Co gniti ve Therap y an d Research, Vol. 23, No . 2, 1999, pp .

119-129

Th e D evelopm ent and Validation of the Psycho logical

Vulnerability Scale
V au gh n G . Sin clair 1 an d K e nne th A . Wallsto n 2,3

The purpose of th is article is to in troduce the Psych olo gical Vuln erab ility Scale (PVS),
a six-item m easu re of a set of cogn itio ns that prom ote harm ful reactio ns to stress.
Malad ap ti ve cogn iti ve reactio ns to in terpersonal events can affect copin g beha vio r
an d psych o lo gical an d physical w ell-b eing. A m easu re of psych olo gical vuln erab ility
that re¯ects detrim ental cogn iti ve beliefs could be valu ab le in id entifyin g in divid uals
m ost in need of cogn iti ve-beha vio ral in terventio ns.
Three sam ples (ns 5 90, 138, an d 137) of in d ivid uals w ith rheum ato id arth ritis
pro vid e e vid ence for the reliab ility an d valid ity of the PVS, w hich has ad equate
in ternal consistency an d test±retest reliab ility. Con vergen t valid ity of the scale is
dem onstrated by positi ve correlatio ns w ith m easu res of perceived h elp lessness, nega-
ti ve affect, m alad ap ti ve pain copin g beha vio rs, an d disease acti vity, an d negati ve
correlatio ns w ith m easu res of positi ve affect, life satisfactio n, ad ap tive p ain -copin g
beha vio rs, perceived social support, an d personal copin g resources such as self-
ef®cacy. The sensiti vity of the PVS to cogn iti ve ±b eha vio ral in terventio n is also dem on-
strated .
KE Y WOR D S: cognitive processe s; vulne rability; stress re sponse s; adjustme nt; instrument deve lopme nt.

Individuals establish habitual cognitive patte rns in re sponse to the e xpe ctations
and treatme nt the y re ceive from othe r people . The se patte rns re ¯ect their be lie f
in the ir value re lative to othe rs. Cognitive reactions to inte rpe rsonal e ve nts can
affe ct individuals’ psychologic al and physical well-be ing. This is particularly the
case for pe ople with chronic illne sse s in which maladaptive cognitions can in¯ue nce
depre sse d mood and, pe rhaps, he alth outcom es (Smith, Peck, Milano, & Ward,
1988; Smith, Christe nsen, Pe ck, & Ward, 1994) . A short, e asy-to-admi niste r measure
of psychologic al vulne rability that re ¯e cts de trimental cognitive be liefs could be
valuable in ide ntifying individuals most in nee d of cognitive ±be havioral inte rve n-
tions. The ability to pre dict which individuals could be ne®t most from cognitive ±
1
Vanderbilt Unive rsity School of Nursing, Nashville , Te nnesse e.
2
Vanderbilt Unive rsity School of Nursing, Nashville , Te nnesse e.
3
Re quests for reprints should be addresse d to Ke n Wallston, Ph.D., School of Nursing, V ande rbilt
University, Nashville , Te nnesse e 37240.

119
0147-5916/99/0400-0119 $16.00/0 Ó 1999 Plenum Publishing Corporation
120 Sinclair an d Wallston

behavioral program s would greatly e nhance the e ffective and e f®cie nt utilization
of these re source s. Such a me asure could also be use d as an indicator of the
effe ctive ne ss of the rape utic inte rventions de signe d to modify maladaptive cogni-
tive be lie fs.
The Psychologica l V ulne rability Scale (PV S) described in this article was de -
signe d to ide ntify individuals with cognitive patte rns that re nde r the m more suscepti-
ble to stress. We de®ne the type of psychologica l vulne rability assessed by this scale
as a patte rn of cognitive belie fs re ¯e cting a depe nde nce on achie ve ment or e xte rnal
source s of af®rmation for one ’s se nse of se lf-worth. Such a de pe nde nce on concre te
achie veme nts and othe r pe ople for self-af®rmation, as oppose d to a belie f in the
worth of one ’s inne r qualitie s and characte r, renders one ’s se nse of se lf-worth
vulne rable to the whimsical tre atme nt of othe rs or the vicissitude s of life that can
marke dly inte rfere with one ’s ability to achie ve goals.
The psychologic al vulne rability measure d by this instrum e nt scree ns for cogni-
tive vulne rability relate d to pe rceptions of de pe nde ncy, pe rfe ctionism , ne gative
attributions, and the ne e d for e xte rnal source s of approval. Psychosocial researchers
have inve stigate d various links be twe en cognitive vulne rability and psychologica l
breakdown, particularly de pre ssion. A cluste r of pe rsonality variable s that have
bee n inve stigate d as speci®c vulne rability factors to depression include depende ncy,
self-blame , pe rfe ctionism , and dysfunction al attributions.
Se lf-orie nted pe rfe ctionism and conce rns about autonom ous achie ve ment have
bee n the focus of many studie s re garding vulne rability to de pre ssion. For the individ-
ual with e xce ssive conce rns about achie ve ment, perceive d failure relate d to impor-
tant achie vements can be an ove rwhelming blow that leads to de pre ssion (Robins,
1995) . In one longitudinal study of 103 patie nts with a history of de pre ssion, re se arch-
ers found that se lf-orie nte d pe rfe ctionism inte racte d with achie vement stre ss to
predict the incide nce of de pre ssion ( Hewitt, Flett, & E dige r, 1996) . A nothe r dime n-
sion of perfe ctionism, socially pre scribed perfe ctionism, has be en shown to inte ract
with inte rpe rsonal stre ss and pre dict de pre ssion (Hewitt & Flett, 1993) .
De pendency and inte rpersonal se nsitivity have also re ceive d atte ntion from
researchers (B oyce & Mason, 1996) . Schill and Sharp (1995) found that individuals
who score d highe r on sociotropy, or social de pe nde nce , also scored highe r on their
measure of ``se lf-de fe ating’’ pe rsonality traits and had more ne gative cognitions
about themse lve s, the world, and the future . The se inve stigators speculate that
threats to relationships may be an important source of de pre ssion for this subse t
of subje cts. Franche and Dobson (1992) found that the ir subje cts with e ithe r a
curre nt or past history of depression scored signi®cantly highe r on measure s of
inte rpe rsonal de pendency and self-criticism than did norm al controls. These re -
searchers sugge st these two factors are fairly stable vulne rability factors for de pre s-
sion be cause the y remain active be yond the e nd of the de pre ssive episode .
O the r inve stigators focuse d on the role of dysfunction al attitude s (i.e., rigid,
global, ne gative thinking and autom atic ne gative thoughts about the se lf, the world,
and the future ) in promoting de pre ssion (Charlton & Powe r, 1995; Ruble , Gre ulich,
Pome rantz, & Gochbe rg, 1993) . In one study comparing de pre sse d individuals
with the ir recove re d depre sse d relative s, the re se arche rs found that score s on a
dysfunctiona l attitude measure returne d to normal in re covered de pre sse d individu-
Psycho logical V uln erability Scale 121

als, where as de pe nde ncy score s re maine d ele vate d (Powe r, Duggan, Lee , & Murray,
1995) . Dysfunctional attitude s may, therefore , be activate d by stressful life e ve nts
and may be le ss stable than social de pende ncy patte rns. Miranda (1992) found
disproportio nate ly high le vels of dysfunctiona l thinking following stre ssful e vents in
adults with a history of de pre ssion, but not in adults who had neve r be e n de pre sse d.
Riskind e t al. (1987) hypothe size d that attributiona l style has an impact on de pres-
sion through the mediating e ffe ct of pe ssimistic expe ctations about future outcom e s.
In the ir study, the y found that the inte raction of attributional style and e xpe ctations
predicte d de pre ssion score s 6 weeks late r, supporting the proposition that vulne ra-
bility factors combine inte ractive ly and qualify the e ffe cts of attributional style .
De pre ssoge nic attributional style s te nd to promote a sense of helple ssness
and hope le ssne ss, and some re searche rs be lie ve that attributional style and othe r
etiological factors cause de pressive symptoms through the ope ration of hope lessness
(A bramson, Me talsky, & A lloy, 1989; Metalsky & Joine r, 1992) . A t le ast one study
demonstrate d that individuals with de pre ssoge nic attributions had more incide nce s
of major de pre ssion, and that the se incide nce s were more seve re ove r a 2-ye ar
period than were incide nces of de pre ssion for subje cts with a healthy and more
positive attributiona l style (A lloy, Lipman, & A bramson, 1992) . Ruble e t al. (1993)
speculate that gender socialization may make attributional issue s more salie nt for
wome n by promoting lower expe ctations of succe ss, more maladaptive causal attri-
butions for succe ss or failure , and ne gative evaluative reactions to failure .
The purpose of this article is to prese nt reliability and validity information
regarding PV S from thre e sample s of individuals with rhe umatoid arthritis. Rhe uma-
toid arthritis (RA ) is an autoimm une condition that is ofte n accompanie d by high
le vels of pain and functional disability (Pincus, 1996) . A lthough the majority of
individuals with RA make a satisfactory adjustm ent to the condition, the re is great
individual variability in how well individuals with RA adjust, e ven controlling for the
amount of pain and disability present (Newman & Reve nson, 1993) . The incide nce of
major de pression in individuals with RA is some what highe r than is the case for
healthy individuals (DeV e llis, 1993) , but there is gre at variability in depressive
symptoms within sample s of individuals with RA . Thus, one indicator of the con-
struct validity of the PV S lie s in its re lationship to me asure s of depressed affe ct
and othe r indicators of psychologica l well-be ing. Psychologic al vulne rability could
also be conce ptualize d as an indicator of a de®cit in pe rsonal coping re sources.
Individuals high in psychologica l vulne rability should have less re source s available
to cope with stre ssors such as the pain associate d with RA . Thus, signi®cant corre la-
tions with measure s of coping re sources such as a sense of optimism , pe rceive d
compete nce , le arne d he lple ssne ss, se lf-ef®cacy, and social support as well as corre la-
tions with me asure s of pain-coping be havior should provide furthe r e vide nce of
the construct validity of the PV S.

SCA LE D E V E LOPME NT A ND R E LIA B ILITY

The PV S was initially de velope d to dete rmine whethe r individuals with RA

adopt more positive be lie f patte rns about their social abilitie s afte r a cognitive ±
122 Sinclair an d Wallston

behavioral inte rvention. ( The de tails of that inte rve ntion have be en de scribe d else -
where : se e Sinclair, Wallston, Dwye r, B lackburn, & Fuchs, 1998) . O riginally, the
®rst author ( V S) wrote 38 ite ms tapping these and similar belie fs. Nine ty wome n
with RA who were enrolle d in the inte rve ntion program (he re afte r re fe rre d to as
sample 1) ®lle d out the 38 items as part of a base line assessment packe t. The y also
comple ted these ite ms thre e more times ove r the course of the study. In addition,
the se cond author (KW) administe re d the same se t of items to 77 men and 200
wome n with RA who were participatin g in two longitudinal pane l studie s of pe rsons
with RA that he was conducting (he re afte r refe rre d to as samples 2 and 3).
Se parate e xploratory principal compone nts factor analyse s with orthogonal
rotation were pe rforme d on data from sample 1 (n 5 90) and sample 2 (n 5 138) .
Thre e factors with e ige nvalue s greate r than 1 e merge d on both analyse s. Two
decision rule s were use d to choose the ite ms for the scale that was subse que ntly
name d the PV S: (1) the ite m had to load $ .30 on factor 1 for bo th sample s; and
(2) the ite m could not have an appre ciable factor loading on e ithe r of the othe r
two factors.

Ite m A nalyse s
The seve n items that met the crite ria mentione d pre viously were then subje cted
to se parate item analyse s for the purpose s of e stablishing the inte rnal consiste ncy
of the scale and to se e if all se ven ite ms made a signi®cant contribution to Cronbach’s
alpha. O ne ite m, ``My pain fre que ntly e nable s me to e scape obligations I’d rathe r
not deal with,’’ was subse que ntly droppe d be cause the alpha re liability in sample
1 without that ite m (.86) was slightly highe r than the alpha including the ite m (.85) .
Dropping this ite m also made the PV S appropriate for individuals without chronic
pain, e nabling more wide spre ad use of the scale . (Se e Table I for a listing of the
ite ms that constitute the PV S.)
The alpha re liability of the six-ite m PV S was some what lowe r (.73) in sample
2 than in sample 1, pe rhaps re ¯e cting the gende r diffe re nce s in the two sample s.
{In fact, the alpha was slightly highe r (.75) for the 108 wome n in sample 2 than for
the total sample.} The inte rnal consiste ncy of the PV S was furthe r de te rmine d to
be .71 in sample 3 ( n 5 137) . O nce again, the alpha re liability for the 92 wome n
in sample 3 was slightly highe r ( .73) than for the total sample , perhaps indicating
that the PV S is more inte rnally consiste nt for women than for men. Thus, in thre e
inde pe nde nt sample s of persons with RA , the inte rnal consiste ncy re liability of the

Tab le I. Ite ms in the Psychological V ulnerability Scale (PV S)

1. If I don’t achie ve my goals, I fee l like a failure as a person.
2. I fe e l entitled to bette r tre atment from others than I ge ne rally re ceive .
3. I am freque ntly aware of fee ling inferior to other pe ople.
4. I nee d approval from others to fe el good about myse lf.
5. I te nd to set my goals too high and become frustrated trying to re ach them.
6. I often fe el rese ntful when others take advantage of me.

Note : The possible response s varie d from 1Ð ``Doe s not de scribe me at all’’ to
5Ð ``Describes me ve ry we ll.’’ (Re sponse options 2, 3, and 4 we re not labeled.)
Psycho logical V uln erability Scale 123

PV S range d from .71 to .86. This le ve l of inte rnal consiste ncy is ade quate for
research purpose s in a scale consisting of as fe w as six ite ms.

Stab ility
The PV S was re administe re d to 88 of the 90 wome n participating in the inte r-
ve ntion study (sample 1) 1 we e k be fore the inte rve ntion began. This occurre d 5±6
we e ks afte r the y ®rst ®lle d out the same se t of ite ms. The te st±re te st correlation
of the PV S ove r this 5±6 week pe riod was .83. The PV S was also re administe re d
to 87 of the wome n at the end of the inte rvention, and those score s corre late d .79
with PV S score s administe re d 5± 6 we e ks e arlie r, just before the inte rve ntion be gan.
Finally, the te st±re te st corre lation of the PV S from the end of the inte rve ntion to
the 3-month follow-up was .81. Thus, the re is e vide nce that the PV S is both inte rnally
consiste nt and stable ove r time. The re fore, it is a reliable instrum ent for research
purpose s.

V A LID ITY

The concurre nt and construct validity of the PV S was initially e stablishe d by

corre lating PV S score s with me asure s of pe rsonal coping re sources, psychologica l
we ll-be ing, social support, and pain-coping be havior in the thre e sample s of pe rsons
with RA . The othe r measure s of personal coping resource s were arthritis help-
le ssne ss (Ste in, Wallston, & Nicassio, 1988) , Pe rceived Health Compe te nce (Smith,
Wallston, & Smith, 1995) , arthritis se lf-ef®cacy (Lorig, Chastain, Ung, Shoor, &
Holman, 1989) , and dispositiona l optimism ( Sche ie r & Carve r, 1985) . Psychologica l
we ll-be ing was asse sse d by the Positive A nd Negative A ffe ct Sche dule (PA NAS;
Watson, Clark, & Te lle gen, 1988) , the Satisfaction with Life Scale (Die ne r, E mmons,
Larse n, & Grif®n, 1985) ; and, for samples 2 and 3, the Cente r for E pide miological
Studie s±Depre ssion Scale (CE S-D; Radloff, 1977) . (The CE S-D was not adm inis-
te re d to subje cts in sample 1.) Pe rceive d social support was assessed by short
measure s of quality of emotional support, availability of instrume ntal support, and
size of the support network (see B rown, Wallston, & Nicassio, 1989) . In sample s
2 and 3, pain-coping behavior was asse sse d by the V ande rbilt Multidim ensional
Pain Coping Inve ntory (V MPCI; Smith, Wallston, Dwye r, & Dowdy, 1997) . A
shorte ne d version of the V MPCI was administe red to subje cts in sample 1.
For all thre e sample s, bivariate correlational analyse s re vealed a consiste nt
patte rn of theoretically pre dictable correlations be twe e n PV S score s and measure s
assessing pe rsonal coping re sources (Table II), psychologica l well-be ing (Table III) ,
social support (Table IV ), and pain coping be havior (Table V ). PV S score s did not
corre late with chronologic al age , le ngth of time since diagnosis, or e ducational leve l
in any of the three sample s.
A s a test of the se nsitivity of the PV S as a pote ntial outcome me asure for
cognitive ±be havioral inte rventions, paire d t-te sts were run on succe ssive administra-
tions of the measure to subje cts in the inte rve ntion study (sample 1). The re was
no signi®cant change in PV S score s ove r the 5±6 wee k pre inte rvention pe riod, but
124 Sinclair an d Wallston

Table II. Correlations with Me asures of Pe rsonal Coping Resource s

Sample 1 Sample 2 Sample 3

Arthritis helplessness 0.51 a 0.32 a 0.34 a
Perce ive d he alth compe tence 20.5 a 2 a
0.3 2 a
0.5
Se lf-ef®cacy (arthritis) 20.44 a 20.23 b 20.14
Se lf-ef®cacy (other symptoms) 20.33 a 20.13 20.27 a
Dispositional optimism 20.43 a 20.45 a 20.44 a
a
p , .01; b p , .05.

PV S score s de cline d signi®cantly ( t(86) 5 1.75; p , .05, one -taile d) as a function of

subje cts participating in the cognitive ±be havioral inte rve ntion. In the 3 months
following the inte rve ntion, PV S score s remaine d essentially unchange d. Thus, the re
is initial e vide nce that the PV S might be se nsitive e nough to de te ct change s brought
about by a cognitive ±be havioral inte rve ntion.
The PV S score s of subje cts in sample 1 at time 1 also contribute d signi®cantly
(B eta 5 .252; t 5 2.77; p , .01) to the pre diction of chan ge in negative affe ct from
time 1 to time 2. The highe r the perceive d vulne rability at the ®rst assessment
period, the more ne gative their mood be came ove r the 5±6 wee k pe riod be fore
the inte rve ntion be gan. A similar, but le ss signi®cant, ®nding occurre d when the
crite rion me asure was change in positi ve affect (B eta 5 2.183; t 5 21.78; p 5 .08) ;
those who were more psychologica lly vulne rable at time 1 de ve lope d le ss positive
affe ct ove r the 5± 6 wee ks prior to the inte rvention. Inte re stingly, the cognitive ±
behavioral inte rve ntion some how affe cted the nature of the re lationship be twee n
psychologica l vulne rability and change in affe ct; time 2 PV S scores did not signi®-
cantly pre dict change in eithe r positive or ne gative affe ct ove r the course of the inte r-
ve ntion.
Give n the relative ly high (.51 to .61) correlation betwe en the PV S and the
negative affe ct score on the PA NA S, it is possible that the amount of variance the y
share in common (approximat ely 30%) is re ¯e ctive of a re sponse -se t bias that has
bee n te rmed neuroticism (Costa & McCrae, 1987) or ne gative affe ctivity (Watson &
Clark, 1984) rathe r than e vide nce that the re is, inde e d, a relationship betwe e n
psychologica l vulne rability and psychologic al well-be ing. If this were the case, most
of the signi®cant corre lations displaye d in Table s II±V might also be re ¯e ctive of
this response se t bias. To te st this proposition, partial corre lations were run on data

Table III. Correlations with Me asures of Psychological Well-Being

Sample 1 Sample 2 Sample 3
a a
Ne gative affe ct 0.61 0.51 0.51 a
Positive affect 20.52 a 20.22 a 20.36 a
Life satisfaction 20.49 a 20.51 a 20.39 a
De pre ssive symptomato logy n/a a 0.52 b 0.46 a
a
p , .01; n/a 5 not available .
Psycho logical V uln erability Scale 125

Tab le IV. Corre lations with Measure s of Social Support

Sample 1 Sample 2 Sample 3

Q uality of e motional support 20.54 a 20.23 a 20.43 a
A vailability of instrume ntal support 20.28 a 20.28 a 20.24 a
Network size n/a a 20.31 a 20.17 a
a
p , .01; n/a 5 not available .

from sample 1 to se e if the 2.49 corre lation be twe en PV S and life satisfaction (LS)
would disappe ar if the PA NAS score s were partialle d out of the re lationship.
Partialling out the ne gative affe ct score droppe d the correlation to 2.31, but it
remaine d signi®cant ( p , .01) . In a similar manne r, partialling out the positive affect
score from the residual still left the re lationship be twee n PV S and LS signi®cant
(r 5 2.24; p , .03) . The se ®ndings were cross-valida te d with data from sample 2.
A lthough the sample 2 corre lation be twe en PV S and LS droppe d from 2.51 to 2.19
afte r controlling for both positive and ne gative affe ct score s, it was still signi®cant
( p , .03) . Furthe rmore , also using data from sample 2, the re lationship be twe en
PV S score s and de pressive symptom atology (as asse sse d by the CE S-D) remaine d
positive and signi®cant (r 5 .21; p , .02) afte r partialling out positive and negative
affe ct an d dispositiona l optim ism. Thus, it is unlike ly that the re lationship betwe e n
PV S scores and othe r measure s can be totally dismisse d as be ing due to re sponse
set biase s or similar me thodological artifacts.

D ISCU SSION

Whe re as positive cognitive belie fs that are re late d to se lf-e f®cacy, optim ism,
and pe rceived compe te nce se rve as he lpful resource s in mee ting challe nge s, the
belie f se t e xpre ssed in the PV S could pre dispose an individual to ne gative e xpe cta-
tions of thwarte d obje ctive s and failure . The items in the PV S (se e Table I) re ¯e ct
many of the pe rsonality variable s discusse d in the psychosocia l lite rature as vulne ra-
bility factors for de pre ssion. A n individual who e ndorse d the ®rst ite m, for instance ,
is re ve aling a te nde ncy toward global, rigid, ne gative attributions. O the r ite ms
re¯e ct te nde ncie s toward pe rfectionism (e .g., unre alistic goal se tting) , critical se lf-
evaluation, and de pendency on othe rs for approval. Ite ms 2 and 6 depict a pe rson
who re pe ate dly fee ls disappointe d ove r unme t e xpe ctations re garding othe rs’ re -

Table V . Correlations with Me asures of Pain-Coping B e havior

Sample 1 Sample 2 Sample 3
Passive coping 0.48 a 0.41 a 0.35 a
Se lf-blame 0.20 0.33 a 0.36 a
Isolation 0.33 a 0.37 a 0.31 a
Catastrophizing 0.58 a 0.41 a 0.37 a
Reappraisal 20.28 a 20.06 20.32 a
a
p , .01.
126 Sinclair an d Wallston

sponse s to him/he r. The passive , re se ntful tone of the se items reve als frustration
and helple ssness regarding social re lationships. A lthough one might que stion the
pote ntial ove rlap of these items and scale s measuring rese ntme nt or hostility, the
analysis strate gy we use d in which we controlle d for ne gative affe ctivity (using an
instrum ent that contains ite ms such as ``hostile ’’ and ``irritable ’’) partially re inforce s
the argum e nt that the PV S is not just a me asure of hostility.
The signi®cant corre lations of the PV S with the maladaptive coping be haviors
of passivity, se lf-blam e, isolation, and catastrophiz ing are note worthy. The PV S
re¯e cts a patte rn of be lie fs that would contribute to the use of the se maladaptive
pain coping be haviors. O the r rese arch done with the subje cts in sample 2 illustrate s
the poor outcom e s for individuals who engage in the se maladaptive pain-coping
behaviors (Smith e t al., 1997) . If individuals de pend on othe rs’ approval for the ir
sense of se lf-worth and re pe ate dly fail to re ceive the af®rmation the y long for from
othe rs, the y could de ve lop ne gative expe ctations with re spe ct to social re lationships
that, in turn, are aligne d with a patte rn of le arne d helple ssne ss. This heighte ne d
negativity and sense of he lple ssne ss could the n re sult in incre ase d use of the mal-
adaptive coping be haviors of passivity, se lf-blame , isolation, and catastrophiz ing.
The se , in turn, can le ad to diminishe d physical and psychologica l well-be ing. O n a
dee pe r le vel, the be lie fs re pre se nted in the PV S may re ¯e ct a pre occupation with
fe ar of failure and disapproval that may represe nt a learne d re sponse to powe rful
negative e xpe rie nce s.
A ll of the se ne gative thoughts and e xpe ctations, couple d with ne gative expe ri-
ences, could contribute to a sense of hope lessness. Me talsky and Joine r (1992)
de®ne hope le ssne ss as ``an e xpe ctation that highly de sire d outcome s will not occur
or that highly ave rsive outcome s will occur and that nothing is going to change this
situation for the be tte r’’ (p. 667) . With respe ct to obtaining af®rmation of se lf-
worth via re lationships or achie ve ments, individuals who e ndorse the be lie fs in
the PV S re¯e ct high le vels of hope le ssne ss. It is not surprising, give n the role of
hope le ssne ss as a causal factor of de pre ssion (Me talsky & Joine r, 1992) , that initial
score s on the PV S predicte d change in negative affe ctivity on the PA NA S, our
proxy measure of de pre ssion, ove r a 6-we ek pe riod of time in sample 1.
The construct validity of the PV S was e stablishe d with pre dictable corre lational
patte rns with a varie ty of preinte rve ntion me asure s in three sample s of individuals
with RA . Pre dictable positive corre lations were e stablishe d be twee n the PV S, arthri-
tis he lple ssne ss, ne gative affectivity, and maladaptive pain-coping be haviors. The
PV S was also negative ly correlate d with all measure s of positive pe rsonal coping
resource s (e .g., arthritis self-e f®cacy, pe rceive d he alth compe te nce), adaptive pain-
coping be haviors, psychologica l well-be ing, and social support. The fact that the
relationship betwe en the PV S and life satisfaction re maine d signi®cant e ve n after
the e ffect of negative affe ctivity was re move d demonstrate d that this relationship
was ``re al,’’ and not signi®cant mere ly because the PV S was inadve rte ntly measuring
neuroticism or some similar re sponse set.
The cognitive ±behavioral inte rve ntion administe re d to subje cts in sample 1 was
designe d to unrave l dysfunctiona l attributions by addre ssing ne gative expe ctations,
unre alistic e xpe ctations (pe rfe ctionism ), common cognitive distortions, and negative
attributional patte rns. The ®nding that PV S score s de crease d signi®cantly ove r the
Psycho logical V uln erability Scale 127

course of the inte rve ntion with subje cts in sample 1 was af®rming of the scale’s se nsi-
tivity.
The inte rve ntion could have disrupte d the relationship be twe e n psychologica l
vulne rability and change in affe ct be cause individuals responde d diffe re ntially to
the inte rvention base d on a varie ty of factors. The impact of the program appare ntly
contribute d to alte re d mood state s post-inte rve ntion, and the se change s were not
relate d to base line scores on the PV S. Instead, the patte rn of improve ment in our
subje cts’ mood was relate d to othe r factors that remain a focus of analytical inve sti-
gation.
O ne pote ntial limitation of the factor analyse s we carried out is that the subje cts-
to-ite ms ratios in both sample s were low (2.37:1 in sample 1 and 3.63:1 in sample
2). If the two samples had bee n combine d, the ratio would have be e n be tter (6:1) ,
but we would not have be e n able to re plicate the analyse s across the two samples
as we did here . O n balance , it is pe rhaps bette r to cross-valida te factor analyse s in
two smalle r sample s than to do only a single e xploratory analysis in a large r sample .
Howe ve r the PV S was de velope d, furthe r analyse s have de monstrate d its utility.
In summary, the PV S is inte rnally consiste nt and stable . There is be ginning
evide nce for the construct and criterion validity of the PV S, although furthe r re -
search with othe r populations is ne cessary. The PV S may prove to be a use ful
indicator of spe ci®c ne gative cognitive be lie fs re garding sources of self-worth, and
may have implications for matching individuals with programs that can addre ss
the ir harm ful be lie f systems. There is no re ason to re strict the use of the PV S to
persons with RA or othe r chronic and painful conditions, although the validity of
its use in othe r populations is ye t to be de monstrate d. If the PV S stands up to
cross-validat ion in othe r sample s, it has the adde d virtue of be ing short and e as-
ily administe re d.

A CK NOWLE D G ME NTS

The re se arch reporte d in this pape r was supporte d by two re se arch grants to
the se cond author (KW): a Clinical Scie nce A ward from the A rthritis Foundation
(sample 1) and grant #R01 NR01007 from the National Institute s of Health, National
Institute of Nursing Rese arch (sample s 2 and 3).

R E FE R E NCE S

Abramson, L. Y., Metalsky, G. I., & A lloy, L. B . (1989) . Hope lessness depre ssion: A theory-based
subtype of depre ssion. Psycho logical Review , 96, 358-372.
Alloy, L. B ., Lipman, A . J., & A bramson, L. Y. (1992) . Attributional style as a vulnerability factor for
de pre ssion: V alidation for past history of mood disorders. Cogniti ve Therapy an d Research , 16,
391-407.
Boyce , P., & Mason, C. (1996) . An overvie w of depression-prone pe rsonality traits and the role of
interpersonal sensitivity. A ustralian and New Zealand Journ al of Psychiatry, 30, 90-103.
Brown, G. K., Wallston, K. A ., & Nicassio, P. N. (1989) . Social support and de pre ssion in rhe umatoid
arthritis: A one -year prospe ctive study. Journ al of A p plied So cial Psycho logy, 19, 1164-1181.
Champion, L. A ., & Power, M. J. (1995). Social and cognitive approache s to depression: Towards a
ne w synthe sis. B ritish Jo urnal o f Clinical Psych olo gy, 34, 485-503.
128 Sinclair an d Wallston

Charlton, P. F., & Powe r, M. J. (1995) . The asse ssme nt of dysfunctional attitudes and their role in the
onset, pe rsistence and re currence of clinical de pression. E urop ean Jo urn al o f Person ality, 9, 379-400.
Costa, P. T. & McCrae, R. R. (1987) . Ne uroticism, somatic complaints, and disease: Is the bark worse
than the bite? Journ al of Perso nality, 55, 299-316.
Costello, C. G. (1994) . Two dimensional views of psychopathology. B eha viou r Research an d Th erapy,
32, 391-402.
De V ellis, B . M. (1993) . De pre ssion in rheumatological diseases. B alliere’s Clinical Rh eum ato logy, 7,
241-257.
Die ne r, E., E mmons, R. A ., Larson, R. J., & Grif®n, S. (1985). The satisfaction with life scale . Jo urn al
of Perso nality A ssessm ent, 49, 71-75.
Franche, R. L., & Dobson, K. S. (1992) . Self-criticism and interpersonal de pe ndency as vulnerability
factors to depression. Cogniti ve Therapy an d Research, 16, 419-435.
He witt, P. L., & Fle tt, G. L. (1993) . Dime nsions of pe rfectionism, daily stress, and de pre ssion: A te st
of the spe ci®c vulnerability hypothe sis. Jou rnal of Ab no rm al Psycholo gy, 102, 58-65.
He witt, P. L., Flett, G. L., & E dige r, E . (1996). Pe rfe ctionism and de pre ssion: Longitudinal assessme nt
of a speci®c vulne rability hypothesis. Jo urn al of A bn orm al Psych olo gy, 105, 276-280.
Joine r, T. E . (1995) . The price of soliciting and rece iving negative fe e dback: Se lf-veri®cation theory as
a vulne rability to de pre ssion theory. Jo urnal o f A bn orm al Psycho logy, 104, 364-372.
Lorig, K., Chastain, R., Ung, E ., Shoor, S., & Holman, H. (1989). De ve lopment and evaluation of a
scale to me asure pe rce ived self-ef®cacy in pe ople with arthritis. A rth ritis & Rh eu m atism , 32, 37-44.
Metalsky, G. I., & Joine r, T. E . (1992) . V ulnerability to de pre ssive symptomatology: A prospe ctive te st
of the diathesis-stress and causal mediation components of the hope lessne ss theory of depression.
Journ al of Perso nality an Social Psych olo gy, 63, 667-675.
Miranda, J. (1992) . Dysfunctional thinking is activated by stre ssful life e vents. Cogniti ve Therap y an d
Research , 16, 473-483.
Moore, R. G., & B lackburn, I. M. (1996) . The stability of sociotropy and autonomy in de pre ssed patients
unde rgoing tre atment. Co gn itive Therap y an d Research, 20, 69-80.
Ne wman, S. P., & Revenson, T. A . (1993) . Coping with rhe umatoid arthritis. B alliere’s Clinical Rh eu mato l-
ogy, 7, 259-280.
Olive r, J. M., Ross, M. J., Maie r, M., & Wiener, R. L. (1991) . Cognitive vulne rability to de pression as
reve ale d by vulne rability markers. Jo urn al of Social B eha vio r an d Perso nality, 6, 585-596.
Persons, J. B ., & Miranda, J. (1992) . Cognitive theories of vulne rability to depression: Re conciling
ne gative evide nce . Co gn itive Therap y and Research, 16, 485-502.
Pincus, T. (1996) . Rhe umatoid arthritis. In S. T. Wegene r, B . L. B e lza, & E . P. Gall (E ds.), Clinical
Care in th e Rheum atic Diseases. A tlanta, GA : A me rican College of Rheumatology.
Powe r, M. J., Duggan, C. F., Lee , A . S., & Murray, R. M. (1995) . Dysfunctional attitudes in de -
pre ssed and recove red de pre ssed patients and their ®rst-degre e relatives. Psycholo gical Medicin e,
25, 87-93.
Radloff, L. S. (1977) . The CES-D Scale: A self-report depression scale for rese arch in the ge ne ral
population. A pp lied Psycho logical Measu rem ent, 1, 385-401.
Riskind, J. H., Rholes, W. S., B rannon, A . M., & B urdick, C. A . (1987) . A ttributions and expe ctations:
A con¯uence of vulne rabilities in mild de pre ssion in a college student population. Jo urnal of
Personality and So cial Psych olo gy, 53, 349-354.
Robins, C. J. (1995) . Pe rsonality±e ve nt interaction models of de pre ssion. E urop ean Jo urnal of Person al-
ity, 9, 367-376.
Ruble, D. N., Greulich, F., Pome rantz, E . M., & Gochbe rg, B . (1993) . The role of ge nder-re late d
proce sses in the de ve lopme nt of se x differences in self-evaluation and de pre ssion. Jo urnal of A ffecti ve
Disorders, 29, 97-128.
Sche ier, M. F. & Carve r, C. S. (1985) . Optimism, coping, and he alth: A ssessmen t and implications of
ge ne ralized outcome e xpe ctancie s. Health Psycholo gy, 4, 219-247.
Schill, T., & Sharp, M. (1995) . Self-defeating personality and depression: A closer look. Psycholo gical
Reports, 76, 1167-1170.
Sinclair, V . G., W allston, K. A ., Dwye r, K. A ., B lackburn, D. S., & Fuchs, H. (1998). E ffects of a
cognitive -behavioral intervention for wome n with rhe umatoid arthritis. Research in Nursin g &
Health , 21, 315± 326.
Smith, C. A., Wallston, K. A ., Dwye r, K. A ., & Dowdy, S. (1997) . B e yond good and bad coping: A
multidimensional examinati on of coping with pain in persons with rhe umatoid arthritis. An nals o f
B eha vioral Med icine, 19, 11-21.
Smith, M. S., W allston, K. A ., & Smith, C. A . (1995) . The developme nt and validation of the Pe rce ived
He alth Compe tence Scale . Health E du cation Research : Th eo ry & Practice, 10, 51-64.
Psycho logical V uln erability Scale 129

Smith, T. W., Pe ck, J. R., Milano, R., & Ward, J. (1988) . Cognitive distortion in rhe umatoid arthritis.
Journ al of Co nsu lting and Clinical Psych olo gy, 56, 412-416.
Smith, T. W., Christense n, A . J., Pe ck, J. R., & Ward, J. R. (1994) . Cognitive distortion, helplessne ss,
and de pre ssed mood in rheumatoid arthritis: A four-ye ar longitudinal analysis. Health Psych ology,
13, 213-217.
Stein, M. J., Wallston, K. A., & Nicassio, P. M. (1988) . Factor structure of the A rthritis He lplessne ss
Index. Jou rnal o f Rheum atolo gy, 15, 427-432.
Watson, D. & Clark, L. A. (1984) . Ne gative affe ctivity: The disposition to expe rie nce ave rsive emotional
state s. Psycho logical B ulletin, 96, 465-490.
Watson, D., Clark, L. A ., & Tellege n, A . (1988). De velopme nt and validation of brief measures of
positive and ne gative affect: The PA NA S scales. Jou rnal of Person ality and Social Ps vch olo gy, 54,
1063-1070.