ERIC ED373073: Application of a Theoretical Model to the Development of a New Instrument for Assessing Self-Destructive Potential: The Firestone Voice Scale for Self-Destructive Behavior.

14 Pages·1993·0.47 MB·English

by  ERIC

DOCUMENT RESUME
TM 021 810
ED 373 073
Firestone, Lisa
AUTHOR
Application of a Theoretical Model to the Development
TITLE
of a New Instrument for Assessing Self-Destructive
Potential: The Firestone Voice Scale for
Self-Destructive Behavior.
Glendon Association, Los Angeles, CA.
INSTITUTION
PUB DATE
93
16p.; Paper presented at the Annual Meeting of the
NOTE
California Association for Counseling and Development
(1993).
Reports Research/Technical (143)
PUB TYPE
Speeches/Conference Papers (150)
MF01/PC01 Plus Postage.
EDRS PRICE
*Behavior Patterns; Correlation; Models; Scores;
DESCRIPTORS
*Self Destructive Behavior; *Test Construction; Test
Reliability; Test Validity; Theories
*Firestone Voice Scale Self Destructive Behavior;
IDENTIFIERS
Firestone Voice Theory
ABSTRACT
The concept of the "voice" as developed by Robert W,
Firestone (1984 and later) has been hypothesized as an essential
mechanism in self-destructive behavior in general and suicide in
particular. This study applies the theory of the voice to the
development of the Firestone Voice Scale for Self-Destructive
Behavior (FVSSDB). The FVSSDB consists of 110 items equally drawn
from 11 levels of progressively self-destructive thoughts. To
investigate the reliability and validity of the FVSSDB, it was
administered to 507 subjects currently in psychotherapy. Respondents
also completed a battery of nine other instruments covering diverse
areas of self-destructiveness in order to assess construct validity.
Results were consistent with a Gutman scale of increasing
self-destructiveness, providing support for the hierarchical and
continuous nature of self-destructiveness. Construct validity was
suggested by significant correlations between the levels and
corresponding instruments. Criterion validity was supported by a high
correlation of scores with past suicide attempts. Overall results
reflect favorably on the voice theory and the hypothesis that
assessing the level of destructive voices contributes to an
understanding of suicide potential. (Contains 38 references and 2
figures.) (Author/SLD)
***********************************************************************
*
*
Reproductions supplied by EARS are the best that can be made
*
*
from the original document.
******************************************* ,**************************
U.S. DEPARTMENT OF EDUCATION
"PERMISSION TO REPRODUCE THIS
Office of Educatatnet Research and Imcworament
MATERIAL HAS BEEN GRANTED
EDUCATIONAL RESOURCES INFORMATION
BY
CENTER (ERIC)
document has been reproduced as
EIKTC"::s
re:timed from O. person or ononizathon
oncanating
0 Mmor chang411% Key. bowl made lo ornprOve
riproduction Qua MY
Points of wear or opmona stated in this docu-
ment Oo not necessenly represent ollicaol
TO THE EDUCATIONAL RESOURCES
OE RI oosaton or peaty
INFORMATION CENTER (ERIC)."
APPLICATION OF A THEORETICAL MODEL TO THE
DEVELOPMENT OF A NEW INSTRUMENT FOR
ASSESSING SELF-DESTRUCT1VE POTENTIAL:
The Firestone Voice Scale for Self-Destructive Behavior
LISA FIRESTONE, Ph.D.
Los Angeles, California
ABSTRACT
INTRODUCTION
Thc concept of the "voice- as developed by
Suicide is a problem of considerable magnitude.
Robert W. Firestone (1984: 1986: 1988: Firestone &
It is estimated that there are several hundred thousand
suicide attempts in this country each year and that a
Seiden. 1987) has been hypothesized to be an essen-
tial mechanism in self-destructive behavior in gen-
total of 5 to 6 million individuals have made suicide
attempts. In 1988, there were slightly more than
eral and suicide, in particular. This study applies R.
Firestone's theory of the "voice- to the development
30.000 suicides annually (83 suicides per day. or I
of the Firestone Voice Scale for Self-Destructive
suicide every 17 minutes). with 12 of every 100,000
Behavior (FVSSDB).
Americans killing themselves (National Center for
The FVSSDB consists of 110 items equally
1990). Follow-up studies have
Health Statistics.
drawn from 11 levels of progressively self-destruc-
shown that 10-20cC of suicide attempters who were
tive thoughts. To investigate the reliability and valid-
hospitalized following the attempt go on to kill them-
selves within a 10-year period (Dorpat & Ripley,
ity of the FVSSDB. it was administered to 507 sub-
jects currently in psychotherapy. Respondents also
1967). In addition, retrospective studies have shown
that between 20 and 65% of those who kill them-
completed a battery of nine other instruments cover-
ing diverse areas of self-destructiveness, in order to
selves have a history of prior attempts. Dorwart and
assess construct validity.
Chartock (1989) unequivocally stated that the best
The results .A.ere consistent with a Guttman scale
predictor of subsequent suicide attempts and comple-
of increasing self-destructiveness, providing support
tions is having a history of previous attempts. Jacobs
for the hierarchical and continuous nature of self-de-
(1989) states that "any suicidal behavior, regardless
structiveness. Construct validity of the FVSS DB was
of severity, places a person at 10 to WO times more
suggested by significant correlations between the
than the normal risk for suicide" (p. 370).
An individual in the grip of a suicidal crisis is
levels and corresponding instruments. Criterion va-
lidity was supported by the high correlation of scores
deeply ambivalent about taking his/her life. He/she is
on the FVSS DB with fast suicide attempts. Logistic
divided within him/herself: one part wants to live,
regression revealed tht the FVSSDB adds signifi-
while another part wants to die. As clinicians, it is our
cantly to the ability to discriminate prior suicide
responsibility to apreal to and support the part of the
attempts beyond the instruments used. Overall, the
person that wants to live. Leonard (1967) summarizes
"Their 'right' is not to
this sentiment by stating:
results reflect favorably on the "voice- theory and the
original hypothesis that assessing the level of de-
commit suicide but to have their need for psychologi-
cal ossistance met so that they may enjoy a satisfying
structive "voices- contributes to an understanding of
life among us-
suicide potential.
(p. 223).
The Glendon Association
[phone]
1
2
The risk of suicide, including attempts and com-
continuum encompassing those behaviors, commu-
pletions, is disproportionately high among mental
nications, attitudes. or life-styles that are self-limit-
health clients already in treatment (Nekanda-Trepka.
ing, threatening, or antithetical to an individual's
Bishop. & Blackburn, 1983). Studies revealed that
emotional well-being and physical health. These be-
one in five practicing clinical psychologists will lose
haviors have been referred to by Firestone and Seiden
a client to suicide, and the number rises to one in two
(1987) as "microsuicidal." A number of theorists
for psychiatrists (Chemtob, Hamada. Bauer. Kinney
have supported this premise. including Menninger
& Torigoe. 1988: Chemtob, Hamada. Bauer, Torigoe,
(1938), Farberow (1980). and Shneidman (1966). (2)
& Kinney, 1988). Many factors have a potential to
Destructive "voices." ranging from mildly critical
interfere with a clinician's judgment in the highly
attitudes to malicious attacks on the self, exist in
emotional situation of dealing with a client who is
conjunction with the above premise. (3) There is a
potentially suicidal.
conflict within each individual between life-affirm-
Therapists tend to rely almost exclusively on
ing propensities to actively pursue goals in the reai
intuitive sense to determine the dangerousness of a
world, and self-denying, self-protective, and eventu-
suicidal crisis (Maltsberger & Buie. 1989: Jobes.
ally self-destructive tendencies that revolve around
Eyman. & Yufit. 1990). Maltsberger (1986) also
internal gratification through fantasy processes. The
states that dangerous misjudgments can result from
latter tendencies represent a defensive process within
excessive reliance on clinical intuition. A comprehen-
the personality. These incorporated parental attitudes
sive study (Bongar & Harmatz, 1989, 1990) found
come to have their own functional autonomy in the
that only 40% of all graduate programs in clinical
adult personality. Suicide represents the acting out of
psychology provide any formal training in the study
the extreme end of this self-destructive part of an
individual. The "voice"
of suicide. Clinicians are often ill-equipped to deal
Freud's
similar
to
is
with a client's suicidal crisis. They lack training as to
(1921/1955) concept of an overly harsh super-ego
how to cope with these life-and-death situations.
and even more closely aligned with Guntrip's (1969)
Therefore, clinicians in particular need a prompt,
concept of an antilibidinal ego.
easy to administer, and thorough clinical assessment
R. Firestone 1988) explained the split or inter-
strategy for assessing suicidal risk, as well as a con-
nal division existing within each individual as con-
ceptual model for understanding the suicidal individ-
sisting of the self-system and the anti-self system.
The self-system is an individual's unique set of
ual.
The concept of the "voice," as developed by Dr.
wants, desires, and priorities, based on his or her
Robert W. Firestone. provides an innovative explana-
physical and mental attributes, as well as the identi-
tion of the underlying dynamics of self-destructive
fication with parents' positive traits that are harmoni-
behavior. In particular, the "voice" has been hypothe-
ously assimilated into the ego or self. The anti-self
sized to be an essential mechanism in suicidal idea-
system is an integrated system of defensive. self-criti-
tion and behavior (R. Firestone, 1986). The voice has
cal, and self-destructive attitudes internalized as the
been defined as an integrated system of negative
"voice." In other words, the voice process represents
the incorporation and internalization of parents'
thoughts and attitudes, antithetical to self and cynical
toward others. The Firestone Voice Scale for Self-De-
negative attitudes and hostility that the child experi-
structive Behavior (FVSSDB) is an outgrowth of a
enced while growing up. Thus, this alien posture
comprehensive theory of psychopathology and a
toward self and others is originally imposed upon the
comparative model of mental health. An empirical
personality from the external world; it persists into
study was undertaken by the present author to esta-
adult life and colors all interactions and pursuits.
blish the scale's construct and criterion validity.
Attitudes of self-hatred, microsuicidel
and sui-
,
There are three premises underlying R. Fire-
cidal propensities cannot be successfully integrated
stone's approach to self-destructive behavior and sui-
ituo the personality, since they are opposed to the
cide: (1) Self-destructive behavior exists along a
ongoing life of the personality. If these forces as
"Microsuicide- refers to behaviors, communications, attitudes, or life-styles that are self-induced and threatening or
inimical to an indtvidual's physical health, emotional well-being, or personal goals (Firestone & Seiden, 1987).
The Glendon Association
2
[phone]
3
clinicians' ability to predict which individuals will
hypothesized have as a function the ultimate destruc-
tion of the personality or even the physical life of the
commit suicide, without intervention. In addition, the
person, how can they be integrated into the personal-
theory underlying the scale provides clinicians with
ity they oppose? Furthermore, these thoughts and
a comprehensive framework for investigating and
attitudes are not innate in the personality: instead they
understanding the problem of self-destructive behav-
were imposed from without through the process of
ior.
introjection and remain as an overlay on the person-
ality.
ADVANTAGES OF THE FVSSDB
The voice process as conceived by R. Firestone
ranges from unconscious or subliminal to fully con-
The FVSSDB is an instrument designed to as-
scious. It represents a discordant force within the
sess the level of self-attacks a person is experiencing
personality wherein the self becomes the object of
along the Continuum of Negative Thought Patterns
attack and punishment. The dynamics predisposing
(Firestone & Seiden. 1990). The items or statements
the critical voice and self-destructive life-styles are
on the FVSSDB consist of attacks underlying various
multidetermined but focus on these principal areas:
levels of self-destructive behavior, ranging from self-
(a) the voice process involv,::: the incorporation of
denial, isolation, eating disorders, substance abuse.
and self-mutilation. to actual injunctions to commit
parental arAudes and defenses (often unconscious):
(b) the voice represents an attempt to protect the
suicide.
individual from feeling anxious and vulnerable
Thus. it was believed the scale could be benefi-
cial in providing information about each of these
t hrough a complicated process of predetermining and
rehearsing negative outcomes, thus discouraging the
issues. For example. clinicians using the scale could
person from engaging in challenging behavior, (c)
determine the level at which a patient endorsed items
with the highest frequency, thereby identifying the
the voice has an additional defensive function in that
it serves as a self-denying accommodation to death
focus of self-destructive behavior for the individual.
Another potential advantage of the scale relates
anxiety: and (d) lastly. and most importantly, it repre-
to the particular format in which the negative
sents an "identification with the aggressor--the
introjection of parents' covert hostility. thoughts are stated on the questionnaire. By present-
ing the voice statements in the second person. the
In his ongoing investigations into the voice pro-
cess. R. Firestone (1988) noted that subjects were
individual brings to the surface a partially subcon-
scious process, allowing for greater insight and in-
able to trace the origins of their self-attacks to early
family interactions. They identified critical voice
creased power to cope. In addition, understanding
patients' voices contributes to immediate rapport.
statements as parental warnings, directions. labels.
definitions, and feelings that they had assimilated
The scale provides a valuable window into the self-
into their own thinking process during their formative
destructive process. in a 15-year longitudinal clinical
years.
study utilizing Voice Therapy as a laboratory proce-
dure, R. Firestone (1986) observed that becoming
Firestone and Seiden (1987) observed that self-
conscious of self-destructive thoughts and attitudes
attacks of the voice vary along a continuum of inten-
sity from mild self-reproach to strong self-accusa-
gave individuals a measure of control over self-de-
tions and suicidal thoughts. The voice thus becomes
structive behaviors that were previously acted out.
the mechanism that regulates and dictates a person's
Answering the FVSSDB could also open up the
self-denying, microsuicidal, and ultimately suicidal
client to discuss his/her negative cognitions with the
therapist from the onset of therapy, a discussion that
behavior.
enhances the therapist's understanding of his or her
Based on this theoretical approach, R. Firestone
and the staff members of the Glendon Association.
client. The client's responses to the scale would pro-
vide valuable information about each person's self-
including the author, instituted plans to develop the
destructive thought process, as well as his/her poten-
Firestone Voice Scale for Self-Destructive Behavior.
We initiated the project with the belief that we could tial for serious self-destructive behavior.
Therefore, the FVSSDB is directly tied to a
develop a valuable instrument by assessing the depth
facilitates the first step of
and degree of the "voices- a person is experiencing.
treatment approach. It
Voice Therapy, which involves the person in identi-
Indeed, it was believed that the scale would enhance
The alendon Association
[phone]
3
4
fying his/her self-attacks. If the therapist were to
Another important reason for this research was
proceed by encouraging the client to say these self-
to provide empirical support for the personality the-
ory advanced by R. Firestone (1988), thus adding
attacks out loud, further information could be gained
as to the seriousness of the person's self-destructive
credence to a comprehensive perspective on human
behavior and human interaction. If the scale dis-
potential. The therapist accomplishes this by asking
played the capacity to successfully distinguish be-
the client to verbalize his/her negative thoughts in the
second person. as though he/she were another person
tween various patterns of self-destructive behavior
talking to the self. The intensity of angry affect asso-
manifested by patients, it would provide this empiri-
ciated with these attacks becomes obvious when they
cal support.
are verbalized. They indicate another important
Information on the scale's ability to distinguish
measurement of the strength of the incorporated hos-
between various patterns of self-destructiveness was
provided from two sources: one, a comparison of
tile point of view. The scale accesses hostility in the
Level 5 items (vicious self-abusive thoughts): the
clients' scores on the levels of the FVSSDB with
therapists' reports of the different forms of self-
Level 9 items (injunctions to inflict injury on self):
and, of course, Level 11 items (injunctions to carry
destructive behavior: and two, a comparison of scores
out suicide plans). As stated earlier, the strength or
with standard measures of these same self-destruc-
intensity of voice attacks reflects the degree to which
tive behavior patterns.
this °negative aspect of the personality is dominant
There are several scales which have been de-
and the seriousness of one's potential for suicide.
veloped to assess elements of suicidal intention. Un-
To summarize. R. Firestone's (1988) approach to
fortunately, none has proven very effective in predict-
ing suicide. This is partially due to the fact that
self-destructive behavior sets forth a conceptual mo-
del that therapists can utilize to better understand
suicide is such a low base rate phenomenon. It entails
their patients. This model provides the therapist with
predicting which persons will exhibit a highly speci-
fic, very infrequently occurring behavior.
ideas about the direction in which the therapy should
Clinicians have revealed that they rarely make
proceed. The use of this instrument (the FVSS DB) in
use of suicide scales, feeling that they can rely on
clinical settings as an adjunct to diagnosis leads natu-
rally to the utilization of Voice Therapy procedures.
clinical intuition (lobes, Eyman, & Yufit, 1990). Part
of the reason for this may be that most of these
In particular, it can be used as a therapeutic tool to
help clients identify the extent and origins of their
self-report measures ask for the same information
negative thought processes and to help therapists
that is already gathered in a clinical interview and
thus would not improve the clinician's ability to
estimate clients' suicide potential.
assess the person's suicide potential. However, as
stated earlier, there are a multitude of factors which
AIMS OF THE STUDY
interfere with clinical judgment when dealing with
suicidal individuals in particular (Maltsberger &
The purpose of the research reported here was to
Buie, 1989). making the use of clinical judgment
investigate the reliability and validity of the Firestone
Voice Scale for Self-Destructive Behavior. The hy-
alone potentially dangerous.
The Firestone Voice Scale for Self-Destructive
pothesis was that this scale would be able to discrimi-
nate those people with a past history of suicide from
Behavior takes a different approach from other scales
those without such a history and therefore relate
by asking the respondent to reveal the negative at-
closely to actual suicide potential, since a history of
tacks he or she experiences directed toward him or
attempts greatly increases the person's risk of dying
herself. The scale, by eliciting statements in the sec-
by suicide (Dorwart & Chartock. 1989: Jacobs.
ond person format, taps a partially unconscious proc-
ess. This particular format also helps the person to
1989). It was also hypothesized that the scale would
identify where a person falls on a continuum of
begin separating his/her negative point of view from
his/her own self-interest. It provides an opportunity
self-destructive potential. since the items on the scale
include a broad spectrum of self-destructive thought
for a person to develop insight into his/her self-criti-
patterns ranging from mild self-criticism to injunc-
cal thoughts and attitudes.
tions to commit suicide.
The Glendon Association
4
[phone]
5
Most of the scales developed to assess suicide
Form. a Therapist Information Form developed for
are empirically derived and gather information which
this study, and, if the patient had made a previous
has been found to be correlated with suicide. In
suicide attempt. the Intent Scale (Beck, Schuyler. &
contrast, the approach used here was based on asses-
Herman, 1974).
sing the voice process hypothesized to underlie sui-
cidal behavior. The FVSSDB is much more broadly
Instrumentation
based than other measures of suicidality. covering a
In addition to socioeconomic information, the
range of concepts related to suicide such as hopeless-
face sheet asked for mental health history on the
ness. detachment, and isolation as well as a variety of
subject's family of origin. Subjects were also asked
self-destructive behavior patterns such as eating dis-
to indicate whether they had engaged in self-harm or
orders and substance abuse. Thus, the FVSS DB may
suicide attempts, or if anyone in their immediate
enhance our ability to predict suicide as well as
families had demonstrated these behaviors (including
identify a full range of self-destructive behavior pat-
completed suicide for family members). This infor-
terns. In addition, it could Fovide clinicians with a
mation was used to help establish where the subject
comprehensive framework for understanding suicide
stood in relation to the "criterion variable" (whether
and self-destructive befu.vior.
or not they had a past history of suicide attempts).
Subjects were also asked to complete the Suicide
METHODS
Probability Scale (Cull & Gill, 1988); the Reasons for
Living Inventory (Linehan. Goodstein, Nielsen. &
Sub:acts
Chiles, 1983); the Beck Hopelessness Scale (Beck &
There were a total of 507 respondents. The sub-
Steer. 1988); a 2-question subset of the Survey on
jects were geographically diverse, living in areas
Self-Harm (Favazza & Eppright. 1986); the Eating
throughout the United States and Western Canada.
Disorders Inventory (Garner & Olmsted, 1984); the
Respondents were drawn from a variety of mental
Inventory of Feelings. Problems & Family Experi-
health settings and were all currently in psycho-
ence (Cook. 1986) (which actually consisted of 3
therapy. Sites included a center for recovering fami-
tests, the Internalized Shame Scale, the Problem His-
lies, several drug treatment programs, local mental
tory Test. and the Family of Origin Scale); the Moni-
health clinics, and outpatient psychotherapy prac-
toring the Future Substance Use Battery (Bachman
& Johnston, 1978); an 11-item Socially Desirable
tices. Respondents ranged in age from 16 to 73, with
an average age of 38. Of the participants, 169 were
Response Set Measure (Hays, Hayashi. & Stewart.
1989); and the CES-D Depression Scale (Rad loff,
male (33%) and 338 were female (67%). The subjects
were predominantly white (89%) even though a con-
1977).
certed effort was made to include minority subjects.
The socioeconomic status of these subjects varied
Procedures
greatly with 51% earning under S30.(k0; 26% earn-
The patient participants were administered the
ing in the range of 530-50.000 and 197 in the above
battery of tests in a private setting with the main
S50.(XX) range. Respondents were asked to partici-
researcher or a research assistant present. These re-
pate voluntarily after permission had been obtained
searchers were present to answer questions and to
from their therapist. It was found that the sample
communicate with subjects who might become dis-
chosen included 93 persons who had made suicide
turbed by feelings aroused during the testing. As a
attempts ahd 414 who had not.
precaution, if subjects appeared to be upset. the re-
searcher notified the therapist to schedule an extra
Design
session for the subject shortly after testing. Following
All 507 subjects were administered a testing
testing, Beck Hopelessness Scale and Suicide Proba-
bility Scale were scored within 24 hours, and the
packet consisting of a Subject Consent Form, a Face
Sheet of socio-economic information, and 10 instru-
therapist was informed if any of the scores were in a
ments, including the FVSS DB. in random order. The
range of concern.
therapists of these 507 subjects each filled out a
therapist packet consisting of a Therapist Consew
The Glendon Association
[phone]
5
6
tions ranging from 0.38 to 0.48) but seeming to
Results
represent a separate concept.
Internal consistency reliability was evaluated by
Factor 1 included Level 1 (everyday self-crit-
estimating Cronbach's (1951) alpha coefficient. This
icisms), Level 2 (self-denial), Level 3 (cynicism), and
method is used in multi-item scales to indicate the
Level 4 (isolation). All of these levels am directly
degree of convergence between items hypothesized
representative of thoughts contributing to low self-
to represent the same construct or level. The results
esteem and inwardness. In addition, they represent
document a high level of internal consistency. The
commonly occurring thought patterns that most peo-
coefficient ranged from 0.78 for Level 2. (self-denial)
ple can relate to, to varying degrees, as indicated by
to 0.97 for Level 1 I (injunctions to commit suicide).
the high level of endorsement they received from the
The estimated internal consistency of the total scale
majority of subjects. Factor I was labeled Low Self-
was very high (alpha = 0.98).
Esteem.
A multi-trait, multi-item (MTMI) correlation
Factor 2 consists of Level 5 (vicious self-abusive
matrix was computc,1 using the Multitrait Analysis
thoughts) and Level 7 (thoughts engendering hope-
Program (Hays & Hayashi, 1990). A majority of the
lessness). The statements from both of these levels
items in each level satisfied to 0.40 convergence
represent the extreme of a self-hating point of view.
recommended standard.
Thus. Factor 2 was named Extreme Self-Hatred.
In order to examine the hierarchical theory of
Factor 3 included Level 8 (giving up on oneself).
self-destructiveness represented by the FVSSDB,
Level 9 (injunctions to sel f-harm). Level 10 (thoughts
Guttman Scalogram Analysis was done using the
planning suicide) and Level 11 (injunctions to sui-
microcomputer program scale (Gilpin & Hays.
cide.) This cluster of levels represents the actual
1990). Level 9 (injunctions to self-harm) were ex-
destruction of self, both psychologically and physi-
cluded for this analysis since these items were en-
cally. Thus, Factor 3 was named Destruction of the
dorsed with the least frequency of all levels. This
Self.
reflects on the low base rate of self-mutilation beha-
Three forms of validity were examined for the
viors and indicates that they are not a necessary
FVSS DB, construct. criterion, and incremental valid-
precursor of suicide. The first three subscales were
ity. Two fundamental aspects of construct validity
collapsed because they were indicated by the MTMI
were explored, convergent and discriminate validity.
matrix to represent a single construct: they all repre-
In order to identify the cut score for the total
sent forms of common, everyday voices. The seal-
FVSSDB scale score that maximized its sensitivity
ability of responses was determined comparing ob-
and specificity to the probability of suicide, cross
served patterns of data with the patterns predicted for
tabulation tables were developed between the various
a Guttman Scale (Figure 1). The level of prevalence
suicide scales. total score, and the "criterion variable"
observed varied somewhat from predictions, with
(suicide attempts as reported by both therapist and
Level 4 (isolation) receiving a higher pret &nee than
client.) The results revealed that 44% of the attemp-
any other level. However, the coefficiem of repro-
ters scored in the top 20% of scores on the FVSSDB.
ducibility (CR) was 0.91. with a v.
90 or
In addition, 80% of the non-attempters group scored
higher considered acceptable. The coe
cient of
below this top 20% of scores. These findings sup-
scalability (CS) for a slightly modified ord;..ring of
ported the criterion validity of the FVSS DB.
the levels (i.e.. by difficulty) was 0.66 with a CS of
In order to optimize specificity and sensitivity, a
0.60 as a standard for acceptability.
This finding
cut score of 24.4 was selected. This score has a
indicates the levels are ordered along a single dimen-
specificity of 71% and a sensitivity of 61%. To verify
sion.
the cut score selected, the sample was randomly
Confirmatory factor analysis was performed us-
divided and new cross-tabulation tables of FVSSDB
ing a computer program (Bentler, 1989). The results
scores (by the criterion variable) were developed for
revealed three factors of increasing self-destructive-
each half. The results supported the cut score chosen
ness that provide an adequate underlying model for
based on the whole sample.
the observed data (Figure 2). Level 6 (addictions) was
separated out as related to the three factors (correla-
The Glendon Association
6
[phone]
7
o
h
t
n
h
t
n
t
a
i
o
g

w
o
l
s
u
l
P

t

e

n
s
n
o
i
,

m
n
d
o e
n
h
o
o
d
i
o
6
T
i
t
m
c
i
i
c
i
i

t 5
t
c
t
e
c
i
A
c
c
i
u
o
u
m
i
n

n
r
e
S
S
C
t
u
e
s
u
k

i
r
n
e
a
n
j
t
9
n
o
x
h
T
I
E
C
t
I
f
,
o
m
l
l

e
l
s
s
a
g
S
s
g
s
e
t
C
l
o
a
n
t

h
n
d
h
r

L
n
i
o
,
i
g
g
t
0
e
n
i

o
a
)
f
c
e
u
u

v
5
l
i
n
c
g
t
e
i
t
o
i
o
a
u
e
s
i
n
a
d
h
v
s
R
l
h
S
i
p
l
n
T
l
e
e
e
P
m
T
i
s
l
e

b
o
a
F
O
C

n
i

m
s
m
n
e
e
-
e
n
l
v
o
o
l t
i
i
i
I
e
v
s
r

f
f
W
p
i
h
s
l

t
e
l
c
e
U
a
c

A
s
r
f
w
a
n
g

e

g
I
a
d
e
o

n
s
r
e
n
r

d
P
O
e
n
r
i
h
o
c
v

o
g
t
i
v
i
o
i

G
W
n
a
"
V
i
F
e
l
i
h
w
t
e
l
s
e
l

a
h
s
a
g
l
p
n
t
n
e

o
m
o
n
i
n
v
t
e
I
s
i
o
o
t
s

P
a
e
e
r
m
s

f
s
R
f

c
e
e
o
o
f

l
n
R
e

l
s
l
C
s
e
o
s
e

e
g
e
p
c
"
v
m
n

i
o
i
o
d
L
O
H
V
n
a

"
y
l
-
n
l
i
h
B
s
o
c
e
s
I
t
n
n

S
i
n
r
g
w
n
D
o
e
a

n
o
f
o

i
u
r
t
o
d
i
3
c
i
i
g
l
"
S
t
f

t
e
i
0

c
n
m
a
c
d
n
n
i
I
g
1
d d
i
S

s
i
e
d
b
s
n
A
d
i
e
c
e
m
E
d
A
V

i
c
g
w
t

A
o
r
i
i
n
r
o
o
C
F
C
t
i
V
f
e
b

e

e
h
s
l
s
a
n
s
t
n
c
e

o
-
o
r
f
c
i
p
l
i
g
e
t
i
o
o
a
s
S
o
n
V
4
s
e
f

l
i
u

4
s

o
r
a
e
c
u
s
1
h
P
v
a
c
o
i
i
A
c
s

i
s
c
,
5
u
y

e
i
d
V
b
0
y
s
g
n
A
9
P
a
a
l
a
.
r
0
e
5
n

v
2
=
a
A
g
6

y
a

f
.
o
l
n
,
r

e t
0
n
t
u
i
n
m
o

s
l

o
e
o
i
e
=
o
(
b
n
i
c
t
n
9

i
a
o
i
t
n
a
t
y
O
c
0
n
l
a
l
E
o
A
t
u
3

e
r
i
l
s
k

l
o
s
d
g
e
l
m
I
c
b

e
s
m
a
o
g
c
o
a
e
t
i
M
r
i
t
i
T
A
o
l
p
s
l
a
V
a
r
e
c
o
R
S
c
d

n
S
f
f
s
o
o
E
m
g
e
n

s
m
t
t
c
i
n
f
l
n
i
r
n
l
o
c
-
e
i
e
e
f
o
e
i
e
a
l

d
n
e
S
e
e
V
i
n
i
y
t
S

c
c
s
m
n
5
C
,
e
7

E
m
i
i
h
g
y
f
i
6
f

0
L
-
d
f
f
g
n
s
a
l
2
e
t
e
l
5
n

u
E
e
i
t
d
c
5
o
o
a
=
o
S

u
i
s
y

.
t
r
C
C
N

l
i
2
h
e
a
w
r
r
G
C
T
c
i
e
o
n
i
o
v
L
e

V
D
E
0 0
0
0
0 0 0
.
5 0
5
0 5 0
1
1 1
3 2 2

e
n
.
r
5
a
o
c
T
u
g
i
F
8
e
n
d
o
i
c
i
t
1
i
6
c
u
4
n
8
S
,
1
u
/

j
o
n
t
I
5
1
9
3
6
4
i
i
f
e
l
e
s
d
n
,
S
i
c
3
c
a
t

e
i
8
l
e
u
P
S
h
t
'

7
f
o

8
3

n
F
1
o
4
m
i
-
t
f
3
c
l
r
s
e
a
7
e
u
S
H
r
t
s
e
D
f
l
p
e
U
s
6

e
0
a
g
7
n
7
e
3
n
3
O
9
9
i
v

i
n
G
5
o
9
3
6
4
2
e
,
s
t
o
a
s
s
e
H
7
j
s
4
l
e
e
-
e
9
n
f
p
n
l
o
e
o
i
H
t
S
2
c
0
F

i
5
d
e
4
d
m
.
A
e
e
s
r
s
v
t
u
e
x
i
o
s
c
7
s
E
u
i
i
8
e
c
o
b
i
V
V
A
_
'
B
-
-
D
0
,
1
1
S
4
8

8
S
n
4
o
V
i
F
t
a

e
l
o
m
h
s
t
i

e
f
o
e

t
s
s
1
i
E
s

m
y
-
l
F
l
f
a
3
l
s
e
e
n
i
2
c
S
A
2
i
n

4
w
r
y
o
C
o
t
c
L
[
a
F

y
l
r
a
-
o
f
i
9
r
n
l
2
t
e
7
a
7
e
e
S .
m
D
r
i
f
n
1
1
o
C
y
,

,
a
s
.
4
2
e
d
8

T
c
y
e
0
i
r
o
r
e
u
1
V
v
'
(
g
E
P
i
F
_
1
The FVSSDB total scale score was significantly
total score, r = 0.75 (p<.05).Therapists' reports of
correlated with the Suicide Probability Scale. = 0.77
clients' level of self hate were also significantly cor-
(p<.05); the Beck Hopelessness Scale. r =
0.60
related with Level 5.i = 0.40, (p.<05),
(p<.05); the Eating Disorders Inventory,
0.62
For Level 1 (self-critical voices), Level 2 (self-
=
(p<.05)t the CES-D Depression Scale. /
0.83
denial), Level 3 (cynical attitudes), and Level 4 (iso-
=
(p<.05); both parts of the Survey on Self-Hann. Self-
lation), the highest correlations were with Internal-
Hann, r = 0.20 (p.<05). Self-Harm types. i = 0.42
ized Shame Scale total scores/ = 0.70, 0.60,0.58, and
(p<,057): the Internalized Shame Scale,
= 0.74
0.60 respectively, (p<.05).
(p<,05); and the Problem History Scale total, = 0.48
The criterion validity for the FVSSDB was evaJ-
(p<.05).
uated by comparing FVSSDB scores with previous
The FVSSDB total scale score was also corre-
suicide attempts. The FVSSDB was found to have a
lated significantly with therapist overall evaluation
hieher correlation with the criterion variable (sub-
of me self-destructiveness of thc clients. r = 0.40,
jects' and therapists' reports of past suicide attempts)
= 0.31 (p<.05) than any of the other measures.
It was predicted that the separate subscales of the
including the Suicide Probability Scale. = 0.26, and
FVSSDB would correlate with specific measures of
the Beck Hopelessness Scale, = 0.18. Steiger ratios
the construct they purport to measure and that these
were calculated to estimate the significance of the
correlations would be higher than those with other
difference of these correlations. The FVSSDB corre-
levels measuring distinct or different constructs.
lation with the criterion variable is significantly
Level 10 (thoughts planning suicide) and Level 11
higher than all other measures except the SPS. Cor-
(injunctions to commit suicide) each correlated r =
relations for the FVSSDB Level 10 (thoughts plann-
(1.69 (p<.05) with the SPS total score. and r = 0.80
ing suicide) and Level 11 (injunctions to suicide) had
(p<.05) with SPS suicidal ideation subscale. The
significantly higher correlations with the criteria than
therapists' reports of clients' suicidal ideation corre-
SPS.
In order to determine whether or not the
lated significantly with Level 101= 0.40 (p<.05). and
with Level 1 1, r = 0.45 (p<.05).
FVSSDB could add significantly to our ability to
Level 9 (injunctions to self-harm) had the high-
determine suicide potential. logistic regression ana-
est correlation with the Survey on Self-Harm. The
lysis was conducted to explore this aspect of incre-
correlation with the Survy on Self-Harm types. was
mental validity. A logistic regression coefficient was
= 0.52 (p<.05), and with the Survey on Self-Harm
obtained using the variables SPS total score. BHS
Times r = .025 (p<.05). Therapists' reports of clients'
total score, age. income, gender, race, employment
self-harm correlated
0.15 (p<.05) with Level 9.
status. marital status as predictors. Subsequently, a
=
Level 7 (thoughts engendering hopelessness) and
logistic regression was run adding the FVSSDB total
Level 8 (giving up on oneself) correlated r = 0.50
score. The difference in resulting logic coefficients
(p<.(15) and I = 0.86 (p<.05). respectively, with the
was compared X' (1, N = 383) = 7.268, p<.05 and
revealed a significant
Hence, the
Beck Hopelessness Scale. In addition. they were both
difference.
significantly correlated with the Hopelessness subs-
FVSSDB total score adds significantly to our ability
cale of the SPS = 0.78 (p<.05).
to discriminate those persons who have made prior
Therapists' reports of clients' hopelessness cor-
suicide attempts and therefore by inference represent
related with Level 7 and Level 8, both i
0.33
a greater potential threat of actual suicide.
=
(p<.05). The highest correlation for Level 6 (addic-
tions) scores was with the addictions subscale of the
DISCUSSION
Problem History Test.
= 0.64 (p.05). In addition,
Level 6 correlated r = 0.40 (j2<.05) with the EDI.
The results of this study provide support for the
Therapists' reports of addictive behaviors also corre-
reliability an,' validity of the Firestone Voice Scale
lated with Level 6: therapists' reports of clients sub-
for Self-Destructive Behavior. Most importantly, the
stance use correlated r = 0.33 (p<.05) and of eating
criterion validity was demonstrated by the FVSSDB
disorders r = 0.17 (p<.05).
having a highly significant correlation with the sub-
Level 5 (vicious self- ...lusive thoughts) had its
ject's past suicide attempts. This correlation was sig-
highest correlation with the Internalized Shame Scale
nificantly higher than all other instruments except the
The Glendon Association
[phone]
9
12