Resiliency and
Risk among Young People of Color
University
of Minnesota Children, Youth and Families Consortium. Permission is
granted to create and distribute copies of this document for non-commercial
purposes provided that the Urban Coalition and University of Minnesota's
adolescent Health Program and MN CYFC receive acknowledgment and this
notice is included.
Several
tables from this report have not been reproduced with this electronic
copy.
1994,
produced by The Urban Coalition, 2610 University Avenue West, Suite
201, St. Paul, MN 55114
(612) 348-8550
ACKNOWLEDGMENTS
This
is the second report produced by the Urban Coalition in collaboration
with the University of Minnesota's Adolescent Health Program. We would
like to thank Dr. Michael Resnick and Dr. Robert Blum, both of the Adolescent
Health Program, and Ms. Linda Harris of the University-affiliated National
Adolescent Health Resource Center, funded by the Maternal and Child
Health Bureau, for their involvement in this collaboration from the
beginning. After working with us on The Next Generation, which the Coalition
published in 1990, they saw the potential for further research on resiliency
and protective factors among youth, conceived the strategy for doing
the analysis, and provided valuable advice and support along the way.
Ms.
Brenda Miller handled the very demanding computer and statistical analysis
for the follow-up study. We deeply appreciate her skill in overcoming
some difficult technical problems and producing solid, well-grounded
results.
Several
foundations contributed financially to this effort. We would like to
thank the McKnight Foundation, the Honeywell Foundation, the Dayton-Hudson
Foundation, the General Mills Foundation and the Pillsbury Foundation
(now the Grand Metropolitan- Food Sector Foundation) for their support.
Finally,
we cannot end this section without recognizing the contributions of
Diane Hedin, a creative researcher and advocate for young people who
passed away while this project was underway. The ideas and funding possibilities
for this project began to crystallize during meetings with Diane while
she was with the Pillsbury Foundation and on leave from her position
at the University's Center for Youth Development and Research, and she
helped build support for the effort. We hope this report reflects the
spirit she brought to her work with young people.
Mission
The
Urban Colition's mission is to increase the capacity of low–income,
African American, American Indian, Asian and Chicamo/Latino persons
and Communities to adress political, social and economic concerns which
they identify.
Founded
in 1968, the Urban Coalition is a non-profit organization that pursues
its mission through research, public policy, technical assistance, advocacy
and capacity-building. In recent years, the Coalition has focused on
education, employment, food and hunger, health, and race relations issues,
although it may become involved in other issues as they arise. In the
spirit of true coalition, it is involved in multiple partnerships and
collaborations with other organizations in the community.
BOARD
OF DIRECTORS
Tracey
Burton, Richard Cathcart, George Chaney, Mayor Norm Coleman, Mark Fiddler,
Luz Maria Fria, Portia Hampton-Flowers, Van Ahn Hoan,g Michael C. Hyter,
Gary R. Johnson, Gaohli Y. Lee, Jeff Levy, Juan Lopez, Patrick McNulty,
Yusef Mgeni, ex-officio Sothea Poch, Peter Rode, ex-officio Mayor Sharon
Sayles-Belton, Arlys J. Stadum, chairperson Joanne Stately, Tom Tefft,
treasurer Genevieve Williams-Murph, corporate secretary, ex-officio
Keyah
Davis program officer, health
JoAnn Cardenas Enos, education advocate
Rachel Fang, program officer, food and hunger
Allan Malkis, research associate, census project
Julia Malmgren, director of administration
Yusef Mgeni, president
Peter Rode, vice-president, research
Elaine Salinas, program officer, education
Robert Taylor, Project ASSIST
Genevieve Williams-Murph, office manager
RoseAnn Zimbro, education advocate
EXECUTIVE
SUMMARY
In
1990, the Urban Coalition published The Next Generation, a portrait
of the health and well-being of African American, American Indian, Asian,
Chicano/Latino and white adolescents from Minneapolis and St. Paul.
The report used data obtained from the University of Minnesota's Adolescent
Health Survey, in which 13,000 7th-12th graders from the two cities
responded to questions about physical health, emotional stress, worries
and concerns, chemical use, sexual behavior, family and community relationships
and other subjects.
This
report uses the same rich database to explore "resiliency,"
the capacity of young people to thrive, to be emotionally healthy, and
to avoid destructive behavior in spite of very difficult circumstances.
Using statistical analysis, this study seeks to identify the factors
and life experiences that uniquely contribute to emotional well-being
and avoidance of damaging behavior.
Three
outcome variables representing feelings as well as behavior were selected
for the analysis: emotional stress, suicide risk, and delinquency. Since
one purpose of the study was to see if different dynamics were at work
in each community, the analysis was done separately for each racial/ethnic
group.
–EMOTIONAL
STRESS
The study found that three factors made significant contributions to
increasing emotional stress in all five racial/ethnic groups. Adolescents
who said that they were bored ("nothing interesting to do"),
who had negative feelings about their bodies, and who were worried about
violence (whether at home, in the neighborhood, at school or in the
country at large) were much more likely to have high emotional stress.
Strong
family connectedness, on the other hand, reduced emotional stress in
all five racial/ethnic groups. Young people who said they had lots of
fun with their families and who felt that their families cared about
them and understood them reported much less emotional stress.
Other
factors made significant contributions in some communities but not others.
Previous experiences of physical abuse was related to higher emotional
stress in most but not all racial/ethnic groups. Likewise, strong connections
with adults and friends outside the family helped reduce emotional stress
in some communities.
–SUICIDE
RISK
The study found that in all five racial/ethnic groups, high emotional
stress was the most powerful factor leading to greater risk of suicide.
Perhaps not surprisingly, students experiencing high emotional stress
were much more likely to report suicidal thoughts or previous suicide
attempts. Another factor linked to suicide risk in all racial/ethnic
groups was recent serious mental health or emotional problems within
the family. Suicide risk was significantly greater among youth who said
that someone in their family had recently experienced serious mental
health or emotional problems or had received treatment for them.
In
some racial/ethnic groups, but not all, suicide risk was also greater
among young people who had experienced physical abuse, who had certain
chronic physical health problems, and who expressed an unwillingness
to seek help from anyone about emotional or relationship concerns. An
unexpected finding was that American Indian students who were doing
well in school reported suicidal thoughts or attempts more often than
those not doing well in school.
On
the other hand, in the American Indian and Chicano/Latino communities,
religiosity or spirituality appeared to reduce suicide risk significantly.
Those who considered themselves to be a religious or spiritual person
were much less likely to report suicidal thoughts or attempts.
–DELINQUENCY
The results for the delinquency outcome variable were less helpful than
the results for emotional stress and suicide risk. Two of the most powerful
contributors to lower levels of delinquent activity were demographic
variables -gender and age. Adolescent girls were much less involved
in delinquent behavior than boys, and older students were less involved
than younger students. The study did not indicate, however, what it
is about being female or being older that leads to less delinquency.
In
some racial/ethnic groups, chemical use by parents (tobacco, alcohol,
marijuana and other drugs) and the presence of recent mental health
or emotional problems in the family appeared to contribute to higher
delinquency. Youths with certain chronic physical health problems were
also more likely to be delinquent in some communities.
Factors
that contributed to less involvement in delinquent behavior in some
communities included willingness to turn to parents for help with emotional
or relationship problems and strong family connectedness.
CONCLUSION
Although the results of the analysis were not always the same for each
racial/ethnic group, it is important to note that many of the factors
identified as having an impact were linked to personal relationships
and to the adolescent's sense of belonging. Strong family connections
was a universal factor helping to reduce emotional stress. In some communities,
caring relationships with people outside the family also helped reduce
emotional stress, and being able to turn to parents for help was associated
with lower delinquency. Conversely, mental and emotional problems among
other family members, physical abuse, and having no one to turn to for
help seemed to contribute to greater stress or destructive behavior.
Addressing this fundamental adolescent, and human, need for caring relationships
and belonging should be an integral feature of public policies, programs
and interventions on behalf of youth.
INTRODUCTION
Some
adolescents feel high levels of emotional distress and/or engage in
behavior that may be damaging to themselves or others, such as drug
and alcohol use, drinking and driving, attempting suicide, unprotected
sexual activity, delinquency, and so forth. Yet, other young people
feel much less stress and rarely engage in risky or damaging behavior.
What accounts for these different kinds of experiences during adolescence?
One of the challenges we face in developing policies and programs for
young people is to understand what factors contribute to their physical,
mental and emotional health. In particular, what contributes to "resilience,"
the ability of some young people to thrive and avoid high-risk behavior
despite facing very difficult circumstances?
These
questions are of great importance to American Indian, African American,
Asian and Chicano/Latino young people, who often must grow up not only
in very low income families but also with messages from the larger culture
that tear away at their confidence and self-worth. By most measures,
the pressures are becoming greater. The number of children of color
growing up in very low-income families in Minneapolis and St. Paul soared
between 1979 and 1989. (See Table 1) Deteriorating prospects for higher
education and good jobs weigh heavily on many young people. Too many
children of color are not doing well in school. Neighborhood safety
is a growing concern. What can be done, then, to help young people deal
with and overcome such adversity?
'Profiles
of Change: Communities of Color in the Twin Cities Area,' published
by the Urban Coalition in 1993, contains social and economic census
data on communities of color in the seven county metropolitan area.
Data is presented separately for Minneapolis, St. Paul, the suburbs,
and the region as a whole. Copies are available from the Coalition by
writing to 2610 University Avenue West, Suite 201, St. Paul, MN 55114
or by calling 612348-8550.
Among
the many reports on the economic realities facing young adults are:
Vanishing Dreams: The Growing Economic Plight of America's Young Families,
by the Children's Defense Fund (1988); and The Forgotten Half: Pathways
to Success for America's Youth and Young Families, by the William T.
Grant Foundation (1988).
The
best recent report on students of color in the Minneapolis and St. Paul
school systems is Children of Color: A Wake-Up Call to the Community
(1993). Copies are available through the Communities of Color Institute,
1501 Hennepin Avenue South, Minneapolis, MN 55403. (Phone: 612-373-2797)
TABLE
1: POVERTY RATE FOR CHILDREN- 1979-1989
| MINNEAPOLIS |
1979 |
1989 |
| African
American |
38.5% |
53.9 |
| American
Indian |
48.9% |
66.3 |
| Asian |
49.0% |
55.8 |
| Chicano/Latino |
31.3% |
40.6 |
| White |
10.3% |
12.9 |
| Total
Population |
18.0 |
30.6% |
Source:
Profiles of Change: Communities of Color in the Twin Cities Area, Urban
Coalition, 1993, p. 17.
PURPOSE
The
purpose of this study is to take advantage of survey data on urban youth
to answer two questions:
- What
are the factors and life experiences that uniquely contribute to emotional
well-being and help young people avoid involvement in highrisk or
damaging behavior?
- Are
different dynamics at work in each community of color? Do we find
that the same factors and life experiences affect highrisk behavior
in each community, or are there significant differences from one community
to another?
This
study builds upon the findings of an earlier Urban Coalition report,
The Next Generation, which provided a portrait of the health and well-being
of young people of color in Minneapolis and St. Paul.
4 Both
studies are based on data obtained from the University of Minnesota's
Adolescent Health Survey, which involved about 13,000 American Indian,
African American, Asian, Chicano/Latino and white 7th-12th grade students
in the two cities.
The
earlier report was descriptive. It showed how prevalent certain worries,
attitudes and behaviors are among adolescents, it described patterns,
and it pointed out where young people of different racial/ethnic groups
differed from one another and where they converged. The current study
is explanatory. It uses the same large, rich database to take the analysis
one step further by finding the factors that seem to have the greatest
independent effect upon emotional stress and destructive behavior.
METHOD
OF THE STUDY
The
method used to find the most important explanatory factors relies heavily
on statistical analysis. The Adolescent Health Survey contained no open-ended
questions and no individual "stories," and therefore the data
are lacking in that personal dimension through which students could
describe and explain their thoughts and feelings in their own words.
But it did contain hundreds of carefully-worded multiple-choice questions
covering a wide range of physical, mental and emotional health concerns,
including body image, stress, suicide risk, worries and concerns, chemical
use, risky behaviors, school experiences and family relationships. The
size and the breadth of coverage of the Adolescent Health database make
it possible to use special statistical methods to try to discover the
factors that make unique contributions to health, well-being and resilience.
Three
outcome measures were chosen as the focal points for the study, primarily
because they reflect different kinds of outcomes involving youth and
because the measures are sound enough to allow advanced statistical
analysis.
The
three outcome variables are:
- Emotional
Stress, which is concerned with feelings and is measured by 17 questions
having to do with depression, anxiety, nervousness, tiredness and
boredom.
- Suicide
Risk, which combines information from questions on both suicidal thoughts
and behavior (in the form of previous suicide attempts).
- Delinquency,
which is concerned with acting-out behavior and is measured by eight
questions on involvement in violence, vandalism, theft, running away
from home, and other delinquency offenses.
A common
plan has been used to guide the analysis of each outcome variable. The
challenge is to sort through the large number of questions and topics
included in the Adolescent Health Survey and find those that seem to
have the strongest independent association with emotional stress, suicide
risk, and delinquency.
This
is done in two stages:
- First,
simple correlations were calculated to determine which factors were
significantly related to the outcome variable.
- Second,
regression analysis was used to eliminate factors which did not add
to our ability to predict an individual's level of emotional stress,
suicide risk and delinquency.
Prediction
is the name of the game here. The goal of this kind of analysis is to
find the smallest number of factors that enables us to make strong predictions
of the outcome variables. The following example describes how the analysis
works.
The
actual method is a bit more complicated and technical than described
here. Those who need more detailed information can contact Pete Rode
at the Urban Coalition (612-348-8550) or Dr. Michael Resnick at the
University of Minnesota (612-626-2726).
Simple
correlations (Stage I of the analysis) show that general physical health
is related to emotional stress. Adolescents who have chronic physical
health problems tend to have high levels of emotional stress. However,
regression analysis (Stage II) shows that, if we already have information
on other variables measured by the survey, the existence of physical
health problems does not help us improve significantly our ability to
predict the level of emotional stress. Since it doesn't help, we eliminate
it from our final model.
On
the other hand, knowing that someone has been physically abused does
increase our ability to predict an individual's level of emotional stress,
even when we already have information on other factors. As a result,
physical abuse remains in the model as an independent predictor of the
outcome variable, emotional stress.
This
process of testing and winnowing been done separately for each of the
outcome variables and for each racial/ethnic group. As we shall see,
the final group of factors that help us predict emotional stress is
not exactly the same as the group of factors that best predict suicide
risk or delinquency. Furthermore, the factors that predict an outcome
variable like suicide risk among American Indians are not exactly the
same as those that predict suicide risk among Asian Americans.
NOTE
ON THE SAMPLE
The Adolescent Health Survey was administered during February 1987 in
Minneapolis Public Schools and May 1987 in St. Paul Public Schools.
Every public secondary school in the Minneapolis district participated
in the survey. In addition, four junior and senior high schools from
St. Paul were allowed to participate by St. Paul district officials.
All students in school on the day of the survey were asked to complete
the survey in a classroom setting. The survey was administered by the
University rather than by classroom teachers. Students could choose
not to participate and parents had the right to refuse to allow their
child to take the survey.
The
original Adolescent Health Survey data base included surveys from 13,879
students in the Minneapolis and St. Paul school districts, or about
69 percent of those enrolled at the beginning of the school year in
participating schools. For the purposes of this study, 964 were removed
from the database because information on racial/ethnic group was not
provided or because there were too many missing responses. The racial/ethnic
breakdown of the final sample was as follows:
| African
American |
2584 |
| American
Indian |
372 |
| Asian |
856 |
| Chicano/Latino |
270 |
| White |
8833 |
| Total |
12,915 |
Based
on enrollment data from the 1986-87 school year, it appears that the
participation rate was highest for white and Chicano/Latino students
and much lower for African American, Asian and American Indian students.
The
most important limitations of the sample:
- Young
people who have dropped out of school are completely unrepresented,
and students who are often truant or sick or skip class are under-represented.
This is a survey of adolescents attending school. Results might have
been different if dropouts and chronic truants had somehow been included.
- Southeast
Asian refugee groups appear to be under-represented within the Asian
student sample. In Minneapolis, Southeast Asian students were enrolled
primarily at schools that offered Limited English Proficiency (LEP)
programs. Survey participation among Asian students was lowest in
the schools with LEP programs. Southeast Asians are also underrepresented
because St. Paul, which had only four schools involved in the survey,
has a larger refugee population than Minneapolis.
- Chicano/Latino
students from St. Paul are under-represented because only four St.
Paul schools participated in the survey and because the schools selected
did not have many Chicano/Latino students. Even though the St. Paul
Public Schools have more Chicano/Latino students than Minneapolis,
only one-third of the Chicano/Latino students in the Twin Cities Adolescent
Health Survey database were from St. Paul.
EMOTIONAL
STRESS
BACKGROUND
In this study, emotional stress can be thought of as a variable which
sums up the overall emotional state of an individual. A single measure
of emotional stress was constructed by combining responses to 17 questions
dealing with nervousness, anxiety, sadness, happiness, and tiredness.
When
students were divided into high, medium and low groups based on their
emotional stress scores, striking differences were found between girls
and boys. In each racial/ ethnic group, girls were much more likely
than boys to report high levels of emotional stress.
The
overall differences between racial/ ethnic groups were not very large.
However, there were important differences on some of the individual
questions that were used to calculate the overall measure of emotional
stress. Compared with white youths, adolescents of color were more likely
to say that their daily lives were uninteresting, that they did not
feel "in control" of their behavior, thoughts or emotions,
and that they felt sad or discouraged.
RESULTS
OF THE ANALYSIS
When the regression analysis was conducted as described in the previous
chapter, eleven variables emerged as having a significant and unique
impact on emotional stress in one or more of the five racial/ethnic
groups. Table 2 summarizes the results of the analysis. The numbers
in the table are known as "Beta" scores. Beta scores can be
thought of as measuring how much unique impact a variable has on an
outcome variable, in this case emotional stress. Variables with higher
Beta scores (whether positive or negative) have
more impact than those with lower scores. In this table, Beta scores
are only included if the contribution they make is statistically significant
and unique.
One
further note on the analysis: It is not valid to compare Beta scores
for one racial/ethnic group with another group. For example, the Beta
score for "Worries about Violence" among Asian students can
only be compared directly with the scores for other variables affecting
Asian students-it cannot be compared directly with the Beta score for
"Worries about Violence" among American Indian youth or any
other group.
PERCENT
INDICATING HIGH EMOTIONAL STRESS
| African
American |
12.8
male |
24.1
female |
| Asian |
15.2
male |
31.2
female |
| Chicano/Latino |
20.0
male |
33.1
female |
| White |
15.8
male |
28.9
female |
| American
Indian |
16.5
male |
25.3
female |
FACTORS
ASSOCIATED WITH HIGHER EMOTIONAL STRESS
As Table 2 shows, there were three variables which appeared as significant
predictors of high emotional stress in all five ethnic groups. These
were boredom, negative body image, and worries about violence:
- Boredom
was measured by one question which asked: "How often do you feel
like 'there's nothing to do' or that you are bored?" The positive
Beta scores mean that students who answered "much of the time"
or "all of the time" were much more likely to have high
emotional stress. Boredom was the most powerful factor related to
emotional stress among American Indian and Chicano/Latino students.
- Negative
body image was based on two questions which indicate how young
people feel about their bodies and their weight. In all racial/ethnic
groups, negative images of one's body were associated with greater
emotional stress.
- Worries
about violence was based on five questions dealing with worries
about violence in one's home, in the neighborhood and in the world.
Greater worries about violence are associated with high levels of
emotional stress. This variable does not directly measure experiences
of violence, but indicates how much young people feel worried or concerned
about various kinds of violence.
Several
other variables were predictors of high emotional stress in some but
not all racial/ethnic groups. These included:
- Physical
abuse: In all communities except the Chicano/Latino community,
previous experiences with physical abuse had an inde pendent relationship
with emotional stress. Youths who said they had been -physically abused
by someone in their family or by anyone else were more likely to have
higher emotional stress.
- Year
or grade in school: In the African American, Asian and White communities,
adolescents in the higher school grades- such as the 10th, 11th and
12th grades-had higher emotional stress.
- Recent
mental/emotional problems in family: In the African American and
White communities, youths who said that someone in their family had
recently experienced mental or emotional problems or had received
treatment for them were more likely to have high emotional stress.
The Beta scores are not very large (.06), but they are large enough
to indicate some independent contribution to our ability to predict
emotional stress.
- Being
female: Among Asian youths, simply knowing that someone was female
increased our ability to predict emotional stress. Girls experienced
much higher emotional stress, even after taking into account such
powerful factors as body image and family connectedness.
- Drug
climate at school: In the white community, students who perceived
lots of drug use and drug dealing at their school had higher emotional
stress.
FACTORS
REDUCING EMOTIONAL STRESS
- Strong
family connectedness, which is based on six questions dealing
with the degree of caring and understanding that adolescents felt
was present in their families, appeared to contribute to lower emotional
stress in all five racial/ethnic groups. (The negative Beta scores
mean that the higher levels of family connectedness were associated
with lower levels of emotional stress.) In the African American (-.29),
Asian (-.32) and White (-.27) communities, family connectedness had
the greatest impact on emotional stress-either positive or negative-
of any variable listed in Table 2. In the American Indian and Chicano/Latino
communities, only boredom had greater impact.
- Strong
connections with adults and friends outside family: In the Chicano/
Latino and White communities, youths who felt that school people,
church leaders, other adults and/or their friends cared about them
experienced lower levels of emotional stress.
CONCLUSIONS
AND DISCUSSION
1. Good relationships within the family clearly contribute to emotional
well-being. Young people who say they have lots of fun with their family
and who feel that their family cares about their feelings, understands
them, and pays a lot of attention to them have much less emotional stress.
In
most racial/ethnic groups, family connectedness had more impact on emotional
health than any other factor included in the study. This should serve
as a signal to policy makers, funders, community leaders and others
that we must confront all the social and economic forces that place
great stress on parents and families and threaten to tear apart the
nurturing potential of families. Greater economic opportunity, higher
wages, good health care (including mental health and chemical dependency
treatment), parent education programs, workplace reforms, and other
program and public policy changes that strengthen families should be
supported.
It
should also be noted that the presence of caring relationships with
adults or friends outside the family was an important factor in some
communities, though not as powerful as family connectedness. Taken together,
these factors suggest how important it is for adolescents to be surrounded
by caring, positive, and nurturing relationships.
2.
Another noteworthy finding of the study was the emergence of boredom-the
feeling that "there is nothing to do"-as an important predictor
of high emotional stress. The challenge for families and communities
is to keep adolescents engaged and interested in recreational activities,
cultural events, school, and community service. That can be difficult
when, as is often the case, the opportunities offered by mainstream
culture do not engage the personal and cultural needs of adolescents
of color. Another challenge is to encourage the development of "quieter"
activities, such as a love of reading, so that adolescents will not
feel lost when stimulating group activities are unavailable.
To
deal effectively with the potential for boredom and disinterest, families
and communities need to focus on at least three things:
- plenty
of recreational, cultural, educational, spiritual and service opportunities;
- a
strong voice for young people in creating and designing those opportunities;
and
- good
communications with adolescents, about feelings as well as interests.
3.
Negative body image is an important contributor to emotional stress
in all communities of color. Much attention has been focused on body
image as an issue facing girls, and far more girls reported negative
body images than did boys. However, negative body image was an important,
unique predictor of emotional stress for boys as well as girls. Cultural
images generated by advertising, media and other sources define what
is "positive", desirable and attractive in ways that do great
damage to the selfesteem of many adolescents. But it's not just the
media. The entire culture needs to be challenged on its simplistic and
unrealistic images that tyrannize the selfperceptions and selfesteem
of many youths. At the same time, that concern needs to be balanced
by an approach that conveys accurate health information about weight
and development, and that points out what teenagers can do to maintain
healthy bodies.
4.
Worries about violence was the only set of worries investigated in this
study that had an independent, unique relationship with emotional stress.
Adolescents were also asked how often they worried about: the future
(jobs, doing well in school, etc.); death (their parents and their own);
economic conditions; relationships with their peers; and broad social
issues (war, poverty, etc.). These worries, while often more prevalent
than concerns about violence, did not prove to be strongly related to
emotional stress after all other factors were taken into account. Worries
about violence, on the other hand, appears to have a direct, immediate
connection with stress.
This
study did not look at only one kind of violence, but tried to elicit
concerns about violence in several settings. Young people were asked
how much they worried about:
- all
the violence in my neighborhood;
- all
the violence that happens in our country;
- all
the violence in my home;
- that
one of my parents will hit me so hard I'll be hurt;
- getting
beat up in school.
The
results underscore the hidden, silent impact violence has had on the
sense of well-being and optimism of young people of all racial/ethnic
groups. The fact that physical abuse is an additional predictor of emotional
stress in several groups underscores the importance of stopping violence
in our communities.
SUICIDE
RISK
BACKGROUND
From 1982 to 1992, an average of 51 Minnesota young people between 10
and 19 years of age have taken their own lives each year. As tragic
as these self-inflicted deaths may be, they represent the tip of an
iceberg, a tiny fraction of young people who have made some attempt
at suicide or have thought seriously about it.
In
the Adolescent Health Survey, one of every eight Minneapolis and St.
Paul adolescents said they had tried to kill themselves at some point
in their lives, and one of twenty said they had made an attempt within
the past year.
More
than one of every ten youths were considered to be at "high risk"
for suicide. The variable called Suicide Risk was created by combining
answers to two questions, one about suicide attempts and one about thoughts.
Suicide Risk was defined as high if the youth agreed that "I would
commit suicide if I had the chance," or if they had tried to commit
suicide in the past year and still had serious thoughts about killing
themselves.
The
survey did not ask for details about how young people had tried to kill
themselves. We do not, therefore, know how lifethreatening their behavior
actually was. What is important, however, is the fact that young people
perceived that they wanted to kill themselves and had taken serious
steps in that direction. Suicidal thoughts and attempts are important
in themselves and indicate significant depression and turmoil.
As
the accompanying chart shows, suicide risk was considerably higher in
communities of color compared to the white community. While nine percent
of white youths were considered to be at high risk of suicide, between
13 and 16 percent of young people of each community of color were in
the high risk group.
Eleven
different variables emerged as; having a significant and independent
effect on suicide risk for one or more of the five racial/ethnic groups.
(See Table 3.)
PERCENT
INDICATING HIGH SUICIDE RISK
(AS INDICATED BY RECENT ATTEMPTS AND SUICIDAL THOUGHTS)
| African
American |
13.5 |
| American
Indian |
14.2 |
| Asian |
14.4 |
| Chicano/Latino |
16.2 |
| White |
9.3 |
Source:
The Next Generation Urban Coalition, 1990, p. 19.
FACTORS
ASSOCIATED WITH GREATER RISK OF SUICIDE
- High
emotional stress: By looking at Table 3, one can see that high
emotional stress had the greatest impact on suicide risk in all five
racial/ethnic groups. (In other words, emotional stress had the highest
Beta scores in each group.) This is not surprising, if we think of
suicidal attempts and thoughts as indicating very serious depressive
or emotional difficulties. But it does indicate that the kind of questions
used to measure emotional stress could be helpful in alerting us to
more dangerous levels of depression and distress.
- Recent
family mental/emotional problems: Only one other variable proved
to be an independent predictor of suicide risk in all five groups-the
presence of recent serious mental or emotional health problems within
the family. The survey asked adolescents if anyone in their family
had a serious mental or emotional problem or had been treated for
such problems. The highest score on this question was given to those
who said this had occurred in the past six months. This variable may
say something about the family environment in which young people live
and trauma to which they may have been exposed. Family mental/emotional
problems was generally much less powerful than emotional stress as
a predictor of suicide risk, except in the American Indian community,
where the Beta score of +.25 for family mental/emotional problems
nearly matched the Beta score of +.29 for emotional stress.
Several
other variables were associated with increased suicide risk in at least
one but not all of the five racial/ethnic groups. These included:
- Previous
history of physical abuse: Physical abuse had an independent effect
on suicide risk in the African American, Asian and white communities,
with students experiencing physical abuse more likely to report suicidal
thoughts or attempts.
- Unwillingness
to seek help for emotional or relationship concerns: For this
variable, adolescents were presented with a series of difficult emotional
or relationship situations, such as "depression," "feeling
out of control," and "relationships with girlfriend or boyfriend."
They were then asked who they would go to first for help with these
problems. Possible answers included parents, friends, school workers,
church leaders, other adults and "no one." Students who
more often said they would seek help from "no one" had higher
suicide risk scores. In the African American and Asian communities,
unwillingness or inability to seek help was an independent predictor
of suicide risk.
- Chronic
health conditions: In the African American, Chicano/Latino and
white communities, having to deal with chronic health conditions was
associated with greater suicide risk. Students were asked if they
had a variety of visible and invisible conditions, such as diabetes,
hypertension, epilepsy, hearing problems and physical disabilities.
Those who reported one or more conditions were more likely to have
suicidal thoughts or attempts.
- Strong
connections with adults and friends outside the family: In the
Asian community, the second most important factor related to suicide
risk was relationships with people outside the family (Beta score
of +.15). However, the relationship runs in the opposite direction
of what we would normally expect. The positive Beta score means that
youth who felt that school people, church leaders, adults and friends
cared about them were more likely to be at risk of suicide.
- Doing
OK in school: This variable combines answers to questions on feelings
about school, average grades, and skipping school. Students who said
they liked school, had average or above grades, and who did not skip
school often were considered to be doing OK in school. In the American
Indian community, however, youths who were doing OK in school by these
criteria were more likely to engage in suicidal thoughts and attempts.
- Being
female: In the white community, adolescent girls were more likely
to be at risk of suicide than boys, even after taking into account
the effects of emotional stress and other factors.
- Religiosity/spirituality:
Adolescents were asked to what extent they considered themselves to
be a religious or spiritual person. No questions were asked about
church affiliation or religious beliefs, nor was there any attempt
to explore what terms like "religion" or "spirituality"
mean to young people. "Religiosity" is therefore a very
subjective variable based solely on the youth's perception of themselves
as religious or spiritual. In the American Indian (-.15) and Chicano/Latino
(-.19) communities, this simple measure strongly contributed to lower
suicide risk. The negative Beta scores mean that youths who thought
of themselves as religious or spiritual were less likely to have suicidal
thoughts or attempts.
- Strong
family connectedness: Students who felt their families cared about
them and paid attention to them were at lower risk of suicidal thoughts
and actions in the Asian (.14) and white (-.08) communities.
- Year
or grade level in school: In the white community, students at
the higher grade levels (senior high) had lower suicide risk scores.
CONCLUSIONS
AND DISCUSSION
1. Several factors that describe emotional turmoil in different ways
were important predictors of high suicide risk. Emotional stress, a
variable that measures feelings, was the most important factor. But
certain experiences or events, such as a history of physical abuse or
the experience of recent emotional or mental health problems within
the family, also made an independent contribution to suicide risk.
The
latter finding reinforces the "whole family" approach to emotional
and mental health. Adolescents can be powerfully affected by the troubles
experienced by other members of the family. The whole family, not just
the person with emotional or mental health difficulties, may need strong
support and caring in these situations.
2.
An absence of strong caring relationships was related to suicidal thoughts
and gestures, at least in some communities. An unwillingness or inability
to seek help for emotional or relationship problems, perhaps having
no one to trust or to confide in, was associated with higher suicide
risk, while strong family connectedness helped to reduce suicide risk
in some communities.
However,
it appears that the variables describing caring relationships, or the
lack thereof, were less powerful in predicting suicide risk than they
were in explaining emotional stress. (See previous chapter.) Furthermore,
we remain puzzled by an unusual result within the Asian community-the
association of caring relationships with other adults and friends with
higher levels of suicide risk.
3.
The emergence of religiosity or spirituality as an important factor
helping to reduce suicide risk, at least among Chicano/Latino and American
Indian youth, leads to several questions that ought to be investigated
further. What are young people thinking of when they describe themselves
as religious or spiritual people? To what extent are young people talking
about organized religions and to what extent are they talking about
spiritual conceptions or experiences that fall outside the realm of
organized religions? What are the qualities of their religious or spiritual
lives that have an effect on their emotional lives? Similarly, the association
of certain chronic health conditions with suicide risk raises interesting
questions about the connections between physical and emotional health.
4.
Finally, it is very disturbing to find that American Indian youth who
are doing well in school are more likely to engage in suicidal thoughts
and gestures than those who are not. This result runs counter to common
wisdom and underscores the importance of looking closely at the different
experiences and dynamics within each racial/ethnic community. We might
well ask what are the pressures on American Indian students who stay
in school and do well that lead to higher suicide risk.
DELINQUENCY
BACKGROUND
Unlike emotional stress and suicide risk, which totally or partly measure
thoughts and feelings, delinquency is based strictly on behavior. Students
were asked how often they had been involved in various kinds of delinquent
behavior in the previous twelve months. The answers to eight of these
questions were combined to create an overall measure of delinquency.
The eight questions dealt with vandalism, fighting, theft, running away
from home and other behavior.
No
attempt was made in this survey to find out how serious each of these
offenses was, and very serious violent offenses such as aggravated robbery,
aggravated assault and homicide were not on the list. The measure of
delinquency used here emphasizes frequent involvement in a variety of
behaviors, rather than seriousness of the offenses. Another point to
keep in mind is that the survey obtained responses only from adolescents
who were in school. The findings reported here may not reflect patterns
of delinquency among young people who have dropped out of school.
Eleven
different variables were independently related to delinquency in at
least one of the five racial/ethnic groups.
FACTORS
ASSOCIATED WITH GREATER INVOLVEMENT IN DELINQUENCY
- Perceptions
of delinquent activity at school: In all communities except the
Asian community, youths who perceived that there was lots of delinquent
activity going on at their school were themselves more likely to be
involved in delinquency. This may be because certain schools had exceptional
amounts of unlawful behavior or because delinquent youths were more
aware of delinquent behavior going on around the school than other
kids.
- Recent
mental/emotional problems within family: This variable was an
important predictor of suicide risk, and in the African American (+.10),
Asian (+.11) and white (+.05) communities it was also a predictor
of delinquency. Delinquency was more prevalent among youth who said
one or more family members had recent mental health or emotional problems
- or had received treatment for such problems.
- Physical
abuse: Students who reported previous physical abuse were more
likely to engage in delinquent activity in the American Indian (+.22)
and white (+.07) communities. Among American Indians, physical abuse
was the most powerful predictor of delinquency.
- Multiple
chemical use by parents: Students were asked if their parents
used four kinds of drugs once a week or more often. The four were
tobacco, alcohol, marijuana and crack or cocaine. A score was created
by counting the number of chemical groups used during a typical week
by the parents. In the Asian (+.24) and African American (+ .1 4)
communities, youths involved in delinquency had parents who were more
likely to be involved with a greater number of different chemicals.
This proved to be the most powerful predictor of delinquency in the
Asian community.
- Chronic
health conditions: Students with one or more chronic health conditions
were more likely to engage in delinquent offenses in the African American
and white communities.
- High
emotional stress: High levels of emotional stress, a powerful
predictor of suicide risk, contributed to delinquency involvement
only in the Asian community.
FACTORS
REDUCING DELINQUENCY
- Being
female: The only variable that predicted the level of delinquency
involvement in all five racial/ethnic groups was female gender. The
negative Beta scores in Table 4 indicate that female adolescents were
much less likely to be involved in delinquency, even after taking
into account all of the other variables or factors related to delinquency.
In the African American (-.19), Chicano/Latino (.34) and white (-.40)
communities, being female was the most important single factor predicting
high or low delinquency involvement.
- Being
older: In each community except the Chicano/Latino community,
age was a predictor of delinquency. Older adolescents were less involved
in delinquency than younger adolescents. Other studies have often
found that delinquency-at least the kinds of behavior that were included
in the Adolescent Health Survey-is more common in early adolescence
and decreases with age. It is also possible that many of the more
delinquency-prone adolescents dropped out during high school and were
not counted in the survey.
- Turning
to parents for help: As mentioned earlier, students were presented
with several difficult emotional or relationship situations and asked
to whom they would go for help. In the African American (-.14), American
Indian (-.17) and white (-.11) communities, students who sought help
first from their parents were less likely to engage in delinquent
activities.
- Strong
family connectedness: Higher levels of family connectedness and
caring were associated with lesser amounts of delinquency in the Asian
and white communities.
- Receiving
free or reduced-price school lunch: In the Asian community, students
who received free or reduced-price lunch an indicator of low family
income - were less likely to be involved in delinquency.
CONCLUSIONS
AND DISCUSSION
1. The results for the delinquency outcome variable were less helpful
than the results for emotional stress and suicide risk. Two of the most
powerful contributors to lower amounts of delinquency were demographic
variables - gender and age - which are unchangeable. The study does
not tell us what it is about being female or being older that leads
to less delinquency.
2.
Measures of caring relationships with other people appear to help reduce
delinquency in some cases. In the African American, American Indian
and white communities, willingness to go to one's parents for help with
emotional or relationship problems was related to lower delinquency.
In the Asian and white communities, strong family connections were associated
with lower delinquency.
3.
Measures of emotional turmoil emerged as independent predictors of delinquency
in some communities. However, emotional stress, which is based on feelings,
was an important predictor in only one community, the Asian community.
Experiences with recent emotional or mental health problems in the family
or experiences with past physical abuse were associated with higher
levels of delinquency in some communities. It appears, however, that
these measures of emotional turmoil are considerably weaker at predicting
delinquency (acting out behavior) than they were in predicting suicidal
thoughts and gestures.
4.
Chemical use by parents, specifically the regular use by parents of
several different kinds of chemicals ranging from tobacco to illegal
drugs, appears to contribute strongly to delinquency in two communities
- the African American and Asian. This is the only occasion in this
study where information related to chemical use was found to be an important
predictor of an outcome variable;
RESILIENCY
AND SOCIAL CHANGE:
THE CONTEXT FOR YOUTH POLICY
Those
who study child and adolescent health use the idea of resiliency to
mean the capacity of young people to thrive, to be emotionally healthy
and to avoid destructive behavior in spite of very difficult circumstances.
There
are several good academic reviews of the research on resiliency. These
include:.
Ann
S. Masten, "Resilience in Individual Development: Successful
Adaptation Despite Risk and Adversity," in M.C. Wang and E. Gordon
(eds.), Education Resilience in Inner-City America: Challenges and
Prospects, Hillsdale, NJ: Lawrence Erlbaum (forthcoming).
J.
David Hawkins, Richard F. Catalano, and Janet Y. Miller, "Risk
and Protective Factors for Alcohol and Other Drug Problems in Adolescence
and Early Adulthood: Implications for Substance Abuse Prevention,"
Psychological Bulletin, 1992, Vol. 112, No. 1: 64- 105.
Ann
S. Masten, Kann M. Best, and Norman Garmezy, "Resilience and
Development: Contributions from the Study of Children Who Overcome
Adversity," Development and Psychopathology, 2 (1990): 425-444.
This
study has tried to identify some of the factors that increase the likelihood
of high emotional stress and damaging behavior and some of the protective
factors that contribute to resiliency in adolescents.
One
purpose of the study was to look at how risk and protective factors
operate within different racial/ethnic groups. We did not want to assume
that the dynamics affecting emotional stress and destructive behavior
would be the same in each community. In fact, unique and varied findings
did emerge in each community. Another glance at Tables 2-4 provides
many examples of factors that were significant predictors of emotional
stress, suicide risk or delinquency for some racial/ethnic groups but
not for others. Special attention to the dynamics in different groups
allowed us to find unexpected relationships that would otherwise have
been lost. One of the most disturbing of these was the finding that
American Indian youth who were doing relatively well in school were
more likely to report suicidal thoughts or attempts.
In
addition to differences, the study also identified several universal
and cross-cutting factors that had an impact on outcomes for adolescents
in every racial ethnic group. This was most true for emotional stress,
where strong connections with the family emerged as a powerful factor
promoting resiliency across-theboard, and where worries about violence,
negative body image, and boredom (nothing to do) were consistently associated
with greater stress. High emotional stress was, in turn, an important
predictor of suicide risk in every racial/ethnic group, as was the experience
of dealing with recent major emotional or mental health problems within
the family.
It
is striking, though, how many of these factors are linked to personal
relationships and to the adolescent's sense of belonging. Strong family
connections has already been mentioned as a universal factor helping
to reduce emotional stress. At least in some communities, caring relationships
with people outside the family also helped reduce emotional stress,
and being able to turn to parents for help was associated with lower
delinquency. Conversely, mental and emotional problems within the family,
physical abuse, and having no one to turn to for help seemed to contribute
to greater stress or destructive behavior. Addressing this fundamental
adolescent, and human, need for caring relationships and belonging must
be an integral feature of all public policies, programs and interventions,
no matter what their specific focus or content.
While
there is much excitement about the search for ingredients that strengthen
resiliency, there is a recognition that this is not a magic cure all
and should not be the sole focus of adolescent health. Our collaborators
on this project from the University of MN point out in a separate article:
Michael Resnick, Linda J. Harris and Robert W Blum, "The Impact
of Caring and Connectedness on Adolescent Health and Wellness,"
in Journal of Pediatrics and Child Health, 1993, 29, Suppl. 1:1-9.
"Caring,
while extraordinarily important in the lives of young people, is not
a substitute for correcting fundamental threats to health, rooted
in the economic disparities that have become increasingly manifest
due to both deliberate government policies and a shifting economic
infrastructure that strains the ability of families and individuals
to thrive or function. In fact, the sense of disaffiliation of growing
numbers of young people, noted with alarm by many commentators, is
accentuated by trends in the economy which make it increasingly harder
for families to earn sufficient income with benefits, to meet the
costs of housing, food and everyday living."
Two
streams of activism, therefore, need to come together. The structural
issues -- creating good jobs, health insurance, child care, education,
job training, tax reform and more -- must move ahead. But they must
be joined by policies and programs that address the specific kinds of
risk and protective factors identified in this study.
