found in Mayer,
Caruso & Salovey (1999). Briefly, the MEIS is a competence-based measure
consisting of 4 branches assessing an individual’s ability to perceive, assimilate,
understand, and manage emotion in himself/herself and in others. In this study, the
MEIS was shortened by excluding the Assimilating Emotions branch (assessing
one’s ability to generate emotions) and including only the Emotional Identification,
Understanding Emotions, and Managing Emotion branches. The Assimilating
Emotions branch was excluded due to time restrictions at schools and because in
previous work we discovered that this was a particularly difficult task to
29
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
comprehend for those with limited English proficiency or English as a second
language. Exploratory factor analysis revealed that this abbreviated measure of El,
consisting of these 3 branches of El, still formed an overall El factor similar to
Mayer et al.’s (1999) overall El model and one used in previous work by Trinidad
and Johnson (2002). A critical evaluation of the overall El construct, as measured
by the MEIS, conducted by Ciarrochi, Chan, and Caputi (2000) showed the MEIS to
be a valid measure of El.
The Emotional Identification branch was assessed with the Faces, Music,
and Stories subtests; the Designs subtest was excluded. Each of these subtests
presented a stimulus to the student (e.g., a picture of a face depicting certain
emotions, a musical selection, or a short vignette) and asked him/her to rate whether
a specific emotion was present in each stimulus. Possible responses were rated on a
5-point Likert-type scale, ranging from “Definitely Not Present” (1) to “Definitely
Present” (5).
The Understanding Emotions branch was assessed with the Relativity
subtest. This test consisted of items describing a conflicting social situation
between two persons. Students were asked to estimate how likely each character
experienced certain feelings. Possible responses were rated on a 5-point Likert-type
scale, ranging from “Extremely Unlikely” (1) to “Extremely Likely” (5).
The Managing Emotions branch was assessed with the Managing subtest.
This test assessed how well students were able to manage their own and others’
emotions by asking them to evaluate the effectiveness of certain courses of action
30
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
given a hypothetical situation. Possible responses were rated on a 5-point Likert-
type scale, ranging from “Extremely Ineffective” (1) to “Extremely Effective” (5).
Consensus scoring was utilized in grading the MEIS. This is the standard
scoring method based on norms of reference groups studied by Mayer et al. (1999).
Since Latino and Asian students were underrepresented in Mayer et al.’s sample
used to create reference norms, we computed norms based on the current sample.
Following the procedure described by Mayer et al., each student’s response was
weighted by the proportion of the sample giving the same response. The resulting
score indicated the percent agreement with the population from which the subject
was drawn. For example, if 75% of the students sampled in this study reported that
happiness was “Definitely Present” in a picture of a face (“5” on the Likert-scale),
then a student who chose “5” would receive a score of 0.75 for the item. If 2%
answered “Definitely Not Present” (“1” on the scale) then that student would
receive a score of 0.02 for the item. Sub-totals for each branch were summed.
Branch totals were then summed to obtain a total El score. Branch scores and the
aggregated El scores were corrected for the number of items comprising each score.
Total El scores were then divided into quartiles.
Psychosocial Risk Factors for Smoking.
Measures of psychosocial risk factors for smoking were contained within the
baseline tobacco use survey. Dependent measures of interest were perceived social
consequences of smoking, refusal self-efficacy, and intentions to smoke in the next
year.
31
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Perceived social consequences of smoking (SC) was an index score
comprised of eight variables measuring a student’s perception of the social
outcomes of smoking (alpha=0.66). Variables included items such as “Smoking
cigarettes is one way to lose friends who are nonsmokers”, “Kids who smoke are
more grown up” and “Smoking cigarettes makes kids look cool.” Possible
responses for each of the eight variables were rated on a 4-point Likert-type scale
with high scores indicating higher negative social consequences (i.e., high scores are
more desirable). The range of responses was 11 to 32.
Refusal self-efficacy (RSE) was assessed by asking the question, “If
someone you just met offered you a cigarette and you did not want it, how easy or
hard would it be to say ‘no’?” Possible responses were rated on a 4-point Likert-
type scale with high scores corresponding to greater refusal self-efficacy (e.g.,
“Very easy”) and low scores corresponding to less refusal self-efficacy (e.g., “Very
hard”).
Intention to smoke in the next year next year (yes/no) was measured using a
rationale derived from Pierce, et al.’s (1996) method of assessing smoking
susceptibility. Students who answered “No, definitely not” to the question, “At any
time in the next year (12 months) do you think you will smoke a cigarette?” were
classified as not intending to smoke in the next year, while those who answered
“Yes”, “Maybe yes”, and “Maybe no” were classified as susceptible.
32
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Covariates.
Relevant covariates included several demographic variables, including age,
gender, ethnicity, socioeconomic status (ratio of number of people per room in a
home), linguistic acculturation (5-point Likert-type scale of language spoken at
home ranging from "English only" to "Only another language"), self-reported
assessment of overall grades received in school last year (e.g., mostly A’s, mostly
A’s & Bs, mostly B’s, etc.), perceived peer attitudes toward smoking, and perceived
social norms of smoking. Self-reported grades was included in analyses as a proxy
for IQ. Previous research has shown El to be related to IQ (Mayer, et al., 1999).
Perceived social norms of smoking was assessed by asking students to
identify the approximate number, out of 100 of similarly aged students, who smoked
cigarettes in the past month (e.g., 10, 20, 30, etc.). Perceived peer attitudes toward
smoking was measured by asking students the question, “Do most people your age
think it’s ok to smoke cigarettes once in a while?” Possible responses were rated on
a 4-point Likert type scale with high scores indicating that peers think it is ok to
smoke cigarettes once in a while and low scores indicating otherwise. Both of these
variables were related to the dependent variables and thus were controlled for in
statistical analyses.
Statistical Analyses
All statistical analyses were performed using SAS version 8.1 (SAS
Institute, 2000). For the examination of the relationship between El and its
component branches and psychosocial risk factors for smoking, regression analyses
33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
were performed. Specifically, for dependent variables with a continuous
distribution, SC and RSE, general linear models were fit. For intentions to smoke in
the next year, a dichotomous outcome, a logistic regression model was fit. All
models controlled for age, gender, ethnicity, acculturation, socioeconomic status,
grades received in school, perceived social norms of smoking, and perceived peer
attitudes toward smoking.
Results
Correcting for the number of items, mean score on the MEIS was 0.36
(sd=0.06). Mean score on the Emotional Identification branch was 0.39 (sd=0.07);
for the Understanding Emotions branch, 0.31 (sd=0.07); and for the Managing
branch, 0.24 (sd=0.04). Cronbach’s coefficient alpha’s for the MEIS was 0.75, for
the Emotional Identification subscale 0.83, for the Understanding Emotions
subscale 0.66, for the Managing subscale 0.76.
Mean score on the MEIS for girls was significantly higher than for boys
(0.37 vs. 0.35; p=0.03). Girls also scored higher on the Emotional Identification
Branch (0.40 vs. 0.38; p=0.02). Girls reported receiving higher grades in school
than boys (p<0.01). A significantly larger proportion of boys were more intended to
smoke in the next year compared to girls (11.5% vs 3.7%; p<0.01). No other
significant differences by gender were found.
Of all students sampled, 32.8% unequivocally viewed negative social
consequences associated with smoking (i.e., 32/32 on the SC index); 60.5% reported
34
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
it would be “Very easy” to refuse a cigarette offer from someone they just met; and
7.5% intended to smoke in the next year. Mean score for the social consequences
index variable was 29.07 (sd=3.59). Mean score for the refusal self-efficacy
variable was 3.24 (sd=T.06).
General linear models controlling for age, gender, ethnicity, acculturation,
socioeconomic status, grades received in school, perceived social norms of
smoking, and perceived peer attitudes toward smoking results indicate that those in
the highest El quartile perceived greater negative social consequences associated
with smoking than those in the other quartiles (Mean: 30.18 vs. 28.71,29.19 &
28.62, respectively; p<0.05). Those in the highest El quartile were also more
efficacious in refusing cigarette offers relative to those in the lowest quartile (Mean:
3.41 vs. 3.05; p<0.05). Logistic regression models revealed that relative to those in
the lowest El quartile those in the 2n d highest El quartile were less likely to intend to
smoke in the next year (OR=0.22; 90% C.I.: 0.06-0.86; p=0.03). These results are
summarized in Tables 3 and 4, and illustrated Figures 1-3.
Table 3. Mean Social Consequences and Refusal
Self-Efficacy Scores by El Quartile
El
Quartile
Social
Consequences
Refusal
Self-Efficacy
Qi
28.62 3.05
Q2 29.19 3.23
Q3 28.71 3.34
Q4 30.18 3.41
Range: 11-32 Range: 1-4
Note: Adjusted for age, gender, SES, language spoken at home, ethnicity, self-reported grades in
school, perceived social norms, and perceived peer attitudes toward smoking.
35
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4. Odds Ratios for Smoking
Intentions by El Quartile
El Quartile OR 90% C.I.
P
Ql
1 - -
Q2 0.92 0.37-2.30 0.44
Q3 0.22 0.06-0.86 0.03
Q4 0.41 0.14-1.22 0.09
Note: Adjusted for age, gender, SES, language spoken at home, ethnicity, self-reported grades in
school, perceived social norms, and perceived peer attitudes toward smoking
Figure 1. Perceived Negative Social Consequences by El Quartile
30.5
30
> 5
-o
a
d1
m
a
29
(38.5
u
o
C Z 3
28
27.5
1st Quartile 2nd Quari|lf Quart ? j@ ^ Quartile 4th Quartile
Note: Adjusted for age, gender, SES, language spoken at home, ethnicity, self-reported
grades in school, perceived social norms, and perceived peer attitudes toward smoking.
36
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 2. Refusal Self-Efficacy by El Quartile
3.5
im
©
CJ
C /5
3.4
>>
CJ
OS
3.3
CJ
IS
w
3.2
< 4 H
8
C /5
3.1
" 3
3
3
©
0 4
2.9
2.8
1st Quartile 2nd Quartile 3rd Quartile
El Quartile
4th Quartile
Note: Adjusted for age, gender, SES, language spoken at home, ethnicity, self-reported
grades in school, perceived social norms, and perceived peer attitudes toward smoking.
Figure 3. Odds Ratios for Intentions to Smoke by El Quartile
2
1.8
& 1 *
0 1*6
£ 1.4
1 1.2
1 1
1 0.8
'■ § 0.6
0 >
E 0.4
h H
0.2
0
Note: Adjusted for age, gender, SES, language spoken at home, ethnicity, self-reported
grades in school, perceived social norms, and perceived peer attitudes toward smoking.
— , 1 1 ........
1st Quartile 2nd Quartile 3rd Quartile 4th Quartile
El Quartile
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Because the continuous dependent outcomes were not normally distributed
(SC and RSE), to further validate findings, both variables were dichotomized and
logistic regression analyses were performed. When SC was dichotomized into those
viewing absolute negative social consequences associated with smoking versus
others, the odds of perceiving absolute negative social consequences associated with
smoking for those in the second highest and the lowest El quartiles were
approximately one-half of to those in the highest El quartile (OR=0.36, 95% Cl:
0.18-0.71; and OR=0.42, 95% Cl: 0.22-0.84, respectively). When refusal self-
efficacy was dichotomized into those absolutely confident in refusing versus others,
those in the lowest El quartile were statistically significantly less able to refuse a
cigarette offer than those in the highest El quartile (OR=T.9, 95% Cl: 1.01-3.59).
Therefore, in line with our findings when the continuously-distributed dependent
variables were dichotomized, those with high El were more likely to perceive
negative social consequences associated with smoking and also more able to refuse
a cigarette offer.
Additional, separate general linear and logistic regression models were fit
with the subscales of El as the independent variable and the aforementioned
psychosocial risk factors as the dependent variables. All models controlled for the
same covariates as did the models described above. Results show that those scoring
in the highest quartile of the Emotional Identification branch were more likely to
perceive greater negative social consequences of smoking than those in the lowest
quartile (Mean: 29.64 vs. 28.57; p<0.05). Those who scored in the highest quartile
38
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
of the Understanding Emotions branch of El perceived greater negative social
consequences associated with smoking than those in the two lowest El quartiles
(Mean: 29.85 vs. 28.88 & 28.42; p<0.05). Those in the highest quartile of the
Understanding Emotions branch were also more efficacious in refusing cigarette
offers relative to those in the second highest and lowest quartiles (Mean: 3.51 vs.
3.12 & 3.17, respectively; p<0.05). Students in the highest quartile of the
Understanding Emotions branch were also at lower risk for intending to smoke in
the next year (OR=0.12; 90% C.I.: 0.02-0.72; p=0.03). Those in the highest quartile
of the Managing Emotions branch were more likely to perceive negative social
consequences of smoking than those in the lowest quartile (Mean: 29.77 vs. 28.77;
p<0.05). Those who scored in the second highest quartile of the Managing
Emotions branch were more able to refuse cigarette offers than those who scored in
the lowest quartile (Mean: 3.37 vs. 3.03; p<0.05).
Discussion
Our findings support the growing body of literature that emotional
intelligence is associated with health-related variables, smoking risk factors in this
case. Results indicate that emotional intelligence is a significant protective factor
against psychosocial risk factors for smoking. Specifically, high El was related to
an increased perception of the negative social consequences of smoking. Those
with high El were more able to refuse cigarette offers, as well as less likely to intend
39
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
to smoke in the next year. These results provide empirical evidence that El is a
protective factor against these psychosocial smoking precursors.
Those with high El perceived greater negative social consequences
associated with smoking. Though non-linear, the trend of our results suggests that
as El level increases so do perceptions of negative social consequences of smoking.
Those in the second highest El quartile appeared to perceive negative social
consequences of smoking at roughly the same level as those in the lowest El
quartile, while those in the second lowest El quartile perceived more consequences
than both of these. Perhaps students scoring in the second highest El quartile were
not as sure of their views on social consequences of smoking as those in the highest
quartile, resulting in less negative perceptions, much like those in the lowest
quartile. Regardless, those in the highest El quartile perceived more negative social
consequences of smoking than those scoring in the other quartiles. A more detailed
examination of the El branches revealed similar and more linear trend. A greater
ability to identify, understand and manage emotions (i.e., scoring high on these
individual branches of El) was statistically significantly associated with perceiving
high negative social consequences associated with smoking. Perhaps students in our
sample who scored high on these subscales were better able to identify, understand
and manage the feelings associated with socially beneficial behaviors through their
own experiences and when observing others, particularly with smoking-related
issues. Therefore they may not have been as likely to view smoking as something
they wanted to do. It is also possible that these students viewed smoking as a
40
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
characteristic of more delinquent behavior instead of more mature behavior and thus
perceived it as a social consequence.
Emotional intelligence was related to high refusal self-efficacy with respect
to cigarette offers. Our findings suggest a dose-response trend that as El level
increases so does the ability to refuse cigarette offers. Students in the highest
quartile of the Understanding Emotions branch were more efficacious in refusing
cigarette offers relative to the third and fourth quartiles. Those in the highest
quartile of the Managing Emotions branch were also more efficacious in refusing
cigarette offers. Students with high El may have been better able to refuse cigarette
offers by understanding more clearly their own feelings about smoking. This,
combined with the increased ability to manage emotions, may contribute to feeling
more confident in articulating a potential refusal response that does not hurt one’s
social position or the feelings of the refused person. These students may also be
more able to understand and manage the potential anxiety associated with having to
refuse an offer.
We hypothesized that those with high El may be more comfortable about
making absolute, committed decisions not to smoke in the future because they may
be able to better identify, understand and manage unwanted peer pressures to smoke
and thus be less susceptible to smoking. This hypothesis was supported by our
findings, albeit in a slightly different fashion. Fitting 1-tailed logistic regression
models (with corresponding 1-tailed, 90% confidence intervals) for this
dichotomous variable, it appears that a quadratic relationship exists between El and
41
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
intentions to smoke in the next year. Specifically, those in the 2n d highest quartile of
El were less likely to intend to smoke in the next year relative to those in the lowest
quartile. Those in the highest quartile of El were not significantly less likely to
smoke in the next year than those in the lowest quartile. Leaving this variable in its
original, continuous form yielded a similar pattern of results. These findings may
suggest that an optimum level of El may be necessary to be protective against
smoking intentions. Initially, it was thought this might be because those in the
highest El quartile might perceive fewer negative social consequences of smoking
and thus be more likely to intend to smoke. However, when this relationship was
tested the path was not found to be significant, perhaps due to a lack of statistical
power. Therefore, it might be the way the El scale is scored (proportion of persons
you agree with in the sample—perhaps an indirect measure of conformity or
collectivism) shows that those who answer most similarly to their peers are more at
risk suggesting that perhaps an ample amount of “independence” from social norms
is ideal to be protective against smoking intentions.
Examining the relationship between the separate ability components (the
branches) of El and the psychosocial risk factors of interest provides additional
information about El's role as a protective factor. Of the three branches, the
Identifying Emotions and Managing Emotions branches were not found to have as
strong negative associations with the outcome variables as did the Understanding
Emotions branch. When the effect of El was examined in greater detail, it appeared
that its associations with the psychosocial risk factors was driven, in relatively large
42
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
part, by the Understanding Emotions branch. Though high scores on the Identifying
Emotions branch, the first branch of El to develop, were associated with lower risks,
this branch did not seem to differentiate between high and low risks, as much as the
Understanding Emotions branch, the second branch of El to develop. Perhaps
higher-level El skills, such as emotional understanding (assessed by the
Understanding Emotions branch), are more important in complex situations such as
the decision-making process regarding social pressures to smoke. High scores on
the Managing Emotions branch, the most advanced El branch, was also associated
with perceiving more negative social consequences of smoking, and played a larger
role in refusal skills than in the other psychosocial risk factors. Intuitively this latter
finding makes sense as the ability to manage one’s emotions is a skill that would
apply more to actual interactions (refusing an offer) than to perceptions (of social
consequences). Therefore, though El is protective against various psychosocial risk
factors for smoking, examining its ability components illustrates that some aspects
contribute more to the associations than others in our sample.
In addition to controlling for demographic variables, our analyses controlled
for grades received in school, perceived social norms of smoking, and perceived
peer attitudes toward smoking. Self-reported grades was included as a covariate in
all of our analyses as a proxy for IQ because previous research has shown El to be
related to IQ (Mayer, et al., 1999). After controlling for self-reported grades, it is
interesting that the associations between El and the psychosocial risk factors
remained significant. Therefore, although intellectual level may play a role in
43
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
reducing smoking risk (Young & Rogers, 1986), El appears to contribute to the
reduction in these risks independently. We controlled for perceived social norms
and perceived peer attitudes toward smoking because these variables were
correlated with our dependent variables of interest.
Emotional intelligence hinges on the conception that certain emotional tasks
or problems have answers that can be judged either correct or incorrect. To obtain
the norms used to score the MEIS consensus scoring was employed. This was done
for several reasons. First, to some extent, emotions are socially based (Mayer &
Salovey, 1997). Second, research has shown that when groups of people’s
judgments are pooled together for items such as these Likert-based MEIS items, the
groups tend to serve as expert judges (Legree, 1995). Third, using various scoring
methods (consensus, expert and target) Mayer, et al. (1999) found that consensus
scoring produced the most reliable measures of EL Mayer, et al.’s expert scoring
involved using answers on the MEIS deemed to be correct by experts in emotion
research. Target scoring involved obtaining correct answers from persons depicted
in each item. For example, for identifying emotions present in a photograph of a
young girl crying (in the Emotional Identification branch), the girl whose photo was
taken was asked what emotions she was feeling. Answers from these three scoring
methods were moderately and highly positively correlated, indicating that indeed a
set of answers that were “more correct” than others exist for items and stimuli such
as those contained within the MEIS.
44
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Future Directions
Though results indicate that high El is a protective factor for these smoking
precursors in adolescents and should be considered when designing prevention
programs, further analyses need to be conducted to better understand the role of El
in the relationship between psychosocial risk factors and smoking in adolescents.
For example, El may moderate the relationship between perceiving negative social
consequences of smoking and one’s intentions to smoke, and may also moderate the
relationship between smoking intentions and smoking experimentation. A clearer
understanding of the role of El in the smoking uptake process may lead to more
effective prevention strategies.
Future work should examine these findings across ethnicities to further
inform about how the associations between El and the psychosocial risk factors vary
across ethnicities, and which of the El branches are more developed at this age for
each ethnicity. Furthermore, examining El in general across various ethnicities may
provide additional valuable information about how ethnicity is related to smoking
behaviors. For example, previous research has shown that Asian adolescents smoke
less than other ethnicities (Chen & Unger, 1999). Understanding how is El different
among Asians versus Latinos, for example, and how such differences might relate to
smoking would then have implications for the tailoring of tobacco prevention
programs.
Likewise the examination of El across various levels of acculturation may be
informative for multicultural smoking prevention programs aimed at diverse
45
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
populations. Past research utilizing a language-based measure of acculturation
revealed that adolescents who spoke more English at home smoked more than those
who spoke another language (Unger, Cruz, Rohrbach, et al., 2000). The
identification of certain aspects of El (mental ability skills) protective against
smoking behaviors at specific levels of acculturation could be used to enhance
multicultural smoking prevention programs.
Limitations
Further studies need to be performed in order to validate the findings of this
research. Causal inferences from this study should be treated with caution because
of its cross-sectional design. Future research regarding El and adolescent smoking
may be conducted in the context of a longitudinal, school-based prevention program
in order to address whether those with high El do actually benefit more from such
programs, and whether a program to increase El would help prevent smoking. For
example, teaching students El skills including how to identify various feelings they
may experience, understand how those feeling may have arisen, and manage those
feelings if needed. Such research would address the temporal nature of the
relationship between El and smoking. The resulting knowledge gained can be used
to develop improved adolescent smoking prevention programs.
Future work may also benefit by utilizing more specific measures of
acculturation than was used in this study. The acculturation process is complex, and
a multi-item scale would have been a more complete measure of acculturation.
Although language usage is only one component of the acculturation process, such
46
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
single-item language measures have been shown to correlate highly with more
comprehensive acculturation scales, accounting for a large proportion of the
variance (Epstein, Botvin, Dusenbury, & Diaz, 1996; Unger et al., 2000).
Therefore, although language-based acculturation measures admittedly do not
capture the richness of the acculturation experience, they tend to be highly
correlated statistically with other dimensions of acculturation and therefore can be
viewed as one of the many manifestations of the acculturation process. Similarly,
our measure of El may not completely reflect all of its aspects, particularly since we
excluded the Assimilating Emotions branch. This branch involved students
generating various emotions and was excluded primarily due to time constraints
within classrooms. Furthermore, in our previous work we discovered that this was a
particularly difficult task to comprehend for those with limited English proficiency
or English as a second language. Though we acknowledge that our abbreviation of
the MEIS may have altered the generally accepted mental ability model of El, factor
analyses performed on our abbreviated version indicated that the abbreviated
version we employed was still very similar to Mayer, et al. (1999). Work by
Sullivan (1999) also utilized a similar abbreviated El measure and found it to be
reflective of the overall MEIS.
Implications & Conclusion
Evidence of the relationship between El and health behaviors is increasing.
With the emerging trend within prevention research towards positive and protective
factors, our findings regarding the protective role of El against smoking risk factors
47
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
are encouraging. The most effective adolescent smoking prevention programs focus
on social influences, such as perceived social consequences of smoking (Hansen &
Graham, 1991; MacKinnon, Johnson, Pentz, et al., 1991). Our findings indicate that
those with high levels of El may be better prepared to process and utilize the
information learned from such social influences-based prevention programs and
thus better benefit from them. Similarly, our findings may be beneficial for
prevention programs that emphasize refusal skills training as those with high El
were better able to refuse cigarette offers. Therefore, adding an El component to
future prevention programs may enhance the targeted effects for social
consequences and refusal skill training. Thus as adolescent smoking prevention
programs evolve, taking into account the novel construct of El may lead to
increased effectiveness.
48
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
Bar-On, R. (1997). The Emotional Quotient Inventory (EQ-il: Technical
Manual. Toronto, Canada: Multi-Health Systems.
Botvin, G.J., Griffin, K.W., Diaz, T., Miller, N., & Ifill-Williams, M.
(1999). Smoking initiation and escalation in early adolescent girls: One-year
follow-up of a school-based prevention intervention for minority youth. Journal of
the American Medical Womens Association. 54. 139-143, 152.
Carvajal, S.C., Wiatrek, D.E., Evans, R.I., Knee, C.R., & Nash, S.G. (2000).
Psychosocial determinants of the onset and escalation of smoking: Cross-sectional
and prospective findings in multiethnic middle school samples. Journal of
Adolescent Health, 27, 255-65.
Chen, X. & Unger, J.,B. (1999). Hazards of smoking initiation among
Asian American and non-Asian adolescents in California: a survival model analysis.
Preventive Medicine, 28, 589-599.
Choi, W. S., Peirce, J. P., Gilpin, E. A., Farkas, A. J., & Berry, C. C. (1997).
Which adolescent experimenters progress to established smoking in the United
States? American Journal of Preventive Medicine,13, 385-391.
Ciarrochi, J. V., Chan, A. Y. C., & Caputi, P. (2000). A critical evaluation
of the emotional intelligence construct. Personality and Individual Differences, 28,
539-561.
Dalton, M.A., Sargent, J.D., Beach, M.L., Bernhardt, A.M., & Stevens, M.
(1999). Positive and negative outcome expectations of smoking: Implications for
prevention. Preventive Medicine, 29, 460-465.
Defino, R.J., Jamner, L.D., & Whalen, C.K. (2001). Temporal analysis of
the relationship of smoking behavior and urges to mood states in men versus
women. Nicotine and Tobacco Research, 3. 235-48.
Distefan, J.M., Gilpin, E.A., Choi, W.S., & Pierce, J.P. (1998). Parental
influences predict adolescent smoking in the United States, 1989-1993. Journal of
Adolescent Health, 22, 466-474.
DuRant, R.H., Treiber, F., Goodman, E., & Woods, E.R. (1996). Intentions
to use violence among young adolescents. Pediatrics, 98, 1104-1108.
49
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Epstein, J.A., Botvin, G.J., Dusenbury, L., & Diaz, T. (1996). Validation of
an acculturation measure for Hispanic adolescents. Psychological Reports. 79, 1075-
1079.
Epstein, J.A., Williams, S.C., Botvin, G.J., Diaz, T., & Ifill-Williams, M.
(1999). Psychosocial predictors of cigarette smoking among adolescents living in
public housing developments. Tobacco Control, 8, 45-52.
Goleman, D. (1995). Emotional Intelligence. New York: Bantam Books.
Hansen, W. B., & Graham, J. W. (1991). Preventing alcohol, marijuana,
and cigarette use among adolescents: peer pressure resistance training versus
establishing conservative norms. Preventive Medicine, 20. 414-430.
Lee, L.L., Gilpin ,E.A., & Pierce, J.P. (1993). Changes in the patterns of
initiation of cigarette smoking in the United States: 1950,1965, and 1980. Cancer
Epidemiology. Biomarkers & Prevention, 2. 593-597.
Legree, P. J. (1995). Evidence for an oblique social intelligence factor
established with a likert-based testing procedure. Intelligence. 21, 247-266.
MacKinnon, D. P., Johnson, C. A., Pentz, M. A., Dwyer, J. H., Hansen, W.
B., Flay, B. R., & Wang, E. Y. (1991). Mediating mechanisms in a school-based
drug prevention program: first-year effects of the Midwestern Prevention Project.
Health Psychology. 10. 164-172.
Mayer, J.D., Carlsmith, K.M., & Chabot, H.F. (1998). Describing the
person’s external environment: Conceptualizing and measuring the Life Space.
Journal of Research in Personality. 32. 253-296.
Mayer, J. D., Caruso D., & Salovey, P. (2000). Models of emotional
intelligence. In R.J. Sternberg (Ed.). Handbook of Intelligence (pp. 396-420).
Cambridge, UK: Cambridge University Press.
Mayer, J. D., Caruso D. R, & Salovey, P. (1999). Emotional intelligence
meets traditional standards for an intelligence. Intelligence, 27, 267-298.
Mayer, J.D., & Salovey, P. (1997). What is emotional intelligence? In P.
Salovey & D. Sluyter (Eds). Emotional Development and Emotional Intelligence:
Implications for Educators (pp.3-31). New York: Basic Books.
Mayer, J. D., Salovey, P., & Caruso, D. R. (1997). Multifactor Emotional
Intelligence Scale, Student Version. Durham, New Hampshire: Authors.
50
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Pallonen, U.E., Prochaska, J.O., Velicer, W.F., Prokhorov, A.V., & Smith,
N.F. (1998). Stages of acquisition and cessation for adolescent smoking: an
empirical integration. Addictive Behaviors, 23, 303-24.
Patton, D., Barnes, D., & Murray, R.P. (1997). A personality typology of
smokers. Addictive Behaviors, 22, 269-273.
Pierce, J.P., Choi, W.S., Gilpin, E.A., Farkas, A.J., & Merritt, R.K. (1996).
Validation of susceptibility as a predictor of which adolescents take up smoking in
the United States. Health Psychology' 15, 355-61.
Rubin, M.M. (1999). Emotional intelligence and its role in mitigating
aggression: A correlational study of the relationship between emotional intelligence
and aggression in urban adolescents. Unpublished manuscript, Immaculata College,
Immaculata, PA.
SAS Institute [Computer software]. (200). Cary, North Carolina: SAS
Institute.
Sullivan, A. K. (1999). The Emotional Intelligence Scale for Children.
Dissertation Abstracts International. Unpublished doctoral dissertation.
Trinidad, D.R. & Johnson, C.A. (2002). The association between
emotional intelligence and early adolescent tobacco and alcohol use. Personality
and Individual Differences, 32, pp.95-105.
Unger, J.B., Cruz, T.B., Rohrbach, L.A., Ribisl, K.M., Baezconde-
Garbanati, L., Chen, X., Trinidad, D.R., & Johnson, C.A. (2000). English language
use as a risk factor for smoking initiation among Hispanic/Latino and Asian
American adolescents in California: Evidence for mediation by tobacco-related
beliefs and social norms. Health Psychology, 19, 403-410.
Unger, J.B., Johnson, C.A., Stoddard, J.L., Nezami, E., & Chou, C.P.
(1997). Identification of adolescents at risk for smoking initiation: validation of a
measure of susceptibility. Addictive Behaviors, 22, 81-91.
Young, T.L., & Rogers, K.D. (1986). School performance characteristics
preceding onset of smoking in high school students. American Journal of Diseases
in Children. 140. 257-259.
Welte, J. W., Barnes, G. M., Hoffman, J. H., & Dintcheff, B. A. (1999).
Trends in adolescent alcohol and other substance use: relationships to trends in peer,
parent, and social influences. Substance Use & Misuse, 34, 1427-1449.
51
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 3:
Emotional intelligence and smoking risk factors in adolescents:
Interactions on smoking intentions
This study was supported by Grant Number 9DT-0174 from the Tobacco
Related Diseases Research Program of the University of California, Office of the
President, and Grant Number 5 P5 CA 84735 from the National Cancer Institute.
All correspondence regarding this manuscript should be directed to Dennis R.
Trinidad, Institute for Health Promotion and Disease Prevention Research and
Department of Preventive Medicine, Keck School of Medicine, University of
Southern California, 1000 S. Fremont Ave., Unit 8, Alhambra, CA 91803. Phone:
626-457-4163. Fax: 626-457-4012. Email:dtrinida@usc.edu.
Abstract
Claims by the popular media of emotional intelligence being the most
important predictor of life success have led to a burgeoning of interest in the field of
emotional intelligence research. Though such claims seem to be overly optimistic,
elucidation of its relationship with specific health behaviors, such as adolescent
smoking, beckons exploration. Of particular interest is the hypothesis that
relationships between smoking intentions and other smoking risk factors may vary
differentially across levels of emotional intelligence (El). Smoking risk factors of
interest include negative social consequences of smoking (SC), refusal self-efficacy
52
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(RSE), ever trying cigarettes, and hostility. El is defined as the ability to: accurately
perceive, appraise, and express emotion; access and/or generate feelings in
facilitating thought; understand emotion and emotional knowledge; and regulate
emotions. El was assessed with a shortened version of the Multifactor Emotional
Intelligence Scale, Adolescent Version, and was administered to 416 6th graders
(53% girls) from middle schools in the Los Angeles area (mean age=11.3 yrs; 32%
Latino, 29% Asian/Pacific Islander, 13% White, 19% Multiethnic, 6% Other).
Logistic regression models were fit controlling for age, gender, ethnicity,
acculturation, SES, grades received in school, perceived social norms of smoking,
and perceived peer attitudes toward smoking. A statistically significant interaction
was found between El and RSE on smoking intentions (p=0.03) in that those with
low El were more likely to intend to smoke if refusal skills were low versus high,
while those with high El did not intend to smoke differentially regardless of refusal
skill level. Ever trying cigarettes interacted with El with regard to smoking
intentions in the next year (p<0.01) with high El adolescents being more likely to
intend to smoke in the next year if they had already tried smoking. Those with high
El who did not intend to smoke in the next year were not likely to have tried
smoking. El interacted with hostility level to affect smoking intentions (p=0.03) in
that for those with low El, high hostility increased the likelihood of intending to
smoke while for those with high El, such a difference was not evident. No
significant interactions were found between El and SC on smoking intentions.
53
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Results indicate that El interacts with risk factors to reduce smoking
intentions and should be considered when designing prevention programs. Those
with high levels of El may be better prepared to process and utilize the information
learned from social influences-based prevention programs and thus better benefit
from them. However, further analyses need to be conducted to better understand the
role of El in the relationship between psychosocial risk factors across
cultures/ethnicities. Future adolescent smoking prevention programs maybe
improved by incorporating aspects of emotional intelligence.
Introduction
Claims by the popular media of emotional intelligence being the most
important predictor of life success, accounting for up to 80% of that variance, have
led to a burgeoning of interest in the field of emotional intelligence research
(Goleman, 1995). Though such claims seem to be overly optimistic, some scientific
evidence is starting to emerge associating high emotional intelligence with increased
workplace performance and decreased aggression (Rice, 1999; Rubin, 1999). As
this exciting field of research continues to grow, elucidation of the relationship of
emotional intelligence with specific health behaviors, such as adolescent smoking,
beckons exploration. Of particular interest is the hypothesis that relationships
between smoking intentions and other smoking risk factors may vary differentially
across levels of emotional intelligence (El). An interaction between El and smoking
risk factors, such as perceived social consequences of smoking or refusal self-
54
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
efficacy, for example, may decrease the risk of smoking intentions for those with
high EL If so, adolescents with high El may better benefit from social-influences
based tobacco prevention programs targeting such factors.
Two differing models conceptualizing El exist: a mental ability model and a
mixed model. The mental ability model is a new construct that focuses primarily on
emotion and its interactions with cognition (Mayer, Camso & Salovey, 2000). This
model of El is conceptualized as the “ability to perceive accurately, appraise, and
express emotion; the ability to access and/or generate feelings when they facilitate
thought; the ability to understand emotion and emotional knowledge; and the ability
to regulate emotions” (Mayer & Salovey, 1997, p. 10). Mixed models of El differ
from ability models in that mixed models combine both mental abilities with
traditional traits, such as optimism, motivation or mood (Goleman, 1995; Bar-On,
1997). The mental ability model is a more focused conceptualization emphasizing
mental skills, as opposed to variables already known to be associated with smoking,
such as optimism and mood (Carvajal, Wiatrek, Evans, Knee & Nash, 2000; Defino,
Jamner & Whalen, 2001; Pallonen, Prochaska, Velicer, Prokhorov & Smith, 1998).
Therefore, we will utilize the mental ability model because it provides an
opportunity to expand our understanding of adolescent smoking and its precursors.
Previous work has explored EI’s direct association with adolescent smoking
(Trinidad & Johnson, 2002) and examined its relation to psychosocial factors that
potentially lead to smoking (Trinidad, Unger, Chou, Azen & Johnson, 2002).
Trinidad and Johnson (2002) found evidence for El as a protective factor against
55
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
smoking behaviors in a sample of 7th and 8th grade students. El was inversely
related to smoking behaviors and accounted for eleven percent of the variance after
controlling for potential confounding variables including age, gender and grades
received in school. Further analyses on the same sample also showed that those
with low El were over two times more likely to have engaged in smoking behaviors
(Trinidad & Johnson, 2001). A subsequent study utilizing data from 6th graders
examined El’s relation to psychosocial factors that potentially lead to smoking
(Trinidad, et al., 2002). The authors found high El to be related to lower risks for
psychosocial smoking risk factors, including lower intentions to smoke in the next
year, higher refusal self efficacy, and higher perceptions of negative social
consequences of smoking.
To build upon previous work on El and smoking behaviors, the present
study will explore the role of El as a moderator of the relationships between
psychosocial smoking risk factors and smoking intentions, and ever trying cigarettes
and smoking intentions. Given the previous findings of El as a protective factor, it
is theoretically plausible that high El may buffer the effects of smoking risk factors
and experimentation on intentions while low El could exacerbate such effects
(explained in greater detail below). Psychosocial risk factors of interest for the
current study include perceived social consequences of smoking, refusal self-
efficacy, and hostility. These variables have been targeted by adolescent smoking
prevention programs and shown to be related to future smoking behaviors (Hansen
& Graham, 1991; MacKinnon, Johnson, Pentz, et al., 1991; Scherwitz, Perkins,
56
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chesney, Hughes, Sidney & Manolio, 1992). We focus on smoking intentions
because our sample of students was relatively young (6th graders; discussed below)
and intentions have been numerously related to future smoking behaviors (e.g.,
Kaplan, Napoles-Springer, Stewart & Perez-Stable, 2001; Skara, Sussman & Dent,
2001).
Perceived Social Consequences of Smoking
Increasing perceptions of negative social consequences associated with
smoking are generally targeted by social-influences based tobacco prevention
programs (Hansen & Graham, 1991; MacKinnon, et al., 1991). Those with high El
perceive greater negative social consequences associated with smoking (Trinidad, et
al. 2002). They may view smoking as a characteristic of more delinquent behavior
as opposed to being a more grown up behavior. Adolescents with high El may
perceive that smokers do not have more friends, or the friends that smokers have are
not the friends that they would want for themselves. These students may also more
likely associate smoking with negative emotional states/consequences that result
from its social costs. Those with low El, on the other hand, may not be as
perceptive of such social consequences. We hypothesize that this difference in
ability to perceive social consequences of smoking, with high El perceiving greater
consequences, will put those with high El at lower risk for future smoking intentions
than those with low EL
57
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Refusal Self-Efficacy
High refusal self-efficacy is associated with a decrease in smoking risk
(Botvin, Griffin, Diaz, Miller, Ifill-Williams, 1999; Epstein, Williams, Botvin, Diaz,
& Ifill- Williams, 1999). High El is associated with high refusal self-efficacy for
cigarette offers (Trinidad, et al., 2002). Those with high El may be better able to
refuse cigarette offers by understanding more clearly their own feelings about
smoking. This greater understanding may contribute to feeling more confident in
articulating a potential refusal response that does not hurt one’s social position or
the feelings of the refused person. Those with low El might posses a less good
understanding of their own or others’ emotions regarding smoking and thus might
be less able to articulate a refusal response. We hypothesize that this variation in
refusal skills across El level (with high El increasing refusal efficacy) decreases
smoking intentions for those with high El more so than for those with low EL
Smoking Experimentation
Ever trying cigarettes, or lifetime smoking, is associated with future
established smoking behaviors and intentions to use in the future (Pierce, Choi,
Gilpin, Farkas, & Merritt, 1996; Unger, Johnson, Stoddard, Nezami & Chou, 1997).
We hypothesize that those with high El who have tried smoking will be more likely
to intend to smoke in the next year, and perhaps transition into regular smoking.
Future smoking intentions may more likely be based on prior use experiences for
those who are more aware of their feelings and what brings them about (high El).
Similarly, having a greater awareness of one’s feelings and what brings them about,
58
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
skills of El, suggests that those who intend to smoke in the future are likely to have
tried smoking in the past. This same relationship is expected to be true for those
with low El but to a lesser degree because their level of emotional awareness and
understanding is lower than those with high EI.
Previous work on EI and adolescent smoking did not include an examination
of the relationship between hostility and EI. This current study explores this
relationship and tests for an interaction with EI on smoking risk factors as hostility
has previously been found to be related to adolescent smoking (Delfino, et al., 2001;
Scherwitz et al., 1992). We hypothesize that those with high EI will be less hostile
due to higher ability to identify and manage their emotions. Also, for high EI
adolescents, high hostility level may not increase smoking intentions because they
may be better able to identify and manage emotions that could increase smoking
intentions. For those with low EI, on the other hand, high hostility may increase
intentions to smoke because they may not be as able to understand their feelings and
less able manage their emotions.
That adolescent cigarette smoking continues to be a major public health
concern despite several psychosocial risk factors being fairly well established and
targeted for intervention underscores the need to explore how novel factors may
affect, and potentially reduce, adolescent smoking (Lee, Gilpin & Pierce, 1993;
Welte, Barnes, Hoffman & Dintcheff, 1999). A more thorough understandi ng of the
link between EI and smoking precursors would provide a better understanding of the
smoking uptake process and potentially identify whether EI should be considered in
59
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
interventions aimed at psychosocial risk factors, smoking behaviors, or both. Such
added detail is the next logical step in identifying ways of how EI may be used to
strengthen smoking prevention programs. As adolescent smoking prevention
programs evolve, new methods of increasing their effectiveness via the
identification and targeting of emerging protective factors, like EI, becomes
important in order to reduce adolescent smoking prevalence.
Methods
The data described in this article are from the baseline portion of a
longitudinal school-based experimental trial of smoking prevention strategies in a
multicultural, urban population of adolescents in Southern California. The purpose
of the baseline survey was to assess tobacco use and related psychosocial and
cultural variables before the implementation of culturally relevant smoking
prevention programs. A measure of emotional intelligence was administered to a
subset of students from the control condition of the main prevention trial. Analyses
were performed on this subset of students.
Subjects
Subjects were 416 actively consented sixth-grade adolescents from public
middle schools in the greater Los Angeles area. This sample population had a mean
age 11.3 years and was 53% female. The ethnic distribution was 32% Latino, 29%
Asian/Pacific Islander, 13% White, 19% Multiethnic, and 6% Other (African
American students were included in this category due to a very small number in our
60
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
sample). Students were considered Latino if they answered "yes" to the question,
"Are you Hispanic/Latino?" Similar coding schemes were utilized for Asians and
Whites. Due to the small number of African Americans and other ethnicities,
students who were not Latino, Asian or White were considered Other. Students
who were more than one ethnicity, including Other, were coded as Multiethnic.
Table 5 summarizes the age, gender and ethnic characteristics of the study
participants.
Table 5. Age & Gender Characteristics
Frequency
(%)
Gender
Male 194 (46.8)
Female 221 (53.2)
Age
10 3 (0.7)
11 300 (72.5)
12 110 (26.6)
13 1 (0.2)
Procedure
Students completed the baseline questionnaire, a 160-item paper-and-pencil
survey, in their classrooms during a single class period (45-50 minutes). Trained
data collectors, who were not previously acquainted with the students, distributed
the surveys. The surveys were identified only by a code number, not with the
students’ names or any other identifying information. Because the students all were
attending English-language schools in which their classes were conducted only in
61
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
English (California state law prohibits bilingual education in public schools), a basic
level of English-language proficiency was assumed and the surveys were provided
only in English. However, students were encouraged to ask the data collectors to
clarify the meanings of any unfamiliar words.
During another classroom session several weeks after the baseline survey,
students completed a measure of EI (about 45-50 minutes). Students completed this
survey individually with the lead author reading each item, including examples, out
loud to the students. This allowed for the maintenance of uniformity in survey
administration and the minimization of differences in participation and reading
skills.
Measures
Emotional Intelligence.
The Multifactor Emotional Intelligence Scale, Adolescent Version (MEIS;
Mayer, Salovey & Caruso, 1997) was employed to assess the EI of the participants.
A thorough description of this test and how it is scored can be found in Mayer,
Caruso & Salovey (1999). Briefly, the MEIS is a competence-based measure
consisting of 4 branches assessing an individual’s ability to perceive, assimilate,
understand, and manage emotion in himself/herself and in others. In this study, the
MEIS was shortened by excluding the Assimilating Emotions branch (assessing
one’s ability to generate emotions) and including only the Emotional Identification,
Understanding Emotions, and Managing Emotion branches. The Assimilating
Emotions branch was excluded due to time restrictions at schools. Exploratory
62
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
factor analysis revealed that this abbreviated measure of EI, consisting of these 3
branches of EI, still formed an overall EI factor similar to Mayer et al.’s (1999)
overall EI model and one used in previous work by Trinidad and Johnson (2002). A
critical evaluation of the construct of EI, as measured by the MEIS, conducted by
Ciarrochi, Chan, and Caputi (2000) showed the MEIS to be a valid measure of EI.
The Emotional Identification branch was assessed with the Faces, Music,
and Stories subtests; the Designs subtest was excluded. Each of these subtests
presented a stimulus to the student (e.g., a picture of a face depicting certain
emotions, a musical selection, or a short vignette) and asked him/her to rate whether
a specific emotion was present in each stimulus. Possible responses were rated on a
5-point Likert-type scale, ranging from “Definitely Not Present”(l) to “Definitely
Present”(5).
The Understanding Emotions branch was assessed with the Relativity
subtest. This test consisted of items describing a conflicting social situation
between two persons. Students were asked to estimate how likely each character
experienced certain feelings. Possible responses were rated on a 5-point Likert-type
scale, ranging from “Extremely Unlikely” (1) to “Extremely Likely” (5).
The Managing Emotions branch was assessed with the Managing subtest.
This test assessed how well students were able to manage their own and others’
emotions by asking them to evaluate the effectiveness of certain courses of action
given a hypothetical situation. Possible responses were rated on a 5-point Likert-
type scale, ranging from “Extremely Ineffective” (1) to “Extremely Effective” (5).
63
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Consensus scoring was utilized in grading the MEIS. This is the standard
scoring method based on norms of reference groups studied by Mayer et al. (1999).
Since Latino and Asian students were underrepresented in Mayer et al.’s sample
used to create reference norms, we computed norms based on the current sample.
Following the procedure described by Mayer et al., each student’s response was
weighted by the proportion of the sample giving the same response. The resulting
score indicated the percent agreement with the population from which the subject
was drawn. For example, if 75% of the students sampled in this study reported that
happiness was “Definitely Present” in a picture of a face (“5” on the Likert-scale),
then a student who chose “5” would receive a score of 0.75 for the item. Sub-totals
for each branch were summed. Branch totals were then summed to obtain an
aggregate MEIS score.
Smoking and Psychosocial Risk Factors.
Lifetime smoking, or ever trying smoking, was assessed by asking students
if they had ever tried cigarettes, even a puff. This was a dichotomous (yes/no) item.
Measures of psychosocial risk factors for smoking were contained within the
baseline tobacco use survey. Dependent measures of interest were perceived social
consequences of smoking, refusal self-efficacy, intentions to smoke in the next year,
and hostility.
Perceived social consequences of smoking (SC) was an index score
comprised of eight variables measuring a student’s perception of the social
outcomes of smoking (alpha=0.66). Variables included items such as “Smoking
64
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
cigarettes is one way to lose friends who are nonsmokers” and “Kids who smoke are
more grown up” and “Smoking cigarettes makes kids look cool.” Possible
responses were rated on a 4-point Likert-type scale with high scores indicating
higher negative social consequences (i.e., high scores are more desirable). The
range of responses was 11 to 32.
Refusal self-efficacy (RSE) was assessed by asking the question, “If
someone you just met offered you a cigarette and you did not want it, how easy or
hard would it be to say ‘no’?” Possible responses were rated on a 4-point Likert-
type scale with high scores corresponding to greater refusal self-efficacy (e.g.,
“Very easy”) and low scores corresponding to less refusal self-efficacy (e.g., “Very
hard”).
Intention to smoke in the next year next year (yes/no) was measured using a
rationale derived from Pierce, et al.’s (1996) method of assessing smoking
susceptibility. Students who answered “No, definitely not” to the question, “At any
time in the next year (12 months) do you think you will smoke a cigarette?” were
classified as not intending to smoke in the next year, while those who answered
“Yes”, “Maybe yes”, and “Maybe no” were classified as susceptible.
Hostility was measured using 4-items from the Buss-Durkee Hostility
Scale’s Irritability subscale (Buss & Durkee, 1957). Items in this measure included
such things as “I lose my temper easily,” and “Sometimes people bother me just by
being around.” Possible responses were rated on a 4-point Likert-type scale with
65
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
high scores indicating more hostility (answering “yes” to the questions) and low
scores indicating less hostility (answering “no” to the questions).
Covariates.
Relevant covariates included several demographic variables, including age,
gender, ethnicity, socioeconomic status (ratio of number of people per room in a
home), linguistic acculturation (5-point Likert-type scale of language spoken at
home ranging from "English only" to "Only another language"), self-reported
assessment of overall grades received in school last year (e.g., mostly A’s, mostly
A’s & Bs, mostly B ’s, etc.), perceived peer attitudes toward smoking, and perceived
social norms of smoking. Self-reported grades was included in analyses as a proxy
for general academic ability. Previous research has shown El to be related to IQ
(Mayer, et al., 1999).
Perceived social norms of smoking was assessed by asking students to
identify the approximate number, out of 100 of similarly aged students, who smoked
cigarettes in the past month (e.g., 10, 20, 30, etc.). Perceived peer attitudes toward
smoking was measured by asking students the question, “Do most people your age
think it’s ok to smoke cigarettes once in a while?” Possible responses were rated on
a 4-point Likert type scale with high scores indicating that peers think it is ok to
smoke cigarettes once in a while and low scores indicating otherwise. Both of these
variables were related to the dependent variables and thus were controlled for in
statistical analyses.
66
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Statistical Analyses
All statistical analyses were performed using SAS version 8.1 (SAS
Institute, 2000). For the examination of the association between El and hostility
linear regression analyses were performed. Interactions between El and
psychosocial risk factors, and El and ever smoking, on smoking intentions were
tested by logistic regression models. Interaction variables were created by
multiplying together the centered, dichotomized (median splits) risk factor variables
and EL For example, for the interaction between El and hostility, the El variable
was dichotomized then centered and the hostility variable was dichotomized and
centered. Both resulting variables were multiplied together to form the interaction
variable. Variables that were split at the median and centered were El, SC, RSE and
hostility; ever smoking was naturally dichotomous. Each of the resulting variables
were multiplied with the dichotomized El variable. This resulted in four separate
logistic regression models testing for interactions. These four models included the
main effect variables, the interaction term and covariates. All regression models
controlled for age, gender, ethnicity, acculturation, socioeconomic status, grades
received in school, perceived social norms of smoking, and perceived peer attitudes
toward smoking.
Results
Correcting for the number of items, mean score on the MEIS was 0.36
(sd=0.06). Mean score on the Emotional Identification branch was 0.39 (sd=0.07);
67
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
for the Relativity branch, 0.31 (sd=0.07); and for the Managing branch, 0.24
(sd=0.04). Cronbach’s coefficient alphas for the MEIS were 0.75, for the Emotional
Identification subscale 0.83, for the Relativity subscale 0.66, for the Managing
subscale 0.76.
Mean score on the MEIS for girls was significantly higher than for boys
(59.2 vs. 57.1; p=0.03). Girls reported receiving higher grades in school than boys
(p<0.01). A significantly larger proportion of boys intended to smoke in the next
year compared to girls (11.5% vs 3.7%; p<0.01). No other significant differences
by gender were found.
Of all students sampled, 5.5% had ever tried cigarettes and 7.5% intended to
smoke in the next year. 60.5% reported it would be “Very easy” to refuse a
cigarette offer from someone they just met; 75% unequivocally viewed negative
social consequences associated with smoking; and 11.5% were very high in hostility
(i.e., highest quartile).
Controlling for age, gender, ethnicity, acculturation, socioeconomic status,
grades received in school, perceived social norms of smoking, and perceived peer
attitudes toward smoking, logistic regression analyses revealed a significant main
effect of ever trying cigarettes on smoking intentions (p<0.01). No significant main
effects of SC, RSE and hostility on smoking intentions were found. These logistic
regression models also tested for interactions between El and the psychosocial risk
factors (SC, RSE, hostility), and El and ever smoking, on smoking intentions. The
results of these separate regression models are summarized in Table 6 and are
68
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
described below. No significant interaction was found between El and SC on
smoking intentions. A statistically significant interaction was found between El and
refusal self-efficacy on smoking intentions (p=0.03) in that those with low El were
more likely to intend to smoke if refusal self-efficacy was low versus high, while
those with high El did not intend to smoke differentially regardless of refusal self-
efficacy level (Figure 4).
Table 6. Summary of separate logistic regression models examining interactions
_______ between smoking risk factors and El on smoking intentions__________
Intention to Smoke
_________________________________________________ Std. P p-value
Model 1: Perceived Social Consequences
Perceived Social Consequences main effect -0.25 0.07
Emotional Intelligence main effect -0.30* 0.02
Perceived Social Consequences X El -0.02 0.45
Model 2: Refusal Self-Efficacv
Refusal Self-Efficacy main effect -0.05 0.39
Emotional Intelligence main effect -0.30* 0.03
Refusal Self-Efficacy X El 0.33* 0.03
Model 3: Ever Tried Cigarettes
Ever Smoked main effect 0.29* <0.01
Emotional Intelligence main effect -0.52* <0.01
Ever Smoked X El 0.24* <0.01
Model 4: Hostility
Hostility main effect 0.15 0.17
Emotional Intelligence main effect -0.19 0.12
Hostility X El -0.30* 0.03
Note. Estimates are controlled for age, gender, SES, language spoken at home, grades received in
school, perceived social norms of smoking & perceived peer attitudes toward smoking.
*p<05
69
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4. Interaction between emotional intelligence and refusal
self-efficacy on smoking intentions
1.2
P $ .
O 1
| 0.8
a
o 0.6
S3
.2 0.4
S3
O
0.2
Interaction
p-value=0.03
a
NH
Low High
Refusal Self-Efficacy
• “♦ “ El Low “ “ • “ ■El High
Ever trying cigarettes interacted with El with regard to smoking intentions in
the next year (p<0.01). Those with high El were more likely to intend to smoke in
the next year if they had already tried smoking. Those with high El who did not
intend to smoke in the next year were not likely to have tried smoking. This
difference in ever trying smoking was not as pronounced for those with low El
(Figure 5).
Linear regression models testing for the association between El and hostility
yielded no statistically significant findings. However, emotional intelligence level
interacted with hostility level to affect smoking intentions (p=0.03). Specifically,
for those with low El, high hostility was increased the likelihood of intending to
smoke. For those with high El, such a difference was not evident (Figure 6).
70
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 5. Interaction between emotional intelligence and ever trying
cigarettes on smoking intentions
8
7
6
5
4
3
2
1
0
Interaction
p-value <0.01
No Yes
Ever Smoked
■"4-Loff El HI--High El
Figure 6. Interaction between emotional intelligence and hostility
on smoking intentions
6
5
Interaction
p-value=0.03
4
3
2
1
0
High Low
Hostility
♦ Low El HU—High El
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Discussion
Though previous research has begun to establish a link between El and overt
smoking behaviors (Trinidad & Johnson, 2002) and El and psychosocial smoking
risk factors (Trinidad, et al., 2002), interactions between El and smoking risk factors
had not yet been explored. Our current findings suggest that El does indeed interact
with several psychosocial smoking risk factors and lifetime smoking to affect future
smoking intentions. Those with low El are more likely to intend to smoke if they
have low refusal skills or are more hostile, while those with high El are more likely
to intend to smoke if they have previously experimented with cigarettes.
The test of interactions revealed that those with low El and low RSE are
more likely to intend to smoke in the next year compared to those with low El but
high RSE, as hypothesized. However, contrary to our hypothesis, for those with
high El, RSE did not seem to make a significant difference with respect to smoking
intentions. Those with high El might already know how to identify and manage
their emotions and social situations in such a way that their smoking intentions are
not modified by their RSE. Thus our findings suggest that refusal skills training
may be more beneficial for those with low EL In other words, those adolescents
who exhibit high El skills (e.g., able to identify self and others’ feelings, know how
to deal with potentially difficult social/emotional situations) or a stronger belief in
their ability to refuse cigarette offers are more protected from smoking. Conversely,
if an adolescent has difficulty with either, then he or she may be more likely to
72
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
intend to smoke. Therefore, increasing refusal skills for adolescents with low El is
likely to be more beneficial than for those with high EL
Those with high El were more likely to intend to smoke in the next year if
they had already tried smoking. This could be because those with high El
understood their behaviors and the associated antecedents better relative to those
with low EL Therefore, increased smoking intentions may have resulted if those
with high El found smoking experimentation to be rewarding and/or viewed few or
no consequences associated with it. It is then plausible that if those with high El
intend to smoke they might be more likely to follow through and do it. Conversely,
for those with low El intentions might not be as likely to translate into use. As
suspected, we found that for those with low El, the relationship between
experimentation and future smoking intentions was not at strong as those with high
EI. Compared to high El students, they may be less likely to intend to smoke even
if they have already tried it because of less awareness and understanding of what
drives their behaviors and intentions (i.e., less likely to “follow through”).
Therefore, if smoking intentions are targeted outcomes of prevention programs
Though no significant main effects between EI and hostility were found, an
interaction was present. It is interesting that EI was not significantly negatively
related to hostility as it has previously been negatively related to anti-social
behavior, another correlate of hostility (Mayer, Carlsmith & Chabot, 1998).
Perhaps this could be because hostility be is more of a personality trait or
disposition rather than a learned skill such as refusal skills or increased perceptions
73
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
of smoking consequences. Although in past research hostility has been shown to be
related to smoking (Scherwitz et al, 1992), hostility level did not seem to increase
intentions to smoke for those with high EI. High EI may buffer the effect of
hostility on smoking intentions. Those with low EI were more likely to intend to
smoke if they were more hostile. Those with low hostility, regardless of EI level,
did not intend to smoke. Therefore, the difference in effect of EI becomes more
apparent as hostility increases. Those with low EI might not be as capable of
identifying and managing their inner feelings or social situations which then could
lead to increased hostility and intentions to smoke.
Our main effect findings show that perceptions of negative social
consequences of smoking were not associated with smoking intentions. However,
perceptions of social consequences associated with smoking were related to high EI,
as in our previous work (Trinidad, et al., 2002). Interestingly, EI did not modify the
relationship between perceptions of social consequences associated with smoking
and smoking intentions. Contrary to our hypothesis, this finding suggests that the
role of EI on smoking intentions is independent from the role of perceiving
consequences on smoking intentions, even though both independent variables act
independently to decrease smoking intentions.
In addition to controlling for demographic variables, our analyses controlled
for grades received in school, perceived social norms of smoking, and perceived
peer attitudes toward smoking. Self-reported grades was included as a covariate in
all of our analyses as a proxy for IQ because previous research has shown EI to be
74
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
related to IQ (Mayer, et al., 1999). After controlling for self-reported grades, it is
interesting that the associations between EI and the psychosocial risk factors
remained significant. Therefore, although intellectual level may play a role in
reducing smoking risk (Young & Rogers, 1986), EI appears to contribute to the
reduction in these risks independently. We controlled for perceived social norms
and perceived peer attitudes toward smoking because these variables were
correlated with our dependent variables of interest.
Emotional intelligence hinges on the conception that certain emotional tasks
or problems have answers that can be judged either correct or incorrect. To obtain
the norms used to score the MEIS consensus scoring was employed. This was done
for several reasons. First, to some extent, emotions are socially based (Mayer &
Salovey, 1997). Second, research has shown that when groups of people’s
judgments are pooled together for items such as these Likert-based MEIS items, the
groups tend to serve as expert judges (Legree, 1995). Third, using various scoring
methods (consensus, expert and target) Mayer, et al. (1999) found that consensus
scoring produced the most reliable measures of EI. Mayer, et al.’s expert scoring
involved using answers on the MEIS deemed to be correct by experts in emotion
research. Target scoring involved obtaining correct answers from persons depicted
in each item. For example, for identifying emotions present in a photograph of a
young girl crying (in the Emotional Identification branch), the girl whose photo was
taken was asked what emotions she was feeling. Answers from these three scoring
methods were moderately and highly positively correlated, indicating that indeed a
75
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
set of answers that were “more correct” than others exist for items and stimuli such
as those contained within the MEIS.
Limitations
Further studies need to be performed in order to validate the findings of this
research. Causal inferences, particularly of a temporal nature, from this study
should be treated with caution because of its cross-sectional design. Future research
regarding EI and adolescent smoking may be conducted in the context of a
longitudinal, school-based prevention program in order to address whether those
with high EI do actually benefit more from such programs, and whether a program
to increase EI would help prevent smoking. Such research would address the
directionality of the relationship between EI, smoking risk factors and smoking
intentions (and, perhaps, smoking behavior). The resulting knowledge gained can
be used to develop improved adolescent smoking prevention programs.
Our measure of EI may not completely reflect all of its aspects, particularly
since we excluded the Assimilating Emotions branch. This branch involved
students generating various emotions and was excluded primarily due to time
constraints within classrooms. Furthermore, in our previous work we discovered
that this was a particularly difficult task to comprehend for those with limited
English proficiency or English as a second language. Though we acknowledge that
our abbreviation of the MEIS may have altered the generally accepted mental ability
model of EI, factor analyses performed on our abbreviated version indicated that the
abbreviated version we employed was still very similar to Mayer, et al. (1999).
76
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Work by Sullivan (1999) also utilized a similar abbreviated EI measure and found it
to be reflective of the overall MEIS. Regardless, the MEIS is a relatively new
measure of EI therefore not much published data exist as a way of comparison,
especially for adolescent samples. Similarly, future work may also benefit by
utilizing more specific measures of acculturation than was used in this study. The
acculturation process is complex, and a multi-item scale would have been a more
complete measure of acculturation.
Future Directions
Though results indicate that EI moderates the relationship between
psychosocial smoking risk factors and one’s smoking intentions, as well as
experimentation and smoking intentions, future work should examine these findings
across cultures/ethnicities and cultural contexts. Results from such work would
further inform about how the associations between EI and the psychosocial risk
factors vary across cultures/ethnicities, and which of the EI branches are more
developed at this age for various groups. Furthermore, examining EI in general
across various cultures/ethnicities may provide additional valuable information
about how one’s culture/ethnicity is related to smoking behaviors. For example,
previous research has shown that Asian adolescents smoke less than other
ethnicities (Chen & Unger, 1999). Understanding how EI varies across Asians,
Latinos and Whites, for example, and how such differences might relate to smoking
would then have implications for the tailoring of tobacco prevention programs.
77
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Likewise the examination of EI across various levels of acculturation maybe
informative for multicultural smoking prevention programs aimed at diverse
populations. Past research utilizing a language-based measure of acculturation
revealed that adolescents who spoke more English at home smoked more than those
who spoke another language (Unger, Cruz, Rohrbach, et al., 2000). The
identification of certain aspects of EI (mental ability skills) protective against
smoking behaviors at specific levels of acculturation could be used to enhance
multicultural smoking prevention programs.
The relationship between EI and smoking intentions might also be modified
by certain cultural values associated with various ethnicities. For example, smoking
relationships between the Asian and Latino values of filial piety may be associated
with lower smoking, while views of traditional male dominance might be associated
with increased smoking, particularly among males. These relationships might be
altered by an individual’s level of EI. An exploration of this relationship may be
informative in terms of tailoring prevention program towards specific cultures.
Finally, as promising as these findings are, our sample of students were not
yet of an age where actual smoking behaviors might have been more prevalent (e.g.,
high school). Our sample of sixth grade students did not exhibit high enough levels
of experimental smoking to detect large differences in analyses. Only about 5.5%
of students had ever tried smoking in our sample, while about 7.5% intended to
smoke. Sampling high school students, for example, may provide greater variability
in smoking behaviors. Replication of this current research on older samples may
78
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
yield more information about how EI moderates the relationship between
psychosocial risk factors and actual smoking behaviors. As just mentioned,
longitudinal studies examining the relationship between EI, psychosocial risk
factors and smoking behaviors would not only address temporal relationships but
also provide stronger conclusions regarding EEs relationship with actual smoking
behaviors.
Implications & Conclusion
Evidence of the relationship between EI and health behaviors is now starting
to increase. With the emerging trend within tobacco research towards positive and
protective factors, our findings regarding the protective role of EI against smoking
risk factors are encouraging; specifically that high EI buffers the effect of several
smoking risk factors on smoking intentions. The most effective adolescent smoking
prevention programs focus on social influences, such as perceived social
consequences of smoking and smoking intentions (Hansen & Graham, 1991;
MacKinnon, et al., 1991). Our findings indicate that those with high levels of EI
may be better prepared to process and utilize the information learned from such
social influences-based prevention programs and thus better benefit from them.
Similarly, our findings may be beneficial for prevention programs that emphasize
refusal skills training as those with high EI were better able to refuse cigarette
offers. Therefore, adding an EI component to future prevention programs may
enhance the targeted effects for variables such as refusal skill training, hostility and
79
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
smoking intentions. Thus as adolescent smoking prevention programs evolve,
taking into account the novel construct of EI may lead to increased effectiveness.
With EI research still in relative infancy, very little published literature
exists regarding its relationship with, and its influence on, adolescent smoking risk
factors. It is our hope that this work will positively add to the body of knowledge
about the risk factors for adolescent smoking and how EI is associated with these
factors.
Acknowledgements: We wish to express our thanks to Drs. Peter Salovey, John
Mayer, and David Caruso for allowing us to use the MEIS and sharing their ideas
and research with us. To them and all others who contributed to this work, we are
very grateful and extend our heartfelt thanks.
80
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
Bar-On, R. (1997). The Emotional Quotient Inventory (EO-i): Technical
Manual. Toronto, Canada: Multi-Health Systems.
Botvin, G.J., Griffin, K.W., Diaz, T., Miller, N„ & Ifill-Williams, M.
(1999). Smoking initiation and escalation in early adolescent girls: One-year
follow-up of a school-based prevention intervention for minority youth. Journal of
the American Medical Womens Association, 54, 139-143, 152.
Buss, A. H., & Durkee, A. (1957). An inventory for assessing different
kinds of hostility. Journal of Consulting Psychology, 21, 343-348.
Carvajal, S.C., Wiatrek, D.E., Evans, R.I., Knee, C.R., & Nash, S.G. (2000).
Psychosocial determinants of the onset and escalation of smoking: Cross-sectional
and prospective findings in multiethnic middle school samples. Journal of
Adolescent Health, 27, 255-65.
Chen. X. & Unger. J..B. (1999). Hazards of smoking initiation among
Asian American and non-Asian adolescents in California: a survival model analysis.
Preventive Medicine. 28, 589-599.
Ciarrochi, J. V., Chan, A. Y. C., & Caputi, P. (2000). A critical evaluation
of the emotional intelligence construct. Personality and Individual Differences. 28.
539-561.
Defino, R.J., Jamner, L.D., & Whalen, C.K. (2001). Temporal analysis of
the relationship of smoking behavior and urges to mood states in men versus
women. Nicotine and Tobacco Research. 3. 235-48.
Epstein, J.A., Williams, S.C., Botvin, G.J., Diaz, T., & Ifill-Williams, M.
(1999). Psychosocial predictors of cigarette smoking among adolescents living in
public housing developments. Tobacco Control. 8, 45-52.
Goleman, D. (1995). Emotional Intelligence. New York: Bantam Books.
Hansen, W. B., & Graham, J. W. (1991). Preventing alcohol, marijuana,
and cigarette use among adolescents: peer pressure resistance training versus
establishing conservative norms. Preventive Medicine, 20, 414-430.
Kaplan, C.P., Napoles-Springer, A., Stewart. S.L.. & Perez-Stable, E.J.
(2001). Smoking acquisition among adolescents and young Latinas: the role of
socioenvironmental and personal factors. Addictive Behaviors, 26, 531-550.
81
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Lee, L.L., Gilpin ,E.A., & Pierce, J.P. (1993). Changes in the patterns of
initiation of cigarette smoking in the United States: 1950, 1965, and 1980. Cancer
Epidemiology. Biomarkers & Prevention, 2. 593-597.
Legree, P. J. (1995). Evidence for an oblique social intelligence factor
established with a likert-based testing procedure. Intelligence, 21. 247-266.
MacKinnon, D. P., Johnson, C. A., Pentz, M. A., Dwyer, J. H., Hansen, W.
B., Flay, B. R., & Wang, E. Y. (1991). Mediating mechanisms in a school-based
drug prevention program: first-year effects of the Midwestern Prevention Project.
Health Psychology, 10. 164-172.
Mayer, J.D., Carlsmith, K.M., & Chabot, H.F. (1998). Describing the
person’s external environment: Conceptualizing and measuring the Life Space.
Journal of Research in Personality, 32, 253-296.
Mayer, J. D., Caruso D., & Salovey, P. (2000). Models of emotional
intelligence. In R.J. Sternberg (Ed.). Handbook of Intelligence (pp. 396-420).
Cambridge, UK: Cambridge University Press.
Mayer, J. D., Caruso D. R, & Salovey, P. (1999). Emotional intelligence
meets traditional standards for an intelligence. Intelligence. 27. 267-298.
Mayer, J.D., & Salovey, P. (1997). What is emotional intelligence? In P.
Salovey & D. Sluyter (Eds). Emotional Development and Emotional Intelligence:
Implications for Educators (pp.3-31). New York: Basic Books.
Mayer, J. D., Salovey, P., & Caruso, D. R. (1997). Multifactor Emotional
Intelligence Scale, Student Version. Durham, New Hampshire: Authors.
Pallonen, U.E., Prochaska, J.O., Velicer, W.F., Prokhorov, A.V., & Smith,
N.F. (1998). Stages of acquisition and cessation for adolescent smoking: an
empirical integration. Addictive Behaviors, 23, 303-24.
Pierce, J.P., Choi, W.S., Gilpin, E.A., Farkas, A.J., & Merritt, R.K. (1996).
Validation of susceptibility as a predictor of which adolescents take up smoking in
the United States. Health Psychology, 15, 355-61.
Rice, C.L. (1999). A quantitative study of emotional intelligence and its
impact on team performance. Unpublished master’s thesis, Pepperdine Unidersity.
82
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Rubin, M.M. (1999). Emotional intelligence and its role in mitigating
aggression: A correlational study of the relationship between emotional intelligence
and aggression in urban adolescents. Unpublished manuscript, hnmaculata College,
Immaculata, PA.
SAS Institute [Computer software], (2000). Cary, North Carolina: SAS
Institute.
Scherwitz, L.W., Perkins, L.L., Chesney, M.A., Hughes, G.H., Sidney, S. &
Manolio, T.A. (1992). Hostility and health behaviors in young adults: The
CARDIA Study. Coronary Artery Risk Development in Young Adults Study.
American Journal of Epidemiology. 136, 135-45.
Skara, S. Sussman. S. Dent, C.W. (2001). Predicting regular cigarette use
among continuation high school students. American Journal of Health Behavior,
25, 147-156.
Sullivan, A. K. (1999). The Emotional Intelligence Scale for Children.
Dissertation Abstracts International. Unpublished doctoral dissertation.
Trinidad, D.R. & Johnson, C.A. (2001). Is low emotional intelligence a risk
factor for adolescent tobacco and alcohol use? Poster Presented at the Society for
Behavioral Medicine and Society for Research on Nicotine and Tobacco Joint
Annual Meeting.
Trinidad & Johnson (2002). The Association Between Emotional
Intelligence and Early Adolescent Tobacco and Alcohol Use. Personality and
Individual Differences, 32, 95-105.
Trinidad, D.R., Unger, J.B., Chou, C.P., Azen, S.P., & Johnson, C.A.
(2002). The protective association of emotional intelligence with psychosocial
smoking risk factors for adolescents. Manuscript submitted for publication.
Unger, J.B., Cruz, T.B., Rohrbach, L.A., Ribisl, K.M., Baezconde-
Garbanati, L., Chen, X., Trinidad, D.R., & Johnson, C.A. (2000). English language
use as a risk factor for smoking initiation among Hispanic/Latino and Asian
American adolescents in California: Evidence for mediation by tobacco-related
beliefs and social norms. Health Psychology, 19,403-410.
Unger, J.B., Johnson, C.A., Stoddard, J.L., Nezami, E., & Chou, C.P.
(1997). Identification of adolescents at risk for smoking initiation: validation of a
measure of susceptibility. Addictive Behaviors, 22, 81-91.
83
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Welte, J. W., Baraes, G. M., Hoffman, J. H., & Dintcheff, B. A. (1999).
Trends in adolescent alcohol and other substance use: relationships to trends in peer,
parent, and social influences. Substance Use & Misuse, 34, 1427-1449.
Young, T.L., & Rogers, K.D. (1986). School performance characteristics
preceding onset of smoking in high school students. American Journal of Diseases
in Children. 140. 257-259.
84
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 4:
Emotional intelligence and ethnicity:
Interactions on smoking intentions in early adolescents
This study was supported by Grant Number 9DT-0174 from the Tobacco
Related Diseases Research Program of the University of California, Office of the
President, and Grant Number 5 P5 CA 84735 from the National Cancer Institute.
All correspondence regarding this manuscript should be directed to Dennis R.
Trinidad, Institute for Health Promotion and Disease Prevention Research and
Department of Preventive Medicine, Keck School of Medicine, University of
Southern California, 1000 S. Fremont Ave., Unit 8, Alhambra, CA 91803. Phone:
626-457-4163. Fax: 626-457-4012. Email:dtrinida@usc.edu.
Abstract
Some previous research has established that EI is protective against smoking
and smoking intentions. However, the link between EI and smoking intentions, a
predictor of future smoking behavior, may vary differentially by culture/ethnicity.
An understanding of the interaction between EI and ethnicity on smoking intentions
may help in designing improved targeted smoking prevention programs for different
ethnic cultures. EI is defined as the ability to: accurately perceive, appraise, and
express emotion; access and/or generate feelings in facilitating thought; understand
emotion and emotional knowledge; and regulate emotions. EI was assessed with a
85
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
shortened version of the Multifactor Emotional Intelligence Scale, Adolescent
Version, and was administered to 416 6 graders (53% girls) from middle schools in
the Los Angeles area (mean age=l 1.3 yrs; 32% Hispanic/Latino (H/L), 29%
Asian/Pacific Islander (A/PI), 13% White, 19% Multiethnic, 6% Other). Logistic
regression analyses revealed that the buffering effect of EI against smoking
intentions across ethnicities approached statistical significance. The trend of the
interaction effect for Whites was stronger compared with A/PI and H/L adolescents
(p=0.098 for A/PI X EI; p=0.059 for H/L X EI). These findings highlight the need
for further research on a larger sample of diverse adolescents and may suggest that
culturally tailored prevention programs aimed at decreasing smoking intentions in
early adolescents could benefit by incorporating modules designed to increase EI.
Introduction
Recent interest in the concept of emotional intelligence (EI) has increased
due to the popular media claiming it to be the most important predictor of life
success, with some suggesting that EI could account for up to 80% of that variance
(Goleman, 1995). Though much scientific and popular literature views such claims
as overly optimistic (Hedlund & Sternberg, 2000; Mayer, Caruso & Salovey, 2000),
this exciting field of research continues to grow. Exploration of the relationship
between EI, or the interaction of emotions and cognition, and specific health
behaviors, such as adolescent smoking, is of particular interest. Some previous
research has established that EI is protective against smoking and smoking
86
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
intentions (Trinidad and Johnson, 2002; Trinidad, Unger, Chou, Azen & Johnson,
2002a). However, the link between EI and smoking intentions, a predictor of future
smoking behavior (; O’Callaghan, Callan, Baglioni, 1999; Flay, Hu, Richardson,
1998; Pierce, Choi, Gilpin, Farkas, & Merritt, 1996), may vary differentially by
culture/ethnicity. An understanding of the interaction between EI and ethnicity on
smoking intentions may help in designing improved targeted smoking prevention
programs for different ethnic cultures.
Adolescent smoking continues to be a major public health problem. Recent
studies have revealed ethnic/cultural differences in adolescent smoking (Chen &
Unger, 1999; Chen, Unger, Cruz & Johnson, 1999; Johnston, O’Malley &
Bachman, 1998). Past research has clearly shown smoking intentions to be highly
predictive of future smoking behaviors (Flay, et al., 1998; O’Callaghan, et al., 1996;
Pierce, et al., 1996). As such, smoking intentions also vary across
cultures/ethnicities and such differences suggest that there maybe variation in the
strength of shared smoking intention predictors across ethnicities.
Previous research on EI and adolescent smoking has looked at EI and
smoking, and EI and smoking risk factors. This body of research has shown EI to
be protective against adolescent smoking (Trinidad & Johnson, 2002), as well as
some psychosocial smoking risk factors (Trinidad, Unger, Chou, Azen & Johnson,
2002a). Trinidad and Johnson (2002) found evidence for EI as a protective factor
against smoking behaviors in a sample of 7th and 8th grade students, and that low EI
adolescents were over two times more likely to have engaged in smoking behaviors
87
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(Trinidad & Johnson, 2001). Another study utilizing data from 6th graders further
suggested that high EI was related to lower risks for psychosocial smoking risk
factors, including decreased intentions to smoke in the next year, increased refusal
self efficacy, and increased perceptions of negative social consequences of smoking
(Trinidad, et al., 2002a). Subsequent research on this 6th grade sample also revealed
that students with low EI were more likely to intend to smoke if they had low
refusal self-efficacy or were more hostile, and that for low EI adolescents, intentions
were not as likely to translate into use (Trinidad, Unger, Chou, Azen & Johnson,
2002b).
As promising as these previous findings are, however, no research to date
has looked at the relationship between EI and smoking intentions across ethnicities.
Given ethnic differences in smoking rates, it is plausible that the relationship
between EI and smoking intentions may differ by ethnicity. The strength of the
association between EI and smoking intentions may be a stronger predictor of
smoking in adolescents from certain cultures/ethnicities. In other words, EI might
then interact with ethnicity to affect smoking intentions and may buffer against
smoking intentions for adolescents of a particular culture/ethnicity but not for
others. This paper will explore these relationships. Such examination across
ethnicities may help explain some of the variation in smoking across
cultures/ethnicities. The resulting knowledge could lead to improved smoking
prevention programs targeting specific adolescent ethnic cultures.
88
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Though several measures of El exist, such as the Bar-On EQi (1997) and the
EQ Map (Cooper, 1997), the most commonly accepted measure of El in academia is
the Multifactor Emotional Intelligence Scale (Mayer, Salovey & Caruso, 1997). As
such, our previous work exploring El and adolescent smoking made use of this
measure. The recommended method of scoring the MEIS, and the only one used in
the few empirical smoking-related studies utilizing it thus far, is consensus scoring
(Mayer, et al., 1997, Trinidad & Johnson, 2002; Trinidad, et al., 2002a, Trinidad, et
al., 2002b). Briefly, and expanded upon in the Methods section, this scoring method
involves weighing a respondent’s answer by the proportion of the sample giving the
same response and aggregating the scores to form an overall El score. This
consensus approach may indirectly measure collectivism (Triandis, 1995).
Collectivism is generally defined as group oriented, low in individualism and
placing much importance in public relations (Triandis, 1995). We hypothesize that
those from collectivistic cultures, Asian/Pacific Islanders (A/PI) and
Hispanic/Latinos (H/L) will score higher on the El measure.
Two differing models conceptualizing El exist: a mental ability model and a
mixed model. The mental ability model is a new construct that focuses primarily on
emotion and its interactions with cognition (Mayer, Caruso & Salovey, 2000). This
model of El is conceptualized as the “ability to perceive accurately, appraise, and
express emotion; the ability to access and/or generate feelings when they facilitate
thought; the ability to understand emotion and emotional knowledge; and the ability
to regulate emotions” (Mayer & Salovey, 1997, p. 10). Mixed models of El differ
89
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
from ability models in that mixed models combine both mental abilities with
traditional traits, such as optimism, motivation or mood (Goleman, 1995; Bar-On,
1997). The mental ability model is a more focused conceptualization emphasizing
mental skills, as opposed to variables already known to be associated with smoking,
such as optimism and mood (Carvajal, Wiatrek, Evans, Knee & Nash, 2000; Defino,
Jamner & Whalen, 2001; Pallonen, Prochaska, Velicer, Prokhorov & Smith, 1998).
Therefore, this research will utilize the mental ability model because it will provide
an opportunity to expand our understanding of adolescent smoking and its
precursors.
Understanding the differences of associations between El and smoking
intentions across cultures/ethnicities may help in designing improved targeted
smoking prevention programs for different ethnic cultures. With the California
population, particularly Los Angeles, becoming increasingly diverse the exploration
of associations between El and smoking relationships across ethnicities is
warranted.
Methods
The data described in this article are from the baseline portion of a
longitudinal school-based experimental trial of smoking prevention strategies in a
multicultural, urban population of adolescents in Southern California. Latino and
Asian adolescents were oversampled. The purpose of the baseline survey was to
assess tobacco use and related psychosocial and cultural variables before the
90
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
implementation of culturally relevant smoking prevention programs. A measure of
emotional intelligence was administered to a subset of students from the control
condition of the main prevention trial. Analyses were performed on this subset of
students.
Subjects
Subjects were 416 actively consented sixth-grade adolescents from public
middle schools in the greater Los Angeles area. Students were considered Latino if
they answered "yes" to the question, "Are you Hispanic/Latino?" Similar coding
schemes were utilized for Asians and Whites. Due to the small number of African
Americans and other ethnicities, students who were not Latino, Asian or White were
considered Other. Students who were more than one ethnicity, including Other,
were coded as Multiethnic.
Procedure
Students completed the baseline questionnaire, a 160-item paper-and-pencil
survey, in their classrooms during a single class period (45-50 minutes). Trained
data collectors, who were not previously acquainted with the students, distributed
the surveys. The surveys were identified only by a code number, not with the
students’ names or any other identifying information. Because the students all were
attending English-language schools in which their classes were conducted only in
English (California state law prohibits bilingual education in public schools), a basic
level of English-language proficiency was assumed and the surveys were provided
91
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
only in English. However, students were encouraged to ask the data collectors to
clarify the meanings of any unfamiliar words.
During another classroom session several weeks after the baseline survey,
students completed a measure of El (about 45-50 minutes). Students completed this
survey individually with the lead author reading each item, including examples, out
loud to the students. This allowed for the maintenance of uniformity in survey
administration and the minimization of differences in participation and reading
skills.
Measures
Emotional Intelligence.
The Multifactor Emotional Intelligence Scale, Adolescent Version (MEIS;
Mayer, Salovey & Caruso, 1997) was employed to assess the El of the participants.
A thorough description of this test and how it is scored can be found in Mayer,
Caruso & Salovey (1999). Briefly, the MEIS is a competence-based measure
consisting of 4 branches assessing an individual’s ability to perceive, assimilate,
understand, and manage emotion in himself/herself and in others. In this study, the
MEIS was shortened by excluding the Assimilating Emotions branch (assessing
one’s ability to generate emotions) and including only the Emotional Identification,
Understanding Emotions, and Managing Emotion branches. The Assimilating
Emotions branch was excluded due to time restrictions at schools. Exploratory
factor analysis revealed that this abbreviated measure of El, consisting of these 3
branches of El, still formed an overall El factor similar to Mayer et al.’s (1999)
92
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
overall El model and one used in previous work by Trinidad and Johnson (2002). A
critical evaluation of the construct of El, as measured by the MEIS, conducted by
Ciarrochi, Chan, and Caputi (2000) showed the MEIS to be a valid measure of EL
The Emotional Identification branch was assessed with the Faces, Music,
and Stories subtests; the Designs subtest was excluded. Each of these subtests
presented a stimulus to the student (e.g., a picture of a face depicting certain
emotions, a musical selection, or a short vignette) and asked him/her to rate whether
a specific emotion was present in each stimulus. Possible responses were rated on a
5-point Likert-type scale, ranging from “Definitely Not Present” (1) to “Definitely
Present” (5).
The Understanding Emotions branch was assessed with the Relativity
subtest. This test consisted of items describing a conflicting social situation
between two persons. Students were asked to estimate how likely each character
experienced certain feelings. Possible responses were rated on a 5-point Likert-type
scale, ranging from “Extremely Unlikely” (1) to “Extremely Likely” (5).
The Managing Emotions branch was assessed with the Managing subtest.
This test assessed how well students were able to manage their own and others’
emotions by asking them to evaluate the effectiveness of certain courses of action
given a hypothetical situation. Possible responses were rated on a 5-point Likert-
type scale, ranging from “Extremely Ineffective” (1) to “Extremely Effective” (5).
Consensus scoring was utilized in grading the MEIS. This is the standard
scoring method based on norms of reference groups studied by Mayer et al. (1999).
93
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Since Latino and Asian students were underrepresented in Mayer et al.’s sample
used to create reference norms, we computed norms based on the current sample.
Following the procedure described by Mayer et al., each student’s response was
weighted by the proportion of the sample giving the same response. The resulting
score indicated the percent agreement with the population from which the subject
was drawn. For example, if 75% of the students sampled in this study reported that
happiness was “Definitely Present” in a picture of a face (“5” on the Likert-scale),
then a student who chose “5” would receive a score of 0.75 for the item. Sub-totals
for each branch were summed. Branch totals were then summed to obtain an
aggregate MEIS score. Cronbach’s coefficient alpha’s for the MEIS was 0.75, for
the Emotional Identification subscale 0.83, for the Understanding Emotions
subscale 0.66, for the Managing Emotions subscale 0.76.
Intention to Smoke
Intention to smoke in the next year next year (definitely not vs. all others)
was measured using a rationale derived from Pierce, et al.’s (1996) method of
assessing smoking susceptibility. Students who answered “No, definitely not” to
the question, “At any time in the next year (12 months) do you think you will smoke
a cigarette?” were classified as not intending to smoke in the next year, while those
who answered “Yes”, “Maybe yes”, and “Maybe no” were classified as susceptible.
Covariates.
Relevant covariates included several demographic variables, including age,
gender, ethnicity, socioeconomic status (ratio of number of people per room in a
94
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
home), linguistic acculturation (5-point Likert-type scale of language spoken at
home ranging from "English only" to "Only another language"), and self-reported
assessment of overall grades received in school last year (e.g., mostly A ’s, mostly
A’s & Bs, mostly B’s, etc.). Self-reported grades was included in analyses as a
proxy for IQ. Previous research has shown El to be related to IQ (Mayer, et al.,
1999).
Statistical Analyses
Logistic regression analyses were performed to assess the interaction
between El and ethnicity on smoking intentions, a dichotomous variable. Each
model included main effects for El, main effects for ethnicity, and El X ethnicity
interaction terms. Interaction variables were created by multiplying the centered El
variable with dummy variables for each ethnicity. For example, for the interaction
between El and H/L, the El variable was centered and then multiplied with the H/L
variable (yes/no). Each interaction term was constmcted to assess whether each
ethnic group differed from Whites on the strength of the association between El and
smoking intentions. A positive interaction term indicates that the association
between El and smoking intentions is stronger in the ethnic minority group than
among Whites. Conversely, a negative interaction term indicates that the
association between El and smoking intentions is weaker in the ethnic minority
group than among Whites. The models were controlled for age, gender,
socioeconomic status, language spoken at home, and self-reported grades received
95
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
in school. All analyses were performed using SAS version 8.1 (SAS Institute,
2000).
Results
Demographic characteristics of respondents
Table 7 summarizes the age, gender and ethnic characteristics of the study
participants. The sample population of 416 adolescents had a mean age 11.3 years
and was 53% female. The ethnic distribution was 32% Hispanic/Latino, 29%
Asian/Pacific Islander, 13% White, 19% Multiethnic, and 6% Other.
Table 7. Demographic Characteristics
Frequency
(%)
Gender
Male 194 46.8
Female 221 53.2
Age
10 3 0.7
11 300 72.5
12 110 26.6
13 1 0.2
Ethnicity
White 53 12.7
Hispanic/Latino 135 32.3
Asian/PI 121 29.1
Multi-ethnic 80 19.2
Other 27 6.5
Gender differences in MEIS score and smoking intentions
Unadjusted mean score on the MEIS for girls was significantly higher than
for boys (59.16 vs. 57.14; p=0.03). Girls reported receiving higher grades in school
96
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
than boys (p<0.01). A significantly larger proportion of boys intended to smoke in
the next year compared to girls (11.46% vs 3.65%; p<0.01). No other significant
differences by gender were found. These results are summarized at the bottom of
Table 8.
Table 8. Emotional Intelligence and Smoking Intention
across Ethnicity and Gender
Intention to Smoke:
Mean Emotional % not answering
Intelligence Score “Definitely Not” (n)
Ethnicity
White 58.93 7.69 (4)
H/L 57.23 13.33 (18)
A/PI 60.52a 2.48 (3)
Multi-ethnic 56.85 3.85 (3)
Other 55.29 11.54 (3)
Chi-
square=T3.09**
Gender
Girl 59.16 3.65
Boy 57.14 11.46
t-value=2.24* Chi-
square=9.21 ***
*p<0.05
**p=0.01
***p<0.01
a A/PI students scored higher on the MEIS than all other students in our sample.
Ethnic variation in MEIS score and smoking intentions
Table 8 also shows the ethnic differences in MEIS scores and in smoking
intentions. A comparison of mean El scores across ethnicity revealed that A/PI
adolescents scored higher on the MEIS than other students in our sample (p<0.05).
No other ethnic differences regarding MEIS scores were statistically significant.
97
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Absence of a firm commitment to not smoke was highest among Hispanic/Latinos
and adolescents of Other ethnicity (Chi-square=13.09; p-value=0.01).
Ethnic variation in the association between MEIS score and smoking intentions
Table 9 shows the association between emotional intelligence, ethnicity and
smoking intentions from logistic regression analyses. All analyses are controlled for
age, gender, socioeconomic status, language spoken at home, and self-reported
grades received in school. The numbers shown in the table are standardized Beta
coefficients.
Table 9. Ethnic differences in El on Intentions to Smoke in the Next Year
Intention
to Smoke
Std. p p-value
Ethnicity main effects
Hispanic/Latino (vs. White) -.649 .096
Asian/Pacific Islander (vs. White) -.274 .505
Multi-ethnic (vs. White) -.259 .458
Other ethnicity (vs. White) -.072 .790
Emotional intelligence main effect
Emotional Intelligence 1.052 .051
Ethnicity X Emotional Intelligence interactions
Hispanic/Latino X Emotional Intelligence -.694 .059
Asian/Pacific Islander X Emotional Intelligence -.501 .098
Multi-ethnic X Emotional Intelligence -.285 .231
Other ethnicity X Emotional Intelligence -.125 .593
Note. Estimates are controlled for age, gender, SES, language spoken at home,
and grades received in school.
Whites are the reference group for all interaction terms.
98
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Main effects of emotional intelligence. Emotional intelligence score on the
MEIS was negatively associated smoking intentions in the next year for Whites
(p<0.05).
Main effects of ethnicity. Compared with Whites, intention to smoke in the
next year was significantly higher among Hispanic/Latinos (p<0.05).
Interactions between MEIS score and ethnicity on smoking intention. The
buffering effect of El against smoking intentions across ethnicities approached
statistical significance. The trend of the effect for Whites was stronger compared
with A/PI and H/L adolescents (i.e., the interaction terms for A/PI X El and H/L X
El were negative; p=0.098 for A/PI X El; p=0.059 for H/L X El). No other
interaction terms were significant (i.e., the associations between El and smoking
intentions did not differ between Whites and any of the other ethnic groups).
Discussion
The association between emotional intelligence and smoking intentions did
not statistically significantly vary across culture/ethnicity, though there was a trend
towards significance. Specifically, the trend we detected was that El might have
been more protective against smoking intentions in the next year for White
adolescents relative to A/PI and H/L adolescents (interaction terms were negative
and approached significance). Additional statistical power in a sample of older
adolescents may have allowed us to come to stronger conclusions regarding the
effect of El on smoking intentions in Whites relative to A/PI and H/L.
99
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Though our findings are far from definitive, they could suggest that
prevention methods incorporating El might be more beneficial for one culture than
for another. Our findings indicate that including an El component (as defined and
measured by the MEIS) in smoking prevention programs might help reduce future
smoking intentions more so for White than A/PI and H/L, particularly given that
Whites have been shown to smoke at higher rates (Johnston, et al., 1998). For
prevention programs aimed at decreasing smoking intentions in early adolescents,
incorporating modules designed to increase El may be beneficial. Modules
emphasizing the most basic components of El, Identifying Emotions and
Understanding Emotions, may be the most efficacious because the other branch of
El, Managing Emotions, may not yet be well developed at such an early age
(Mayer, Caruso & Salovey, 2000; Trinidad, et al., 2002a).
El score for A/PI adolescents was significantly higher than for all other
ethnicities in our sample. To verify, we also compared MEIS scores across
ethnicity after controlling for potential confounding variables, including age,
gender, SES, grades and English language use. After controlling for these variables,
A/PI students still scored higher on the MEIS compared with other ethnicities
except White. These findings suggest that the scoring method of the MEIS might
indirectly measure collectivism, a traditional feature of A/PI and H/L cultures
(Triandis, 1995). However, given that H/L students did not score higher than other
ethnicities suggests that El is not solely a measure of collectivism, and that its
collectivistic aspect was not the lone reason for the MEIS score variation across
100
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ethnicity. Perhaps other features of A/PI culture contributed to increased scores for
A/PI adolescents in our sample. For example, features of Confuci an-based cultures,
such as A/PI, emphasize sensitivity to non-verbal communication strategies. This
includes things such as being attentive to a person’s facial expression and vocal
tone, and may increase one’s ability to identify emotions in faces and stories, two
components of the El scale. Therefore, we examined correlations between El
subscales and ethnicity and found that indeed A/PI ethnicity was significantly
associated with higher scores on the Identifying Emotions subscale of the MEIS. In
other words, higher scores on the Identifying Emotions branch, the first branch of El
to develop, were associated with being of A/PI culture/ethnicity. No other El
branch scores were significantly correlated with ethnicity/culture, even when the
effects of potentially confounding variables were partialed out, including age,
gender, SES, grades and language spoken at home. Though score on the MEIS was
higher for A/PI adolescents and yet was not significantly protective is interesting. It
may be that since A/PI students are already at lower risk of intending to smoke
compared with Whites, El as a protective factor against smoking intentions might be
more pronounced for White adolescents.
In line with our recent work (Trinidad, et al., 2002ab), we also examined
associations between El and smoking risk factors, including perceived social
consequences of smoking and refusal self-efficacy, across ethnicity and found no
statistically significant results or trends toward statistical significance. This lack of
significant interactions suggests that for variables such as perceived social
101
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
consequences and refusal self-efficacy, the association of those variables with El do
not vary by ethnicity. However, the current findings regarding a trend in variation
in the association between El and smoking intentions across ethnicity suggest that
some smoking risk factor variables may vary by ethnicity and El. Further research
on other such variables is warranted in light of the potential to improve smoking
prevention programs for adolescents.
We also tested interactions between El and cultural values on smoking
intentions, perceived social consequences and refusal self-efficacy. This was done
to determine whether the previously found significant, protective associations
between El and smoking risk factors (Trinidad, et al., 2002a) would be modified by
cultural value-variables associated with certain ethnicities. However, no significant
interactions were found between El and any cultural value-variables in our data,
including filial piety, machismo and familism. It may be that 6th grade students in
our sample did yet not have crystallized opinions or conceptions regarding these
cultural values.
Validity of El Scoring Method
Emotional intelligence hinges on the conception that certain emotional tasks
or problems have answers that can be judged either correct or incorrect. To obtain
the norms used to score the MEIS consensus scoring was employed. This was done
for several reasons. First, to some extent, emotions are socially based (Mayer &
Salovey, 1997). Second, research has shown that when groups of people’s
judgments are pooled together for items such as these Likert-based MEIS items, the
102
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
groups tend to serve as expert judges (Legree, 1995). Third, using various scoring
methods (consensus, expert and target) Mayer, et al. (1999) found that consensus
scoring produced the most reliable measures of EL Mayer, et al.’s expert scoring
involved using answers on the MEIS deemed to be correct by experts in emotion
research. Target scoring involved obtaining correct answers from persons depicted
in each item. For example, for identifying emotions present in a photograph of a
young girl crying (in the Emotional Identification branch), the girl whose photo was
taken was asked what emotions she was feeling. Answers from these three scoring
methods were moderately and highly positively correlated, indicating that indeed a
set of answers that were “more correct” than others exist for items and stimuli such
as those contained within the MEIS.
Limitations
As research into El’s relationship with health behaviors, particularly
adolescent smoking, is still in its infancy further studies need to be performed in
order to validate the findings of this research. Future research regarding El and
adolescent smoking may be conducted in the context of a longitudinal, culturally
tailored, school-based prevention program in order to address whether those with
high El do actually benefit more from such programs, and whether a program to
increase El would help prevent smoking across cultures. Furthermore, longitudinal
studies would not only address temporal issues regarding the development of El
across ethnicities and their relation to smoking, but also provide stronger
conclusions regarding El and ethnicity interactions with actual smoking behaviors.
103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The resulting knowledge gained can be used to develop subsequently improved
adolescent smoking prevention programs.
As promising as these findings are, our sample of students were not yet of an
age where actual smoking behaviors might have been more prevalent (e.g., high
school). Our sample of sixth grade students did not exhibit high enough levels of
experimental smoking to detect large differences in analyses. Only about 5.5% of
students had ever tried smoking in our sample, while about 7.5% intended to smoke.
Sampling high school students, for example, may provide greater variability in
smoking behaviors. Replication of this current research on older samples may yield
more information about how El and culture/ethnicity interact to modify smoking
behaviors.
Our measure of El may not completely reflect all of its aspects, particularly
since we excluded the Assimilating Emotions branch. This branch involved
students generating various emotions and was excluded primarily due to time
constraints within classrooms. Furthermore, in our previous work we discovered
that this was a particularly difficult task to comprehend for those with limited
English proficiency or English as a second language. Though we acknowledge that
our abbreviation of the MEIS may have altered the generally accepted mental ability
model of El, factor analyses performed on our abbreviated version indicated that the
abbreviated version we employed was still very similar to Mayer, et al. (1999).
Work by Sullivan (1999) also utilized a similar abbreviated El measure and found it
to be reflective of the overall MEIS. Regardless, the MEIS is a relatively new
104
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
measure of El therefore not much published data exist as a way of comparison,
especially for adolescent samples.
In addition to controlling for demographic variables, our analyses controlled
for self-reported grades received in school. Self-reported grades was included as a
covariate in all of our analyses as a proxy for IQ because previous research has
shown El to be related to IQ (Mayer, et al., 1999). This is somewhat crude and may
not completely represent true intellectual level. A more comprehensive measure of
IQ would be desirable in future research.
Since our language-based measure of acculturation is associated with
ethnicity we included it in our statistical analyses as a covariate. The acculturation
process, however, is complex and a multi-item scale would have been a more
complete measure of acculturation. Although language usage is only one
component of the acculturation process, such single-item language measures have
been shown to correlate highly with more comprehensive acculturation scales,
accounting for a large proportion of the variance (Epstein, Botvin, Dusenbury, &
Diaz, 1996; Unger, Cruz, Rohrbach, et al., 2000).
Future Directions
Though our results are only suggestive that the relationship between El and
smoking intentions may vary by ethnicity, future work should examine these
findings across levels of acculturation (making use of more comprehensive
acculturation measures). The examination of El across various levels of
acculturation may be informative for multicultural smoking prevention programs
105
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
aimed at diverse populations than ethnicity alone. Examination of these
relationships across acculturation levels, in addition to ethnicity, would provide
valuable information. Future studies would benefit by recruiting larger samples in
order to more easily detect true variation across culture/ethnicities. Given that
smoking rates and risk factors differ by ethnicity, so might they differ by
acculturation level. Past research examining acculturation to the US culture, using a
language-based measure, revealed that US acculturated adolescents smoked more
than those who were not as acculturated (Unger, et al., 2000). Perhaps El and
acculturation interact to affect smoking. If acculturation to the US culture is a
smoking risk factor then perhaps El may offset or negate such risk. Therefore,
future work examining El’s interaction with acculturation may benefit by utilizing a
more comprehensive measure than was used in this study as a covariate, and be
more informative than examinations by ethnicity alone.
Implications
Evidence of the relationship between El and health behaviors is continuing
to increase. With the emerging trend within tobacco research towards positive and
protective factors, our findings regarding a trend towards the protective role of El
against smoking intentions is encouraging. Many adolescent smoking risk factors
cannot be changed, such as ethnicity, socioeconomic status, or having friends that
smoke. However, El is a modifiable factor that can be improved (i.e., it can be
taught) (Mayer & Salovey, 1997) and thus may help to curb future adolescent
smoking behaviors, particularly among White adolescents who are more likely to
106
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
smoke. Therefore as adolescent smoking prevention programs evolve, adding novel
El-enhancing components to future prevention programs may lead to increased
effectiveness. As the dynamics of the U.S. population shift and become
increasingly culturally diverse, it becomes increasingly important to identify
protective variables and design adolescent smoking prevention programs that will be
more effective for adolescents of diverse ethnic and cultural backgrounds.
107
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
Bar-On, R. (1997). The Emotional Quotient Inventory (EQ-il: Technical
Manual. Toronto, Canada: Multi-Health Systems.
Carvajal, S.C., Wiatrek, D.E., Evans, R.I., Knee, C.R., & Nash, S.G. (2000).
Psychosocial determinants of the onset and escalation of smoking: Cross-sectional
and prospective findings in multiethnic middle school samples. Journal of
Adolescent Health, 27. 255-65.
Chen. X. & Unger. J..B. (1999). Hazards of smoking initiation among
Asian American and non-Asian adolescents in California: a survival model analysis.
Preventive Medicine. 28, 589-599.
Chen, X., Unger, J.B., Cruz, T.B., & Johnson, C.A. (1999). Smoking
patterns of Asian-American youth in California and their relationship with
acculturation. Journal of Adolescent Health, 24, 321-328.
Ciarrochi, J. V., Chan, A. Y. C., & Caputi, P. (2000). A critical evaluation
of the emotional intelligence construct. Personality and Individual Differences, 28,
539-561.
Cooper, R.K. (1997). EQ Map. San Francisco: AIT and Essi Systems.
Defino, R.J., Jamner, L.D., & Whalen, C.K. (2001). Temporal analysis of
the relationship of smoking behavior and urges to mood states in men versus
women. Nicotine and Tobacco Research, 3, 235-48.
Epstein, J.A., Botvin, G.J., Dusenbury, L., & Diaz, T. (1996). Validation of
an acculturation measure for Hispanic adolescents. Psychological Reports, 79, 1075-
1079.
Flay, B.R., Hu, F.B., & Richardson, J. (1998). Psychosocial predictors of
different stages of cigarette smoking among high school students. Preventive
Medicine. 27,9-18.1998.
Golem an, D. (1995). Emotional Intelligence. New York: Bantam Books.
Hedlund, J. & Sternberg, R.J. (2000). Too many intelligences? Integrating
social, emotional, and practical intelligence. In R. Bar-On & J.D. Parker (Eds). The
Handbook of Emotional Intelligence (pp. 136-167). San Francisco: Jossey-Bass
Inc.
108
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Johnston, L.D., O'Malley, P.M., & Bachman, J.G. (1998). National survey
results on drug use from the Monitoring the Future study. 1975-1997. Volumne I:
Secondary school students. (NIH Publication No. 98-4345). Rockville, MD:
National Institute on Drug Abuse.
Legree, P. J. (1995). Evidence for an oblique social intelligence factor
established with a likert-based testing procedure. Intelligence. 21, 247-266.
Mayer, J. D., Caruso D. R, & Salovey, P. (1999). Emotional intelligence
meets traditional standards for an intelligence. Intelligence, 27, 267-298.
Mayer, J. D., Caruso D., & Salovey, P. (2000). Models of emotional
intelligence. In R.J. Sternberg (Ed.). Handbook of Intelligence (pp. 396-420).
Cambridge, UK: Cambridge University Press.
Mayer, J.D., & Salovey, P. (1997). What is emotional intelligence? In P.
Salovey & D. Sluyter (Eds). Emotional Development and Emotional Intelligence:
Implications for Educators (pp.3-31). New York: Basic Books.
Mayer, J. D., Salovey, P., & Caruso, D. R. (1997). Multifactor Emotional
Intelligence Scale, Student Version. Durham, New Hampshire: Authors.
O’Callaghan, F.V., Callan, V.J., & Baglioni, A. (1999). Cigarette use by
adolescents: attitude-behavior relationships. Substance Use & Misuse, 34,455-468.
Pallonen, U.E., Prochaska, J.O., Velicer, W.F., Prokhorov, A.V., & Smith,
N.F. (1998). Stages of acquisition and cessation for adolescent smoking: an
empirical integration. Addictive Behaviors. 23. 303-24.
Pierce, J.P., Choi, W.S., Gilpin, E.A., Farkas, A.J., & Merritt, R.K. (1996).
Validation of susceptibility as a predictor of which adolescents take up smoking in
the United States. Health Psychology, 15,355-61.
Rice, C.L. (1999). A quantitative study of emotional intelligence and its
impact on team performance. Unpublished master’s thesis, Pepperdine Unidersity.
Rubin, M.M. (1999). Emotional intelligence and its role in mitigating
aggression: A correlational study of the relationship between emotional intelligence
and aggression in urban adolescents. Unpublished manuscript, Immaculata College,
hnmaculata, PA.
SAS Institute [Computer software]. (200). Cary, North Carolina: SAS
Institute.
109
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Sullivan, A. K. (1999). The Emotional Intelligence Scale for Children.
Dissertation Abstracts International. Unpublished doctoral dissertation.
Triandis, H.C. (1995). Individualism and collectivism. Boulder, CO:
Westview Press.
Trinidad, D.R. & Johnson, C.A. (2001). Is low emotional intelligence a risk
factor for adolescent tobacco and alcohol use? Poster Presented at the Society for
Behavioral Medicine and Society for Research on Nicotine and Tobacco Joint
Annual Meeting.
Trinidad & Johnson (2002). The Association Between Emotional
Intelligence and Early Adolescent Tobacco and Alcohol Use. Personality and
Individual Differences, 32, 95-105.
Trinidad, D.R., Unger, J.B., Chou, C.P., Azen, S.P., & Johnson, C.A.
(2002a). The protective association of emotional intelligence with psychosocial
smoking risk factors for adolescents. Manuscript submitted for publication.
Trinidad, D.R., Unger, J.B., Chou, C.P., Azen, S.P., & Johnson, C.A.
(2002b). Emotional intelligence and smoking risk factors in adolescents:
Interactions on smoking intentions. Manuscript submitted for publication.
Unger, J.B., Cruz, T.B., Rohrbach, L.A., Ribisl, K.M., Baezconde-
Garbanati, L., Chen, X., Trinidad, D.R., & Johnson, C.A. (2000). English language
use as a risk factor for smoking initiation among Hispanic/Latino and Asian
American adolescents in California: Evidence for mediation by tobacco-related
beliefs and social norms. Health Psychology, 19, 403-410.
110
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 5:
Emotional intelligence and acculturation to the US:
Interactions on the perceived social consequences of smoking in early adolescents
This study was supported by Grant Number 9DT-0174 from the Tobacco
Related Diseases Research Program of the University of California, Office of the
President, and Grant Number 5 P5 CA 84735 from the National Cancer Institute.
All correspondence regarding this manuscript should be directed to Dennis R.
Trinidad, Institute for Health Promotion and Disease Prevention Research and
Department of Preventive Medicine, Keck School of Medicine, University of
Southern California, 1000 S. Fremont Ave., Unit 8, Alhambra, CA 91803. Phone:
626-457-4163. Fax: 626-457-4012. Email:dtrinida@usc.edu.
Abstract
Encouraging empirical research has shown that high emotional intelligence
(El) is associated with decreased adolescent health risk behaviors, including
smoking and its risk factors. As this exciting field of research continues to grow,
the relationship of El with adolescent smoking requires further clarification,
particularly in that high El may buffer the relationship between acculturation to the
US culture and perceptions of the social consequences of smoking (PSC). If so,
adolescents with high El may better benefit from culturally tailored, social-
influences based tobacco prevention programs targeting such factors.
I ll
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
El is defined as the ability to: accurately perceive, appraise, and express
emotion; access and/or generate feelings in facilitating thought; understand emotion
and emotional knowledge; and regulate emotions. A measure of El was
administered to a subset of students from the control condition of the baseline
portion of a longitudinal school-based experimental trial of smoking prevention
strategies in a multicultural, urban population of adolescents in Southern California.
El was assessed with a shortened version of the Multifactor Emotional Intelligence
fh
Scale, Adolescent Version, and was administered to 416 6 graders (53% girls)
from middle schools in the Los Angeles area (mean age-11.3 yrs; 32%
Hispanic/Latino (H/L), 29% Asian/Pacific Islander (A/PI), 13% White, 19%
Multiethnic, 6% Other).
As hypothesized, there was a significant El X US acculturation interaction
(p<0.01) suggesting that the protective effect of El on PSC is stronger for those
more US oriented vs. less US oriented. Our results lend support to the mounting
evidence for emotional intelligence as a protective factor against adolescent
smoking risk factors in multicultural settings and indicate that El buffers PSC for
those acculturated to the US culture. As the dynamics of the US population shift
and become increasingly culturally diverse, it becomes increasingly important to
identify protective variables and design adolescent smoking prevention programs
that will be more effective for adolescents of diverse ethnic and cultural
backgrounds.
112
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Introduction
Recent lay and academic interest in emotional intelligence (El), or the
interaction of emotion and cognition, stems from claims by the popular media that
El may possibly be the most important predictor of life success, accounting for up to
80% of that variance (Goleman, 1995). It is improbable that the effect of El could
be of such magnitude in most life domains. Nonetheless, encouraging empirical
research has shown that high El is associated with decreased adolescent health risk
behaviors, including tobacco and alcohol use and aggression (Rubin, 1999; Trinidad
& Johnson, 2002). As this exciting field of research continues to grow, the
relationship of El with specific health behaviors, such as adolescent smoking,
requires further clarification. We are particularly interested in the hypothesis that
high El may buffer the relationship between acculturation to the US culture and
adolescent smoking risk factors, especially perceptions of the social consequences
of smoking. If so, adolescents with high El may better benefit from culturally
tailored, social-influences based tobacco prevention programs targeting such factors.
Many risk factors for adolescent smoking have been identified. However,
one of the more interesting risk factors for future smoking is acculturation to the US
culture. Across various ethnic groups, acculturation to the US culture has been
identified as a risk factor for health-compromising behaviors in adolescents,
including smoking (Unger, Cruz, Rohrbach, et al., 2000; Epstein, Botvin, & Diaz,
1998). As people of varying cultural backgrounds, such adolescents from non-US
cultures, come into contact with the US culture, an interchange of cultural attitudes
113
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and behaviors can occur generally referred to as acculturation (Berry, 1980). This
contact and the resulting attempt to coexist with and/or conform to the new culture,
can cause individuals to change their social behavior, attitudes and beliefs. One of
the behaviors that might change as acculturation to the US occurs is smoking
(Unger, et al., 2000). As students acculturate to the US culture and attempt to
navigate between the two cultures, they may perceive smoking behavior less as a
social consequence and perhaps more as socially neutral and accepted, or beneficial
and desirable, thus putting themselves at an increased risk for smoking.
One of the smoking risk factors frequently associated with future smoking
behaviors is low perceptions of the negative social consequences associate with
smoking. As such, increasing perceptions of negative social consequences
associated with smoking (PSC) are generally targeted by social-influences based
tobacco prevention programs (Hansen & Graham, 1991; MacKinnon, Johnson,
Pentz, et al., 1991). Encouragingly, our recent research has found that those with
high El perceive greater negative social consequences associated with smoking
(Trinidad, Unger, Chou, Azen & Johnson, 2002a). Adolescents we sampled with
high El were more likely to view smoking as not a grown up behavior, perceived
that smokers did not have more friends, and may have associated smoking with
negative emotional states/consequences that result from its social costs. We suspect
that high El buffers the risk of perceiving low social consequences associated with
smoking more for those who are acculturated to the US culture versus those not US
acculturated. Specifically, we hypothesize that high US acculturation will be
114
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
associated with decreased PSC, and that high El will increase PSC for those
acculturated to the US culture. Put another way, the association between El and
PSC will be stronger for those high in US acculturation than for those not US
acculturated.
Past research on El and adolescent smoking indicate high El to be protective
a protective factor. El has been shown to be protective against adolescent smoking
behaviors (Trinidad & Johnson, 2002), as well as some psychosocial smoking risk
factors (Trinidad, et al., 2002a). Trinidad and Johnson (2002) also found evidence
for El as a protective factor against smoking behaviors in a sample of 7th and 8th
grade students, and that low El adolescents were over two times more likely to have
engaged in smoking behaviors (Trinidad & Johnson, 2001). Another study utilizing
data from 6th graders further suggested that high El was related to lower risks for
psychosocial smoking risk factors, including decreased intentions to smoke in the
next year, increased refusal self efficacy, and increased perceptions of negative
social consequences of smoking (Trinidad, et al., 2002a). Subsequent research on
this 6th grade sample also revealed that students with low El were more likely to
intend to smoke if they had low refusal self-efficacy or were more hostile, and that
for low El adolescents, intentions were not as likely to translate into use (Trinidad,
Unger, Chou, Azen & Johnson, 2002b). More recent research examined the
association between smoking intentions and El across ethnicity (Trinidad, Unger,
Chou & Johnson, 2002). Trinidad and colleagues (2002) uncovered only a trend in
the difference in associations between El and smoking intentions in Whites,
115
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Asian/Pacific Islanders and Hispanic/Latinos. Specifically, the trend detected was
that El might have been more protective against smoking intentions in the next year
for White adolescents relative to A/PI and H/L adolescents (p<0.10). An
exploration of the relationship between El, acculturation to the US culture, and PSC
may yield additional valuable information about El as a protective factor across
cultures, particularly since US acculturation is associated with smoking after
controlling for ethnicity (Unger, et al., 2000), and given that previous research on El
and ethnicity did not yield statistically significant results (Trinidad, et al., 2002).
Two differing models conceptualizing El exist: a mental ability model and a
mixed model. The mental ability model is a new construct that focuses primarily on
emotion and its interactions with cognition (Mayer, Caruso & Salovey, 2000). This
model of El is conceptualized as the “ability to perceive accurately, appraise, and
express emotion; the ability to access and/or generate feelings when they facilitate
thought; the ability to understand emotion and emotional knowledge; and the ability
to regulate emotions” (Mayer & Salovey, 1997, p. 10). Mixed models of El differ
from ability models in that mixed models combine both mental abilities with
traditional traits, such as optimism, motivation or mood (Goleman, 1995; Bar-On,
1997). The mental ability model is a more focused conceptualization emphasizing
mental skills, as opposed to variables already known to be associated with smoking,
such as optimism and mood (Carvajal, Wiatrek, Evans, Knee & Nash, 2000; Defino,
Jamner & Whalen, 2001; Pallonen, Prochaska, Velicer, Prokhorov & Smith, 1998).
Therefore, this research will utilize the mental ability model because it will provide
116
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
an opportunity to expand our understanding of adolescent smoking and its
precursors.
Previously found associations between acculturation and adolescent smoking
indicate that it is important to gain information on the role that El may play in
mitigating this link. Understanding the differences of associations between El and
perceived social consequences of smoking across US acculturation level may help
in designing improved cross-cultural smoking prevention programs for adolescents.
The potential to improve smoking prevention programs is increasingly important
with the California population, particularly Los Angeles, becoming increasingly
diverse.
Methods
The data described in this article are from the baseline portion of a
longitudinal school-based experimental trial of smoking prevention strategies in a
multicultural, urban population of adolescents in Southern California. Latino and
Asian adolescents were oversampled. The purpose of the baseline survey was to
assess tobacco use and related psychosocial and cultural variables before the
implementation of culturally relevant smoking prevention programs. A measure of
emotional intelligence was administered to a subset of students from the control
condition of the main prevention trial. Analyses were performed on this subset of
students.
117
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Subjects
Subjects were 416 actively consented sixth-grade adolescents from public
middle schools in the greater Los Angeles area. This sample population had a mean
age 11.3 years and was 53% male. The ethnic distribution was 32% Latino, 29%
Asian/Pacific Islander, 13% White, 19% Multiethnic, and 6% Other. Students were
considered Latino if they answered "yes" to the question, "Are you
Hispanic/Latino?" Similar coding schemes were utilized for Asians and Whites.
Due to the small number of African Americans and other ethnicities, students who
were not Latino, Asian or White were considered Other. Students who were more
than one ethnicity, including Other, were coded as Multiethnic. Table 1 summarizes
the age, gender and ethnic characteristics of the study participants.
Procedure
Students completed the baseline questionnaire, a 160-item paper-and-pencil
survey, in their classrooms during a single class period (45-50 minutes). Trained
data collectors, who were not previously acquainted with the students, distributed
the surveys. The surveys were identified only by a code number, not with the
students’ names or any other identifying information. Because the students all were
attending English-language schools in which their classes were conducted only in
English (California state law prohibits bilingual education in public schools), a basic
level of English-language proficiency was assumed and the surveys were provided
only in English. However, students were encouraged to ask the data collectors to
clarify the meanings of any unfamiliar words.
118
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
During another classroom session several weeks after the baseline survey,
students completed a measure of El (about 45-50 minutes). Students completed this
survey individually with the lead author reading each item, including examples, out
loud to the students. This allowed for the maintenance of uniformity in survey
administration and the minimization of differences in participation and reading
skills.
Measures
Emotional Intelligence.
The Multifactor Emotional Intelligence Scale, Adolescent Version (MEIS;
Mayer, Salovey & Caruso, 1997) was employed to assess the El of the participants.
A thorough description of this test and how it is scored can be found in Mayer,
Caruso & Salovey (1999). Briefly, the MEIS is a competence-based measure
consisting of 4 branches assessing an individual’s ability to perceive, assimilate,
understand, and manage emotion in himself/herself and in others. In this study, the
MEIS was shortened by excluding the Assimilating Emotions branch (assessing
one’s ability to generate emotions) and including only the Emotional Identification,
Understanding Emotions, and Managing Emotion branches. The Assimilating
Emotions branch was excluded due to time restrictions at schools. Exploratory
factor analysis revealed that this abbreviated measure of El, consisting of these 3
branches of El, still formed an overall El factor similar to Mayer et al.’s (1999)
overall El model and one used in previous work by Trinidad and Johnson (2002). A
119
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
critical evaluation of the construct of El, as measured by the MEIS, conducted by
Ciarrochi, Chan, and Caputi (2000) showed the MEIS to be a valid measure of EI.
The Emotional Identification branch was assessed with the Faces, Music,
and Stories subtests; the Designs subtest was excluded. Each of these subtests
presented a stimulus to the student (e.g., a picture of a face depicting certain
emotions, a musical selection, or a short vignette) and asked him/her to rate whether
a specific emotion was present in each stimulus. Possible responses were rated on a
5-point Likert-type scale, ranging from “Definitely Not Present” (1) to “Definitely
Present” (5).
The Understanding Emotions branch was assessed with the Relativity
subtest. This test consisted of items describing a conflicting social situation
between two persons. Students were asked to estimate how likely each character
experienced certain feelings. Possible responses were rated on a 5-point Likert-type
scale, ranging from “Extremely Unlikely” (1) to “Extremely Likely” (5).
The Managing Emotions branch was assessed with the Managing subtest.
This test assessed how well students were able to manage their own and others’
emotions by asking them to evaluate the effectiveness of certain courses of action
given a hypothetical situation. Possible responses were rated on a 5-point Likert-
type scale, ranging from “Extremely Ineffective” (1) to “Extremely Effective” (5).
Consensus scoring was utilized in grading the MEIS. This is the standard
scoring method based on norms of reference groups studied by Mayer et al. (1999).
Since Latino and Asian students were underrepresented in Mayer et al.’s sample
120
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
used to create reference norms, we computed norms based on the current sample.
Following the procedure described by Mayer et al., each student’s response was
weighted by the proportion of the sample giving the same response. The resulting
score indicated the percent agreement with the population from which the subject
was drawn. For example, if 75% of the students sampled in this study reported that
happiness was “Definitely Present” in a picture of a face (“5” on the Likert-scale),
then a student who chose “5” would receive a score of 0.75 for the item. Sub-totals
for each branch were summed. Branch totals were then summed to obtain an
aggregate MEIS score.
Acculturation.
To measure acculturation, we used the recently developed AHIMSA (Unger,
Gallaher, Shakib, Ritt-Olson, Palmer & Johnson, in press). AHIMSA is an acronym
that stands for Acculturation, Habits and Interests Multicultural Scale. This novel
measure of acculturation is a multi-dimensional scale written to assess aspects of
ethnic interaction, cultural heritage, and ethnic behaviors from the perspective of an
early adolescent. It was developed for use in a new, large, randomized-controlled
adolescent smoking prevention curriculum in multicultural settings for adolescents
of varying ethnicities (the mother study of this current research). Therefore it is
appropriate for use in this study. The AHIMSA is a more appropriate and
comprehensive measure of acculturation than the language-based acculturation
measures we have used in the past (Trinidad, et al., 2002a, Trinidad, et al., 2002b,
121
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Trinidad, et al., 2002). A more detailed description of this scale can be found
elsewhere (Unger, et al., in press). We will only provide a brief description here.
Sample items from the 8-item AHIMSA scale include, for example, “I am
most comfortable being with people from...,” “My best friends are from.. and
“The food I eat at home is from...” Forced-choice response options were, “The
United States,” “The country my family is from,” “Both,” and “Neither.” The
AHIMSA scale generates four scores based on the four responses: US Orientation
(the total number of “United States” responses); Other Country Orientation (the
total number of “The country my family is from” responses); Biculturalism (the
total number of “Both” responses); and Neither Country Orientation (the total
number of “Neither” responses). The total number of “The United States”
responses indicates the respondent’s United States Orientation score. The score for
each orientation can range from 0 to 8. Because of the forced-choice format, the
sum of the four orientation scores always will equal eight (the total number of
statements on the scale). This renders it not possible to include all four orientation
scores as independent variables in the same multiple regression model. This is
because the fourth score always will equal eight minus the sum of the other three
orientation scores, thus creating a linear dependence among the independent
variables in the model.
Unger and colleagues (in press) recommend that researchers use one or more
of the four orientation scores depending on their research questions. For this current
research, because of our hypothesis, we used the US Orientation score. Based on
122
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Unger, et al.,’s (in press) work those who were low on US Orientation were most
likely to be high on Biculturalism. Previous work validating the AHIMSA by
Unger and her colleagues revealed a strong negative correlation (r=-0.86) between
those high in US orientation and those high in biculturalism. The authors concluded
in their preliminary work that there was not much variance in the Other Country
Orientation scale and in the Neither Country Orientation scale. The current study
sample is different from the Unger et al. AHIMSA sample; however students in our
current sample were from the same geographic region (Los Angeles area). To
validate this particular finding by Unger, et al., we calculated correlation
coefficients from students in our sample for all of the AHIMSA subscales. We
found an almost identically strong negative correlation (r=-0.85) between those high
in US orientation and those high in Biculturalism as well as the small variance in the
Other Country and Neither Country Orientation scales. Such strong negative
correlations suggest that a student high in US orientation is highly likely to be low
in biculturalism. Therefore our hypotheses are made, and based only, on the US
Orientation subscale of the AHIMSA. In this study when we refer to AHIMSA, we
refer to its US Orientation scale.
Perceived Social Consequences o f Smoking.
Perceived social consequences of smoking (PSC) was an index score
comprised of eight variables measuring a student’s perception of the social
outcomes of smoking (alpha=0.66). Variables included items such as “Smoking
cigarettes is one way to lose friends who are nonsmokers” and “Kids who smoke are
123
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
more grown up” and “Smoking cigarettes makes kids look cool.” Possible
responses were rated on a 4-point Likert-type scale with high scores indicating a
greater perception of negative social consequences (i.e., high scores are more
desirable). The range of responses was 11 to 32. This measure was contained
within the baseline tobacco use survey.
Covariates.
Relevant covariates included several demographic variables, including age,
gender, ethnicity, socioeconomic status (ratio of number of people per room in a
home), English language use (5-point Likert-type scale of language spoken at home
ranging from "English only" to "Only another language"), and self-reported
assessment of overall grades received in school last year (e.g., mostly A’s, mostly
A’s & Bs, mostly B’s, etc.). English language use was included in analyses because
it is correlated with the US Orientation and Biculturalism scales of the AHIMSA.
Self-reported grades was included in analyses as a proxy for IQ. Previous research
has shown El to be related to IQ (Mayer, et al., 1999).
Statistical Analyses
Multiple regression analyses were performed to assess the interaction
between El and the AHIMSA US Orientation scale on PSC. Analyses included
main effects for El, main effects for the AHIMSA US Orientation scale, and El X
AHIMSA US Orientation scale interaction terms. The interaction variable was
created by multiplying the centered El variable with centered AHIMSA variable.
Analyses controlled for age, gender, socioeconomic status, language spoken at
124
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
home, and self-reported grades received in school. All analyses were performed
using SAS version 8.1 (SAS Institute, 2000).
Results
Demographic characteristics of respondents
Table 10 summarizes the age, gender and ethnic characteristics of the study
participants. The sample population of 416 adolescents had a mean age 11.3 years
and was 53% female. The ethnic distribution was 32% Hispanic/Latino, 29%
Asian/Pacific Islander, 13% White, 19% Multiethnic, and 6% Other.
Table 10. Demographic Characteristics_______
Frequency (%)
Gender
Male 194 46.8
Female 221 53.2
Age
10 3 0.7
11 300 72.5
12 110 26.6
13 1 0.2
Ethnicity
White 53 12.7
Hispanic/Latino 135 32.3
Asian/PI 121 29.1
Multi-ethnic 80 19.2
Other 27 6.5
Gender differences in El, US Orientation and PSC
Unadjusted mean score on the MEIS for girls was significantly higher than
for boys (59.16 vs. 57.14; p=0.03). Girls reported receiving higher grades in school
than boys (p<0.01). No other significant differences by gender were found in PSC
125
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
or the AHIMSA US Orientation. These results are summarized at the bottom of
Table 11.
Ethnic differences in El. US Orientation and PSC
Table 11 also shows the ethnic differences in MEIS, US Orientation and
PSC scores. A/PI students scored higher on the MEIS than all other students
(p<0.05). White students scored higher on US Orientation than all other students;
A/PI students scored lower on US Orientation than all other students, as did H/L
students (except vs. Other ethnicity) (p<0.05). A/PI students perceived greater
social consequences associated with smoking than students of Other ethnicity
(p<0.05).
Table 11. Mean scores for Emotional Intelligence, AHIMSA
US Orientation, and Perceived Social Consequences
_________of Smoking across Ethnicity and Gender__________
Emotional
Intelligence
US Orientation Perceived Social
Consequences
Ethnicity
White 58.93 5.74b 29.73
H/L 57.23 4.03b 28.88
A/PI 60.523 3.28b 29.47c
Multi-ethnic 56.85 4.64 28.81
Other 55.29 4.42 27.63c
Gender
Girl 59.16 3.98 29.35
Boy 57.14 4.43 28.81
t-value=2.24* t-value—1.92 t-value=1.53
Note: Higher scores mean higher El, more US Orientation, more Biculturalism, more PSC
*p<0.05
a A/PI scored higher on the MEIS than all other students (p<0.05).
b Whites were higher on US Orientation than all other students; A/PI were lower on US Orientation
than all other students, as were H/L (except vs. Other ethnicity) (p<0.05).
c A/PI perceived greater social consequences associated with smoking than students of Other
ethnicity (p<0.05).
126
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Acculturative variation in the association between MEIS score and PSC
Table 12 contains information from a regression models and shows the
association between emotional intelligence, AHIMSA US Orientation scale, and
PSC from multiple regression analyses. All analyses are controlled for age, gender,
socioeconomic status, language spoken at home, and self-reported grades received
in school. The numbers shown in the table are standardized Beta coefficients.
Table 12. US Acculturation differences in Emotional Intelligence on
________ Perceptions of Negative Social Consequences of Smoking
Perceived Negative Social
Consequences of Smoking
________________________________________Std. { 3 ______ p-value
US Orientation main effect -.126* .031
Emotional Intelligence main effect .166* .003
US Orientation X Emotional Intelligence_____ .183*________ .001
Note. Estimates are controlled for age, gender, SES, language spoken at home,
and grades received in school.
*p<.05
Main effects of emotional intelligence. Higher motional intelligence score
on the MEIS was associated with greater negative PSC in the model testing for an
interaction between US Orientation and El (Model 1; p<0.01).
Main effects of AHIMSA US Orientation scale. Orientation to the US
culture was negatively associated with PSC (p<0.05).
Interactions between El and US Orientation scale on PSC. As hypothesized,
the El X US Orientation interaction term was significant (p<0.01) suggesting that
the protective effect of El on PSC is stronger for those more US oriented vs. less US
oriented.
127
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
For illustrative purposes only, and to gain a more clear understanding of
these interactions, we graphed this relationship (represented in Figure 7). For ease
of interpretation of the resulting graph we performed median splits of the DV (PSC)
and the IVs (High El, Low El; High US Orientation, Low US Orientation). These
median-split IVs were then centered and multiplied to form interaction terms. We
then performed logistic regression analyses to obtain odds ratios for the resulting
graph (again for ease of interpretation and illustration). Figure 7 illustrates that
those who are more US acculturated are less likely to perceive social consequences
of smoking (bottom line in Figure 1) than those not as acculturated to the US (top
line in Figure 7)
Figure 7. Odds Ratio of Perceiving Social Consequences of Smoking by
Emotional Intelligence Level by US Acculturation
« L2
® 1
| 0.8
§f 0.6
05
a
3 ° - 4
1 0.2
©
High Low
El Level
i US “ • “ Hi US
128
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Discussion
There is now mounting evidence for emotional intelligence as a protective
factor against adolescent smoking and smoking risk factors (Trinidad & Johnson,
2002; Trinidad, et al., 2002abc). Our results lend support to this and indicate that El
buffers risk factors for smoking, particularly perceived social consequences, for
those acculturated to the US culture. In other words, acculturation to the US culture
is a known risk factor for adolescent smoking (Unger, et al., 2000), as was shown in
our current results. However, as El increases then PSC increases for those who are
US acculturated. The buffering effect of El on PSC is stronger for those more US
oriented versus less US oriented. Examining our illustrative graph to gain a more
clear understanding of these interactions simplifies our results even further. Figure
1 shows that those who are more US acculturated are less likely to perceive social
consequences of smoking than those not as acculturated to the US. As El increases
for US acculturated students, their odds of perceiving negative social consequences
of smoking increases.
The mental ability skills of El (emotional identification, understanding,
management) may contribute to the increased perceptions of social consequences,
particularly for those are US acculturated. Past research has shown that those with
high El perceived greater negative social consequences associated with smoking
(Trinidad, et al. 2002). But past research has also shown that those more US
acculturated were more likely to smoke (Unger, et al., 2000; Eptstein, et al., 1998).
These highly US acculturated students may have smoked more in part because their
129
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
perceptions of the negative social consequences of smoking were low. However,
our findings suggest that for those who are US acculturated, having high El
increases views of smoking as characteristic of more delinquent behavior as
opposed to being a more grown up behavior, and perceptions that smokers do not
have more friends, or the friends that smokers have are not the friends that they
would want for themselves. These students may also more likely associate smoking
with negative emotional states/consequences that result from its social costs. Those
who were US acculturated with low El, on the other hand, may not be as perceptive
of such social consequences. Furthermore, attempts by adolescents to conform to
the US culture might have led to increased stress, which might have been mitigated
by high EL In other words, high El increases PSC for those acculturated to the US
culture and may help alleviate acculturative stress. Increasing perceptions of
negative social consequences associated with smoking are generally targeted by
social-influences based tobacco prevention programs (Hansen & Graham, 1991;
MacKinnon, et al., 1991) and our findings suggest that adding an El component to
social-influences based prevention programs may increase their efficacy,
particularly for those who are US acculturated.
To add support for our main finding, we performed analyses using the
Biculturalism scale of the AHIMSA. As mentioned above, students high in US
Orientation were very likely to be low in Biculturalism and vice versa because both
of these variables were very negatively correlated (r=-0.85). Results from such
analyses were as expected, that those not acculturated to the US culture perceived
130
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
greater PSC, and that high El increased PSC for those who were US acculturated.
Furthermore, the strong negative correlation between these two of the four scales on
the forced-choice AHIMSA measure suggest that the variances for the Neither and
Other Country Orientation scales of the AHIMSA would be limited. This is why
our hypotheses were made, and based only, on the US Orientation subscale of the
AHIMSA.
Our analyses also controlled for English language use because English
language use is related to the AHIMSA (Unger, et al., in press) and the items in the
MEIS might have favored those more acculturated to the US and thus more
proficient in English (i.e., being more understandable and relatable to US
acculturated adolescents). We also performed additional analyses using an El X
English Language Use interaction term but found no significant interactions.
Though English Language Use and the AHIMSA subscales are correlated, that we
detected interactions using the AHIMSA suggests that the AHIMSA captures
variance unique from, perhaps additional to, English Language Use.
Finally, it should also be noted that our results remained significant even
when controlling for ethnicity, but for sake of model parsimony we decided to not
include the ethnicity dummy variables in our final, presented results. Our findings
by acculturation in this study build on our previous work that detected a trend
toward significance for El as a protective factor against smoking intentions in
Whites. Results by acculturation reveal additional information to results by
131
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ethnicity, suggesting that acculturative factors play a unique but related role in the
El—smoking relationship.
Possible Mechanisms
In line with previous research (Trinidad et al., 2002ab) low El was
associated with an increased risk for smoking risk factor variables, PSC in this case.
Those with low El may be cast aside from the mainstream of social group
interactions, especially difficult circumstances during adolescence. The alienated
adolescent may then feel depressed (Tani, Chavez. Deffenbacher, 2001). As the
result, he or she may self-medicate to feel better (Carmody, 1989). In other words,
those with low El might be more prone to social isolation and thus use tobacco as a
method of coping. Alternatively, the social isolate may increase efforts to seek out
peers who would accept him or her. The accepting peers may be relatively tolerant
smokers, with the criterion for acceptance being tobacco or drug use rather than a
certain level of social or emotional capability. These are two potential, related
pathways that may lead to smoking in adolescents with low EL
Validity of El Scoring Method
Emotional intelligence hinges on the conception that certain emotional tasks
or problems have answers that can be judged either correct or incorrect. To obtain
the norms used to score the MEIS consensus scoring was employed. This was done
for several reasons. First, to some extent, emotions are socially based (Mayer &
Salovey, 1997). Second, research has shown that when groups of people’s
judgments are pooled together for items such as these Likert-based MEIS items, the
132
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
groups tend to serve as expert judges (Legree, 1995). Third, using various scoring
methods (consensus, expert and target) Mayer, et al. (1999) found that consensus
scoring produced the most reliable measures of EL Mayer, et al.’s expert scoring
involved using answers on the MEIS deemed to be correct by experts in emotion
research. Target scoring involved obtaining correct answers from persons depicted
in each item. For example, for identifying emotions present in a photograph of a
young girl crying (in the Emotional Identification branch), the girl whose photo was
taken was asked what emotions she was feeling. Answers from these three scoring
methods were moderately and highly positively correlated, indicating that indeed a
set of answers that were “more correct” than others exist for items and stimuli such
as those contained within the MEIS.
Individual El scores are weighted by the sample population and thus suggest
that the association between El and smoking risk factors may differ for a multi
ethnic sample relative to a more homogeneous one. However, associations between
El and smoking risk factors remained significant when we utilized norms from
Mayer et al.’s (1997) predominantly White (79%), suburban, adolescent sample.
This suggests that El may be a construct that cuts across cultures. In this case, it is
possible that perceptions and actions considered emotionally intelligent among
adolescents from White, suburban cultures are very similar to what would be
considered emotionally intelligent among multi-ethnic cultures.
Roberts, Zeidner and Matthews (2001) assert that it may be inappropriate, at
least in the realm of emotions, to insist that items assessing emotional capabilities
133
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
have absolute correct/incorrect answers. Mayer and colleagues, however, do not
make such strong claims and maintain that some answers to emotional items are
only more correct than others. Thus, an alternative method of conceptualizing
emotional skills falling under the umbrella of emotional intelligence may be more
aptly termed “emotional competence”. Finally, consideration of alternative scoring
methods, such as an expert consensus versus the sample consensus may also diffuse
some of the criticisms regarding validity of the MEIS’s scoring (Roberts et al.,
2001).
Limitations
As research into El’s relationship with health behaviors, particularly
adolescent smoking, is still in its infancy further studies need to be performed in
order to validate the findings of this research. Future research regarding El and
adolescent smoking may be conducted in the context of a longitudinal, culturally
tailored, school-based prevention program in order to address whether those with
high El do actually benefit more from such programs, and whether a program to
increase El would help prevent smoking across cultures. Furthermore, longitudinal
studies would not only address temporal issues regarding the development of El in
conjunction with acculturation across ethnicities and their relation to smoking, but
also provide stronger conclusions regarding El and acculturation/ethnicity
interactions with actual smoking behaviors. The resulting knowledge gained can be
used to develop subsequently improved adolescent smoking prevention programs.
134
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
As promising as these findings are, our sample of students were not yet of an
age where actual smoking behaviors might have been more prevalent (e.g., high
school). Our sample of sixth grade students did not exhibit high enough levels of
experimental smoking to detect large differences in analyses. Only about 5.5% of
students had ever tried smoking in our sample, while about 7.5% intended to smoke.
Sampling high school students, for example, may provide greater variability in
smoking behaviors (though we hope not). Replication of this current research on
older samples may yield more information about how El and culture/ethnicity
interact to modify smoking behaviors.
Our measure of El may not completely reflect all of its aspects, particularly
since we excluded the Assimilating Emotions branch. This branch involved
students generating various emotions and was excluded primarily due to time
constraints within classrooms. Furthermore, in our previous work we discovered
that this was a particularly difficult task to comprehend for those with limited
English proficiency or English as a second language. Though we acknowledge that
our abbreviation of the MEIS may have altered the generally accepted mental ability
model of El, factor analyses performed on our abbreviated version indicated that the
abbreviated version we employed was still very similar to Mayer, et al. (1999).
Work by Sullivan (1999) also utilized a similar abbreviated El measure and found it
to be reflective of the overall MEIS. Regardless, the MEIS is a relatively new
measure of El therefore not much published data exist as a way of comparison,
especially for adolescent samples.
135
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In addition to controlling for demographic variables, our analyses controlled
for self-reported grades received in school. Self-reported grades was included as a
covariate in all of our analyses as a proxy for IQ because previous research has
shown El to be related to IQ (Mayer, et al., 1999). This is somewhat crude and may
not completely represent true intellectual level. A more comprehensive measure of
IQ would be desirable in future research.
Implications and Future Directions
Despite these limitations, evidence of a protective link between El and
smoking-related factors is continuing to increase. With the emerging trend within
tobacco research towards positive and protective factors, our findings regarding a
trend towards the protective role of El against smoking intentions is encouraging.
Many adolescent smoking risk factors cannot be changed, such as ethnicity and
socioeconomic status. However, El is a modifiable factor that can be improved (i.e.,
it can be taught) (Mayer & Salovey, 1997) and thus may help to curb future
adolescent smoking behaviors, particularly for those who are US acculturated.
Based on the results of this and previous studies, it is plausible that the
plateau recently experienced by social-influences based programs may be overcome.
El is a novel variable that can be informative regarding improving social-influences
based smoking prevention programs. Mayer and his colleagues assert that emotions
are socially based (Mayer et al., 1999). Therefore, programs that center around
social processes and emphasize improving emotional skills may lead to reductions
in smoking beyond basic social-influences based programs. Improved emotional
136
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
skills may improve adolescents' processing of social-influences based prevention
information and thus build on standard self-efficacy techniques. For example,
existing and previously successful social-influences based smoking prevention
programs such as Project SMART (Graham. Johnson. Hansen. Flay. & Gee. 1990),
can be improved by taking El into account. Modules within such programs
emphasizing the detection of motivations behind peer pressures to smoke increase
emotional awareness (a component of El). Such program modules can be expanded
upon and/or more strongly emphasized. For example, pointing out that a solicitor
employing social pressures or teasing tactics might be doing so in order to feel
better about himself or make himself look good in front of his friends.
Furthermore, identifying and targeting low El adolescents is a reasonable
goal given these and previous findings (Trinidad et al., 2002ab). Those with high El
appear to be more protected against smoking risk factors. Therefore, administration
of an El survey at the baseline data collection phase of a smoking prevention
program can help identify higher risk (i.e., low El) adolescents for targeted, tailored
interventions. For example, raising El level for US acculturated adolescents could
increase perceptions of the negative social consequences of smoking.
As adolescent smoking prevention programs evolve, adding novel EI-
enhancing components to future prevention programs may lead to increased
effectiveness. As the dynamics of the U.S. population shift and become more
culturally diverse, it becomes increasingly important to identify protective variables
137
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and design adolescent smoking prevention programs that will be more effective for
adolescents of diverse ethnic and cultural backgrounds.
138
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
Bar-On, R. (1997). The Emotional Quotient Inventory (EQ-i): Technical
Manual. Toronto, Canada: Multi-Health Systems.
Berry, J.W. (1980). Acculturation as varieties of adaptation. In A. Padilla
(Ed.), Acculturation: theory, models and new findings (pp. 9-25). Boulder, CO:
Westview.
Carmody, T.P. (1989). Affect regulation, nicotine addiction, and smoking
cessation. Journal of Psychoactive Drugs, 21(3), 331-342.
Carvajal, S.C., Wiatrek, D.E., Evans, R.I., Knee, C.R., & Nash, S.G. (2000).
Psychosocial determinants of the onset and escalation of smoking: Cross-sectional
and prospective findings in multiethnic middle school samples. Journal of
Adolescent Health, 27, 255-65.
Ciarrochi, J. V., Chan, A. Y. C., & Caputi, P. (2000). A critical evaluation
of the emotional intelligence construct. Personality and Individual Differences, 28,
539-561.
Defino, R.J., Jamner, L.D., & Whalen, C.K. (2001). Temporal analysis of
the relationship of smoking behavior and urges to mood states in men versus
women. Nicotine and Tobacco Research, 3, 235-48.
Epstein, J.A., Botvin, G.J., Dusenbury, L., & Diaz, T. (1996). Validation of
an acculturation measure for Hispanic adolescents. Psychological Reports, 79, 1075-
1079.
Hansen, W. B., & Graham, J. W. (1991). Preventing alcohol, marijuana,
and cigarette use among adolescents: peer pressure resistance training versus
establishing conservative norms. Preventive Medicine, 20, 414-430.
Goleman, D. (1995). Emotional Intelligence. New York: Bantam Books.
Graham, J.W., Johnson, C.A.. Hansen, W.B., Flay, B.R., & Gee, M. (1990).
Drug use prevention programs, gender, and ethnicity: evaluation of three seventh-
grade Project SMART cohorts. Preventive Medicine, 19, 305-13.
Legree, P. J. (1995). Evidence for an oblique social intelligence factor
established with a likert-based testing procedure. Intelligence, 21, 247-266.
139
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
MacKinnon, D. P., Johnson, C. A., Pentz, M. A., Dwyer, J. H., Hansen, W.
B., Flay, B. R., & Wang, E. Y. (1991). Mediating mechanisms in a school-based
drug prevention program: first-year effects of the Midwestern Prevention Project.
Health Psychology. 10.164-172.
Mayer, J. D., Caruso D. R, & Salovey, P. (1999). Emotional intelligence
meets traditional standards for an intelligence. Intelligence. 27, 267-298.
Mayer, J. D., Caruso D., & Salovey, P. (2000). Models of emotional
intelligence. In R.J. Sternberg (Ed.). Handbook of Intelligence (pp. 396-420).
Cambridge, UK: Cambridge University Press.
Mayer, J.D., & Salovey, P. (1997). What is emotional intelligence? In P.
Salovey & D. Sluyter (Eds). Emotional Development and Emotional Intelligence:
Implications for Educators (pp.3-31). New York: Basic Books.
Mayer, J. D., Salovey, P., & Caruso, D. R. (1997). Multifactor Emotional
Intelligence Scale, Student Version. Durham, New Hampshire: Authors.
Pallonen, U.E., Prochaska, J.O., Velicer, W.F., Prokhorov, A.V., & Smith,
N.F. (1998). Stages of acquisition and cessation for adolescent smoking: an
empirical integration. Addictive Behaviors. 23, 303-24.
Rice, C.L. (1999). A quantitative study of emotional intelligence and its
impact on team performance. Unpublished master’s thesis, Pepperdine Unidersity.
Rubin, M.M. (1999). Emotional intelligence and its role in mitigating
aggression: A correlational study of the relationship between emotional intelligence
and aggression in urban adolescents. Unpublished manuscript, Immaculata College,
hnmaculata, PA.
SAS Institute [Computer software]. (2000). Cary, North Carolina: SAS
Institute.
Sullivan, A. K. (1999). The Emotional Intelligence Scale for Children.
Dissertation Abstracts International. Unpublished doctoral dissertation.
Tani, C.R., Chavez, E.L., Deffenbacher, J.L. (2001). Peer isolation and
drug use among white non-Hispanic and Mexican American adolescents.
Adolescence, 36, 127-39.
140
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Trinidad, D.R. & Johnson, C.A. (2001). Is low emotional intelligence a risk
factor for adolescent tobacco and alcohol use? Poster Presented at the Society for
Behavioral Medicine and Society for Research on Nicotine and Tobacco Joint
Annual Meeting.
Trinidad & Johnson (2002). The Association Between Emotional
Intelligence and Early Adolescent Tobacco and Alcohol Use. Personality and
Individual Differences. 32, 95-105.
Trinidad, D.R., Unger, J.B., Chou, C.P., Azen, S.P., & Johnson, C.A.
(2002a). The protective association of emotional intelligence with psychosocial
smoking risk factors for adolescents. Manuscript submitted for publication.
Trinidad, D.R., Unger, J.B., Chou, C.P., Azen, S.P., & Johnson, C.A.
(2002b). Emotional intelligence and smoking risk factors in adolescents:
Interactions on smoking intentions. Manuscript submitted for publication.
Unger, J.B., Cruz, T.B., Rohrbach, L.A., Ribisl, K.M., Baezconde-
Garbanati, L., Chen, X., Trinidad, D.R., & Johnson, C.A. (2000). English language
use as a risk factor for smoking initiation among Hispanic/Latino and Asian
American adolescents in California: Evidence for mediation by tobacco-related
beliefs and social norms. Health Psychology. 19.403-410.
Unger, J.B., Gallaher, P., Shakib, S., Ritt-Olson, A., Palmer, P.H., &
Johnson, C.A. (in press). The AHIMSA Acculturation Scale: A new measure of
acculturation for adolescents in a multicultural society. Journal of Early
Adolescence.
141
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 6:
The association between emotional intelligence and
adolescent smoking risk factors:
Summary and Conclusion
Emotional intelligence is a modifiable factor that is protective against
traditional smoking risk factors, such as perceived social consequences of smoking,
refusal self-efficacy, smoking intentions, and hostility. The buffering effect of El
may be even greater for certain ethnicities and US acculturation levels that have
been shown to be at an increased risk for smoking. These results of this dissertation
are summarized below and in Figure 8.
Results from analyses of associations between El and psychosocial smoking
risk factors (chapter 2) indicated that El was significantly protective against such
factors. Specifically, high El was related to an increased perception of the negative
social consequences of smoking, higher refusal self-efficacy toward potential
cigarette offers, and lower smoking intentions.
Additional analyses revealed an interaction between El and factors related to
future established smoking (chapter 3). Those with low El were more likely to
intend to smoke if they had low refusal skills or were more hostile, while those with
high El were more likely to intend to smoke if they have previously experimented
with cigarettes.
142
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 8. Summary of Relationship Between El and Smoking Risk Factors
EMOTIONAL
INTELLIGENCE
• Identifying Emotions
• Understand Emotions
• Emotional Management Eth
US-Ac
Smk
Experiment
Hostility
RSE
PSC
Smk
Intention
Chapter 4 examined the interaction of El and ethnicity on smoking
intentions. The association between emotional intelligence and smoking intentions
did not statistically significantly vary across culture/ethnicity, though there was a
trend towards significance. The trend suggested that El might have been more
143
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
protective against smoking intentions in the next year for White adolescents
compared to A/PI and H/L adolescents.
The final set of analyses (chapter 5) examined interactions between El and
US acculturation on the perceived social consequences of smoking. As El increased
so did perceptions of the social consequences of smoking for those who were
acculturated to the US culture.
As mentioned in each of the chapters, future research on El and adolescent
smoking would benefit by sampling older adolescents. Future versions of the MEIS
can be improved with the addition of even more diverse items, such as pictures of
faces, musical selections, and vignettes. Consideration of alternative scoring
methods, such as an expert consensus versus the sample consensus may also diffuse
some of the criticisms regarding validity of the MEIS’s scoring. Nonetheless, the
results of the preceding four chapters indicate there is now mounting evidence for
emotional intelligence as a protective factor against adolescent smoking and
smoking risk factors. Much research on adolescent smoking prevention programs
has identified risk factors, many of which cannot be modified such as ethnicity.
However, inclusion of an El-enhancing component in smoking prevention programs
may be beneficial. Improving El skills such as identifying emotions within oneself
and in others, understanding how emotions come about, and managing emotions can
augment existing prevention efforts. With the recent plateau in the success of
adolescent smoking prevention programs, identification of such a novel, protective
factor as El brings hope for further reducing adolescent tobacco use.
144
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
MASTER REFERENCE LIST
Bar-On, R. (1997). The Emotional Quotient Inventory (EO-it: Technical
Manual. Toronto: Multi-Health Systems.
Berry, J.W. (1980). Acculturation as varieties of adaptation. In A. Padilla
(Ed.), Acculturation: theory, models and new findings (pp. 9-25). Boulder, CO:
Westview.
Bickel, W. K., Odum, A. L., & Madden, G. J. (1999). hnpulsivity and
cigarette smoking: delay discounting in current, never, and ex-smokers.
Psychopharmacology. 146, 447-454.
Botvin, G.J., Griffin, K.W., Diaz, T., Miller, N., & ffill-Williams, M.
(1999). Smoking initiation and escalation in early adolescent girls: One-year
follow-up of a school-based prevention intervention for minority youth. Journal of
the American Medical Womens Association. 54. 139-143, 152.
Buss, A. H., & Durkee, A. (1957). An inventory for assessing different
kinds of hostility. Journal of Consulting Psychology, 21, 343-348.
Carmody, T.P. (1989). Affect regulation, nicotine addiction, and smoking
cessation. Journal of Psychoactive Drugs. 21(3). 331-342.
Carvajal, S.C., Wiatrek, D.E., Evans, R.I., Knee, C.R., & Nash, S.G. (2000).
Psychosocial determinants of the onset and escalation of smoking: Cross-sectional
and prospective findings in multiethnic middle school samples. Journal of
Adolescent Health. 27. 255-65.
Chen, K., & Kandel. D. B. (1995). The natural history of drug use from
adolescence to the mid-thirties in a general population sample. American Journal of
Public Health. 85. 41-47.
Chen, X. & Unger, J.,B. (1999). Hazards of smoking initiation among
Asian American and non-Asian adolescents in California: a survival model analysis.
Preventive Medicine, 28, 589-599.
Chen, X., Unger, J.B., Cruz, T.B., & Johnson, C.A. (1999). Smoking
patterns of Asian-American youth in California and their relationship with
acculturation. Journal of Adolescent Health, 24. 321-328.
Choi, W. S., Pierce, J. P., Gilpin, E. A., Farkas, A. J., & Berry, C. C. (1997).
Which adolescent experimenters progress to established smoking in the United
States? American Journal of Preventive Medicine,13, 385-391.
145
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ciarrochi, J. V., Chan, A. Y. C., & Caputi, P. (2000). A critical evaluation
of the emotional intelligence construct. Personality and Individual Differences, 28.
539-561.
Conrad, K. M., Flay, B. R., & Hill, D. (1992). Why children start smoking
cigarettes: Predictors of onset. British Journal of Addiction. 8 7 ,1711-1724.
Cooper, R.K. (1997). EQ Map. San Francisco: AIT and Essi Systems.
Dalton, M.A., Sargent, J.D., Beach, M.L., Bernhardt, A.M., & Stevens, M.
(1999). Positive and negative outcome expectations of smoking: Implications for
prevention. Preventive Medicine. 29. 460-465.
Davies, M., Stankov, L., & Roberts, R.D. (1998). Emotional intelligence:
In search of an elusive construct. Journal of Personality and Social Psychology. 75,
989-1015.
Defino, R.J., Jamner, L.D., & Whalen, C.K. (2001). Temporal analysis of
the relationship of smoking behavior and urges to mood states in men versus
women. Nicotine and Tobacco Research, 3, 235-48.
Distefan, J.M., Gilpin, E.A., Choi, W.S., & Pierce, J.P. (1998). Parental
influences predict adolescent smoking in the United States, 1989-1993. Journal of
Adolescent Health. 22. 466-474.
DuRant, R.H., Treiber, F., Goodman, E., & Woods, E.R. (1996). Intentions
to use violence among young adolescents. Pediatrics. 98. 1104-1108.
Epstein, J.A., Botvin, G.J., Dusenbury, L., & Diaz, T. (1996). Validation of
an acculturation measure for Hispanic adolescents. Psychological Reports, 79,1075-
1079.
Epstein, J.A., Williams, S.C., Botvin, G.J., Diaz, T., & Ifill-Williams, M.
(1999). Psychosocial predictors of cigarette smoking among adolescents living in
public housing developments. Tobacco Control. 8, 45-52.
Flay, B.R., Hu, F.B., & Richardson, J. (1998). Psychosocial predictors of
different stages of cigarette smoking among high school students. Preventive
Medicine. 27.9-18. 1998.
Gardner, H. (1983). Frames of Mind (10th Anniversary Edition). New
York: Basic Books.
146
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Gardner, H. (1999). Intelligence Reframed: Multiple Intelligences for the
21s t Century. New York: Basic Books.
Goleman, D. (1995). Emotional Intelligence. New York: Bantam Books.
Graham, J.W., Johnson, C.A., Hansen, W.B., Flay, B.R., & Gee, M. (1990).
Drag use prevention programs, gender, and ethnicity: evaluation of three seventh-
grade Project SMART cohorts. Preventive Medicine. 19. 305-13.
Griffin, K. W., Botvin, G. J., Doyle, M. M., Diaz, T., & Epstein, J. A.
(1999). A six-year follow-up study of determinants of heavy cigarette smoking
among high-school seniors. Journal of Behavioral Medicine, 22, 271-284.
Hansen, W. B., & Graham, J. W. (1991). Preventing alcohol, marijuana,
and cigarette use among adolescents: peer pressure resistance training versus
establishing conservative norms. Preventive Medicine. 20. 414-430.
Hedlund, J. & Sternberg, R.J. (2000). Too many intelligences? Integrating
social, emotional, and practical intelligence. In R. Bar-On & J.D. Parker (Eds). The
Handbook of Emotional Intelligence (pp. 136-167). San Francisco: Jossey-Bass
Inc.
Jessor, R. (1984). Adolescent development and behavioral health. In J.D.
Matarazzo, S. M. Weiss, J. A. Herd, N. E. Miller, & S. M. Weiss (Eds.), Behavioral
Health, A Handbook of Health Enhancement and Disease Prevention (pp.69-90).
New York: John Wiley & Sons.
Johnston, L.D., O'Malley, P.M., & Bachman, J.G. (1998). National survey
results on drug use from the Monitoring the Future study. 1975-1997. Volumne T :
Secondary school students. (NIH Publication No. 98-4345). Rockville, MD:
National Institute on Drug Abuse.
Johnston, L.D., O'Malley, P.M., & Bachman, J.G. (Dec. 1999). Cigarette
smoking among American teens continues gradual decline. University of Michigan
News and Information Services: Ann Arbor, MI.[On-line], Available:
www.monitoringthefuture.org; accessed 05/23/00.
Kaplan, C.P., Napoles-Springer, A., Stewart, S.L., & Perez-Stable, E.J.
(2001). Smoking acquisition among adolescents and young Latinas: the role of
socioenvironmental and personal factors. Addictive Behaviors. 26, 531-550.
Kaukiainen, A., Bjorkqvist, K., Lagerspetz, K., Osterman, K., Salmivalli, C.,
Rothberg, S., & Ahlbom, A. (1999). The relationships between social intelligence,
empathy, and three types of aggression. Aggressive Behavior, 25, 81-89.
147
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Lee, L.L., Gilpin ,E.A., & Pierce, J.P. (1993). Changes in the patterns of
initiation of cigarette smoking in the United States: 1950, 1965, and 1980. Cancer
Epidemiology. Biomarkers & Prevention, 2. 593-597.
Leeper, R. (1948). A motivational theory of emotions to replace “emotions
as disorganized response.” Psychological Review. 55. 5-21.
Legree, P. J. (1995). Evidence for an oblique social intelligence factor
established with a likert-based testing procedure. Intelligence. 21. 247-266.
MacKinnon, D. P., Johnson, C. A., Pentz, M. A., Dwyer, J. EL, Hansen, W.
B., Flay, B. R., & Wang, E. Y. (1991). Mediating mechanisms in a school-based
drug prevention program: first-year effects of the Midwestern Prevention Project.
Health Psychology. 1 0 .164-172.
Mayer, J.D., Carlsmith, K.M., & Chabot, H.F. (1998). Describing the
person’s external environment: Conceptualizing and measuring the Life Space.
Journal of Research in Personality. 32. 253-296.
Mayer, J. D., Caruso, D. R., & Salovey, P. (1997). Multifactor Emotional
Intelligence Scale, Student Version. Durham, New Hampshire: Authors.
Mayer, J.D., Caruso, D.R., & Salovey, P. (1999). Emotional intelligence
meets traditional standards for an intelligence. Intelligence. 27. 267-298.
Mayer, J.D., & Salovey, P. (1993). The intelligence of emotional
intelligence. Intelligence. 17. 433-442.
Mayer, J.D.,& Salovey, P. (1997). What is emotional intelligence? In P.
Salovey & D. Sluyter (Eds). Emotional Development and Emotional Intelligence:
Implications for Educators (pp.3-31). New York: Basic Books.
Mayer, J. D., Salovey, P., Caruso D. (2000). Models of emotional
intelligence. In R.J. Sternberg (Ed.). Handbook of Intelligence. Cambridge, UK:
Cambridge University Press.
Noland, M.P., Kryscio, R.J., Riggs, R.S., Linville, L.H., Ford, V.Y., &
Tucker, T.C. (1998). The effectiveness of a tobacco prevention program with
adolescents living in a tobacco-producing region. American Journal of Public
Health. 88. 1862-1865.
O’Callaghan, F.V., Callan, V.J., & Baglioni, A. (1999). Cigarette use by
adolescents: attitude-behavior relationships. Substance Use & Misuse. 34. 455-468.
148
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Pallonen, U.E., Prochaska, J.O., Velicer, W.F., Prokhorov, A.V., & Smith,
N.F. (1998). Stages of acquisition and cessation for adolescent smoking: an
empirical integration. Addictive Behaviors. 23. 303-24.
Patton, G. C., Carlin, J. B., Coffey, C., Wolfe, R., Hibbert, M., & Bowes, G.
(1998). The course of early smoking: a population-based cohort study over three
years. Addiction, 93. 1251-1260.
Patton, D., Barnes, D., & Murray, R.P. (1997). A personality typology of
smokers. Addictive Behaviors. 22. 269-273.
Pierce, J.P., Choi, W.S., Gilpin, E.A., Farkas, A.J., & Merritt, R.K. (1996).
Validation of susceptibility as a predictor of which adolescents take up smoking in
the United States. Health Psychology, 15, 355-61.
Rice, C.L. (1999). A quantitative study of emotional intelligence and its
impact on team performance. Unpublished master’s thesis, Pepperdine Unidersity.
Rubin, M.M. (1999). Emotional intelligence and its role in mitigating
aggression: A correlational study of the relationship between emotional intelligence
and aggression in urban adolescents. Unpublished manuscript, Immaculata College,
Immaculata, PA.
SAS Institute [Computer software]. (200). Cary, North Carolina: SAS
Institute.
Salovey, P. & Mayer, J.D. (1990). Emotional intelligence. Imagination,
Cognition, and Personality, 9 ,185-211.
Scherwitz, L.W., Perkins, L.L., Chesney, M.A., Hughes, G.H., Sidney, S. &
Manolio, T.A. (1992). Hostility and health behaviors in young adults: The
CARDIA Study. Coronary Artery Risk Development in Young Adults Study.
American Journal of Epidemiology, 136,135-45.
Skara, S., Sussman, S., & Dent, C.W. (2001). Predicting regular cigarette
use among continuation high school students. American Journal of Health
Behavior. 25. 147-156.
Sternberg, R. J. (1997). The concept of intelligence and its role in lifelong
learning and success. American Psychologist. 52. 1030-1045.
Sullivan, A. K. (1999). The Emotional Intelligence Scale for Children.
Dissertation Abstracts International. Unpublished doctoral dissertation.
149
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Sussman, S., Lichtman, K., Ritt, A., & Pallonen, UE. (1999). Effects of
thirty-four adolescent tobacco use cessation and prevention trials on regular users of
tobacco products. Substance Use & Misuse. 3 4 ,1469-1503.
Tani, C.R., Chavez, E.L., & Deffenbacher, J.L. (2001). Peer isolation and
drug use among white non-Hispanic and Mexican American adolescents.
Adolescence. 3 6 ,127-39.
Thorndike, E. (1920). Intelligence and its uses. Harper’s Magazine, 140,
227-235.
TIME (October 2,1995). [Cover]. New York: Time Warner.
Triandis, H.C. (1995). Individualism and collectivism. Boulder, CO:
Westview Press.
Trinidad, D.R. & Johnson, C.A. (2001). Is low emotional intelligence a risk
factor for adolescent tobacco and alcohol use? Poster Presented at the Society for
Behavioral Medicine and Society for Research on Nicotine and Tobacco Joint
Annual Meeting.
Trinidad, D.R. & Johnson, C.A. (2002). The association between
emotional intelligence and early adolescent tobacco and alcohol use. Personality
and Individual Differences, 32, pp.95-105.
Trinidad, D.R., Unger, J.B., Chou, C.P., Azen, S.P., & Johnson, C.A.
(2002a). The protective association of emotional intelligence with psychosocial
smoking risk factors for adolescents. Manuscript submitted for publication.
Trinidad, D.R., Unger, J.B., Chou, C.P., Azen, S.P., & Johnson, C.A.
(2002b). Emotional intelligence and smoking risk factors in adolescents:
Interactions on smoking intentions. Manuscript submitted for publication.
Unger, J.B., Cruz, T.B., Rohrbach, L.A., Ribisl, K.M., Baezconde-
Garbanati, L., Chen, X., Trinidad, D.R., & Johnson, C.A. (2000). English language
use as a risk factor for smoking initiation among Hispanic/Latino and Asian
American adolescents in California: Evidence for mediation by tobacco-related
beliefs and social norms. Health Psychology. 19, 403-410.
Unger, J.B., Gallaher, P., Shakib, S., Ritt-Olson, A., Palmer, P.H., &
Johnson, C.A. (in press). The AHIMSA Acculturation Scale: A new measure of
acculturation for adolescents in a multicultural society. Journal of Early
Adolescence.
150
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Unger, J.B., Johnson, C.A., Stoddard, J.L., Nezami, E., & Chou, C.P.
(1997). Identification of adolescents at risk for smoking initiation: validation of a
measure of susceptibility. Addictive Behaviors. 22, 81-91.
Welte, J. W., Barnes, G. M., Hoffinan, J. H., & Dintcheff, B. A. (1999).
Trends in adolescent alcohol and other substance use: relationships to trends in peer,
parent, and social influences. Substance Use & Misuse. 3 4 ,1427-1449.
Wills, T. A., Vaccaro, D., & McNamara, G. (1994). Novelty seeking, risk
taking, and related constructs as predictors of adolescent substance use: an
application of Cloninger’s theory. Journal of Substance Abuse, 6, 1-20.
Young, T.L., & Rogers, K.D. (1986). School performance characteristics
preceding onset of smoking in high school students. American Journal of Diseases
in Children. 140. 257-259.
151
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Correlates of smoking behaviour among Egyptian adolescents

PDF

Aggression and victimization as predictors of drug use in early adolescence

PDF

Adolescent stress and styles of coping: Predictors and moderators of psychological distress

PDF

Biopsychosocial factors in major depressive disorder

PDF

Correlates of compliance with mammography screening guidelines among low-income Latinas: An exploratory study

PDF

Patterns and risk factors for adolescent smoking progression

PDF

Cigarette smoking, metabolic syndrome, and subclinical atherosclerosis

PDF

Artery wall injury and LDL-cholesterol in early atherosclerosis: The Los Angeles atherosclerosis study

PDF

Adaptation to a life -threatening diagnosis: Dispositional optimism and pessimism and posttraumatic growth among patients with HIV

PDF

Psychosocial and behavioral ractors associated with emotional eating in adolescents

PDF

Objectively measured physical activity and related factors in minority youth

PDF

Cognitive attributions for smoking and their roles on subsequent smoking progression and regression

PDF

Interaction of dietary fiber and serum cholesterol on early atherosclerosis

PDF

Frequency and risk factors of poststroke dementia

PDF

Adolescent exposure to alcohol advertising: A prospective extension of Strickland's model

PDF

The vicious cycle of inactivity, obesity, and metabolic health consequences in at-risk pediatric populations

PDF

Implicit cognition and dissociative experiences as predictors of adolescent substance use

PDF

Contributions of interpersonal communication within a mass media environment: Exploring the role of social influence in a statewide tobacco control media campaign

PDF

Adoption and implementation of evidence-based tobacco use prevention curricula and programming

PDF

Intrapersonal and environmental factors associated with Chinese youth alcohol use experimentation and binge drinking behaviors

Asset Metadata

Creator Trinidad, Dennis Ryan (author)

Core Title Emotional intelligence and smoking risk factors in early adolescents

School Graduate School

Degree Doctor of Philosophy

Degree Program Preventive Medicine/Health Behavior Research

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag health sciences, public health,OAI-PMH Harvest,psychology, social

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Johnson, C. Anderson (committee chair), Azen, Stanley (committee member), Chou, Chih-Ping (committee member), Gatz, Margaret (committee member), Unger, Jennifer (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-545265

Unique identifier UC11339934

Identifier 3094375.pdf (filename),usctheses-c16-545265 (legacy record id)

Legacy Identifier 3094375.pdf

Dmrecord 545265

Document Type Dissertation

Rights Trinidad, Dennis Ryan

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA