ERIC ED378471: The Life Attitudes Schedule: A Theoretically Derived Scale To Assess Adolescent Life-Enhancing, Life-Compromising, and Suicidal Behaviors.

55 Pages·1994·0.59 MB·English

by  ERIC

DOCUMENT RESUME
CG 025 869
ED 378 471
Lewinsohn, Peter; And Others
AUTHOR
The Life Attitudes Schedule: A Theoretically Derived
TITLE
Scale To Assess Adolescent Life-Enhancing,
Life-Compromising, and Suicidal Behaviors.
PUB DATE
[94]
56p.
NOTE
Research/Technical (143)
Reports
PUB TYPE
MF01/PC03 Plus Postage.
EDRS PRICE
Adolescents; At Risk Persons; *Attitude Measures;
DESCRIPTORS
Evaluation Methods; High Risk Students; Satisfaction;
Secondary Education; Self Injurious Behavior;
*Student Attitudes; *Suicide; Well Being
ABSTRACT
Adolescent engagement in risk-taking and
life-compromising behaviors has become a public health concern. This
forms of a new
paper describes the development of three alternate
scale to measure suicidal and risk-taking behavior, the Life
Attitudes Schedule (LAS). The LAS, which was developed and piloted on
1,539 high-school students and young adults, is based on a broad
conceptualization of suicidal behavior. It is unique because it is
derived from a broad theoretical perspective. The LAS measures four
(2) health-related;
different content categories: (1) death-related;
injury-related; and (4) self-related. Each content category
(3)
includes an equal number of items designed to assess actions,
thoughts, and feelings. The LAS was resigned to measure a bipolar
continuum of behavior; consequently, an equal number of positive
(life-enhancing) and negative (life-compromising) items are included.
Three approaches were used to evaluate the validity of the LAS:
correlational analyses, construct and criterion validity analyses,
and structural equation modeling to test several theoretical models
underlying the LAS. The results indicate both that this instrument
has excellent reliability and validity properties and provides
validation of the construct of life-compromising and life-enhancing
behaviors. Six tables present sample items and statistical summaries.
Contains 40 references. (Author/RJM)
***********************************************************************
Reproductions supplied by EDRS are the best that can be made
from the original document.
************************************************************).**********
The Life Attitudes SC:a
The Life Attitudes Schedule:
A theoretically derived scale to assess adolescent
life-enhancing, life-compromising, and suicidal behaviors
Peter Lewinsohn
Oregon Research Institute, Eugene, OR
Jennifer Langhinrichsen-Rohling
University of Nebraska, Lincoln
Richard Langford
Humboldt State University, Arcata, CA
John Seeley
Paul Rohde
Oregon Research Institute, Eugene, OR
Jean Chapman
University of Wisconsin, Madison
LIFE ATTITUDES SCHEDULE
Running Head:
BEST COPY AVAILABLE
DRAFT -- Comments welcome
U.S. DEPARTMENT OF EDUCATION
"PERMISSION 10 REPRODUCE THIS
Office of Educational Research and Improvement
MATERIAL HAS BEEN GRANTED BY
EDUCATIONAL RESOURCES INFORMATION
CENTER (ERIC)
L ow
p,
I N SagA/
0 This document has been reproduced as
received from the person or organization
originating it
0 Minor changes havr been made lo improve
reproduction quality
Points of view or opinions stated in this docu
TO THE EDUCATIONAL RESOURCES
ment do not necessarily represent official
OE RI position or policy
INFORMATION CENTER (ERIC);
The Life Attitudes Schedule
Page 2
Abstract
This paper describes the development of a new scale to
behavior, the Life Attitudes
measure suicidal and risk-taking
The LAS, which was developed and piloted on
Schedule (LAS).
based on a broad
1,539 high school students and young adults, is
The schedule measures
conceptualization of suicidal behavior.
death-related, health-
four different content categories:
Each content category
related, injury-related, and self-related.
actions,
includes an equal number of items designed to assess
The LAS was designed to measure a
thoughts, and feelings.
equal number of
bipolar continuum of behavior; consequently, an
items
positive (life-enhancing) and negative (life-compromising)
has
The results indicate both that this instrument
are included.
provides
excellent reliability and validity properties and
life-
validation of the construct of life-compromising and
enhancing behaviors.
3
The Life Attitudes Schedule
Page 3
A theoretically deri-ved scale
The Life Attitudes Schedule:
life-enhancing, life-compromising,
to assess adolescent
and suicidal behaviors
life-compromising
Adolescent engagement in risk-taking and
Adolescent
behaviors has become a public health concern.
risk-taking behaviors in
suicidal behavior in particular and
(e.g., Scott &
general appear to be increasing in recent years
Suicide is currently the second most frequent
Cabral, 1988).
15 to 24 years of age, and
cause of death for acolescents
with risk-taking
accidents, which are frequently associated
of death in
behaviors and substance use, are the leading cause
As
Statistics, 1991).
this age group (National Center for Health
of risk-taking in
yet another indicator of the high rate
is often a consequence of
teenagers, the rate of AIDS, which
.intercourse, IV drug use,
behaviors such as unprotected sexual
inhibitions, is thought to
and the use of substances which reduce
(Biglan et al., 1990;
be doubling each year in adolescents
It is important to understand
Brooks-Gunn, Boyer, & Hein, 1988).
(Irwin & Millstein, 1986), as
the etiology of these behaviors
assessing youth who are at
well as identify better methods of
life-threatening behaviors.
high risk for suicidal or potentially
the development of
Consequently, in this paper, we report on
instrument, the Life Attitudes
three parallel versions of a new
The LAS was designed on the basis of a
Schedule (LAS).
at risk for engaging in
theoretical model to identify adolescents
4
The Life Attitudes Schedule
Page 4
life-compromising and life-enhancing
a broad array of potentially
behaviors.
aid in
Psychometrically sound instruments are necessary to
life-compromising
the prediction and prevention of suicidal and
available
However, in a recent review of the
behaviors.
Marsteller,
assessment devices (Garrison, Lewinsohn,
Langhinrichsen, & Lann, 1991; Lewinsohn, Garrison,
Langhinrichsen, & Marsteller, 1989), we concluded that current
instruments have a number of limitations and conceptual concerns.
suicidal
First, researchers vary widely in their definition of
Some suicide researchers measure related constructs
behavior.
Weissman,
such as depression and hopelessness (e.g., Beck,
1987), while others
Lester, & Trexler, 1974; Harter & Nowakowski,
behaviors such
choose to define the construct broadly, including
risk-taking, asceticism,
as accident-proneness, unnecessary
the
polysurgery, behavior leading to victimization, and even
one's life
failure to engage in behaviors that serve to protect
More narrow
death or injury.
or reduce the probability of
definitions of suicidal behavior restrict the construct to
overdose) or
behaviors that lead to immediate death (e.g., drug
oneself or self-
at least serious self-injury (e.g., cutting
mutilation) .
models underlying
The variety of often implicit theoretical
available instruments
these measures has made the comparison of
currently
For example, all the following terms are
difficult.
5
The Life Attitudes Schedule
Page 5
the range of behaviors
recognized within the field and represent
ending
completed suicide (consciously
that can be assessed:
in
suicide (similar to completed suicide
one's life); attempted
actions have taken place; however,
that intentional conscious
suicide gesture or parasuicidal
they did not lead to death);
is not actually life-threatening);
behavior (a symbolic act that
something that indicates a self-
suicide threat (saying or doing
ideation (thoughts about
destructive desire); and suicidal
killing oneself).
of
noted was that the vast majority
A second limitation we
from inattention to validity
the current instruments suffer
it has often been
Where validity has been assessed,
issues.
instrument with measures of
limited to correlating the new
depression and hopelessness (e.g.,
related constructs such as
Mizruchi, & Weiner, 1982; Pfeffer,
Pfeffer, Solomon, Plutchik,
In our initial work,
1984).
Zuckerman, Plutchik, & Mizruchi,
the LAS scores and measures of
expected relations between
Thus, we did not expect
found.
depression and hopelessness were
unassociated
scale that was completely
to be able to create a
needs to be
However, we believe that more
with these constructs.
unique
validity and to determine the
done to establish construct
by the
of suicidal behavior as assessed
aspects of the domain
discriminate and criterion validity.
instrument, as well as its
instruments have not been given to
Finally, many of the existing
establish item
number of adolescents to
a large enough
The Life Attitudes Schedule
Page 6
and gender scale
reliability, item to scale relations, and age
norms.
literature we concluded
On the basis of our review of the
reliable and valid
that there continues to be a need for a
identify adolescents at
assessment device that can efficiently
Ideally, this scale should include a
risk for suicidal behavior.
behaviors, be
broad range of potentially life-compromising
of depression
moderately but not highly correlated with measures
attempt within the limits
and hopelessness, and predict a suicide
Finally such an
imposed by the low base rate of such behaviors.
is can be
instrument should be relatively short so that
and non-clinical
administered quickly and easily in clinical
(e.g., school) settings.
to test a theoretical
A second goal of the present study was
suicide-proneness, which was
construct that we initially labeled
propensity at a point in time
assumed to be to reflect a person's
In keeping with others (e.g.,
to engage in suicidal behavior.
19.85; Smith & Crawford,
Farberow, 1992; Menninger, 1938; Pfeffer,
included subtle/non-
1986), our definition of suicidal behavior
violence-provoking
obvious self-destructive behaviors (e.g.,
behaviors that are
behaviors), risk-taking behaviors, as well as
at oneself).
obviously and overtly suicidal (e.g., pointing a gun
expanded further by the
The domain of suicidal behavior was
by life-extending
addition of a positive pole characterized
review and theoretical
On the basis of our literature
behaviors.
7
The Life Attitudes Schedule
Page 7
considerations, we generated operational definitions for four
content categories which were assumed to be sufficient to
death/life-related
encompass the domain of suicidal behavior:
(DR), health/illness-related (HR), injury-related (IR), and self-
For each content category both a negative (life-
related (SR).
compromising) and a positive (life-enhancing) pole were
Following traditional distinctions in psychology
identified.
Staats & Staats,
(e.g., Bandura, 1977; Smith & Crawford, 1986;
thoughts, and
1963), we also distinguished between actions,
feelings.
Test Construction Framework
The LAS was constructed using a sequential multi-stage
As has been recommended by Millon (1983), Jackson
process.
(1970), and Tellegen (Zevon & Tellegen, 1982) among others, we
maintained a close relationship between theory, test development,
As
reliability, validity, and eventual clinical utilization.
of
already indicated, we began by delineating the broad construct
On theoretical
suicide-proneness as precisely as possible.
grounds, we hypothesized the presence of a general bipolar factor
that would appear as a first principal component with significant
We also theorized that suicide-
loadings on all of the items.
population
proneness would be normally distributed in the general
being identified either as
with only a small percentage of people
Items were
dimension.
extremely high or extremely low on the
categories and we expected
generated to fit into the four content
The Life Attitudes Schedule
Page 8
factors.
the latter to emerge as relatively orthogonal
(e.g., actions,
Similarly, we expected the three behavior types
factors.
thoughts, and feelings) to define separate
Method
Subjects
Life
In 1989, six initial versions of the
Sample 1.
separate
Attitudes Schedule (LAS) were administered on two
who were
occasions to 757 predominantly Caucasian students
California.
enrolled at a large, urban high school in northern
female.
Slightly more than half of the subjects (58%) were
the mean age of
Subjects ranged in age from 13 to 18 years with
There were approximately equal numbers of freshman,
16.0 years.
sophomores, juniors, and seniors.
The second sample consisted of 104 college
Sample 2
in 1989 to partially
undergraduates who voluntarily participated
their introductory
fulfill a research credit requirement for
Seveny-two percent of the subjects were
psychology courses.
with a mean
Subjects ranged in age from 18-42 years old
female.
Asian or
Most (78%) were Caucasian, 14% were
age of 20.8 years.
2% were Hispanic,
Pacific Islanders, 2% were African American,
and 4% were other.
recruited from two
In 1990, 32 adolescents were
Sample 3.
described in greater detail
ongoing research studies which are
& Andrews, 1993;
Seeley,
elsewhere (Lewinsohn, Hops, Roberts,
All
& Seeley, in press).
Lewinsohn, Clarke, Rohde, Hops,
The Life Attitudes Schedule
Page 9
(n = 16) were
Half of the subjects
subjects were Caucasian.
participants in a treatment outcome study of adolescent
depression (mean age = 15.3; 73% female) who met DSM-III-R
A
criteria for major depression (n = 11) or dysthymia (n =
5).
recruited
matched (age and gender), control group (n = 16) was
study of
from an ongoing community based epidemiological
In addition to matching on age and gender, the
psychopathology.
psychiatric diagnoses.
control subjects had no past or current
All subjects were paid for their participation.
In 1993, another 646 high school students from
Sample 4.
participated to
the same high school utilized in Sample 1
versions of
determine the psychometric properties of three new
Approximately half of the subjects were female (48%).
the LAS.
There were approximately
The majority (80%) were Caucasian.
(28%), juniors (20%),
equal numbers of freshmen (32%), sophomores
After one month, 412 (64%) of the
and seniors (20%).
participants repeated the procedure.
Measures
As illustrated
Prototype for the Life Attitudes Schedule.
life-
in Table 1, half the items were designed to assess
half assessed
compromising (negative) behaviors while the other
Four content category
life-enhancing (positive) behaviors.
death-related (DR),
subscales were incorporated into the LAS:
self-related (SR).
health-related (HR), injury-related (IR), and
and death-related items
The DR items included traditional suicide