This article was downloaded by: [Erciyes University]

On: 19 December 2014, At: 21:24

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office:
Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Aging & Mental Health

Publication details, including instructions for authors and subscription
information:
http://www.tandfonline.com/loi/camh20

Measurement and correlates of family

caregiver self-efficacy for managing
dementia
a b c
R. H. Fortinsky , K. Kercher & C. J. Burant
a
Division of Geriatrics and Center on Aging , University of Connecticut
Health Center , Farmington, CT, USA
b
Department of Sociology , Case Western Reserve University , Cleveland,
OH, USA
c
Department of Bioethics , Case Western Reserve University , Cleveland,
OH, USA
Published online: 09 Jun 2010.

To cite this article: R. H. Fortinsky , K. Kercher & C. J. Burant (2002) Measurement and correlates
of family caregiver self-efficacy for managing dementia, Aging & Mental Health, 6:2, 153-160, DOI:
10.1080/13607860220126763

To link to this article: http://dx.doi.org/10.1080/13607860220126763

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”)
contained in the publications on our platform. However, Taylor & Francis, our agents, and our
licensors make no representations or warranties whatsoever as to the accuracy, completeness, or
suitability for any purpose of the Content. Any opinions and views expressed in this publication
are the opinions and views of the authors, and are not the views of or endorsed by Taylor &
Francis. The accuracy of the Content should not be relied upon and should be independently
verified with primary sources of information. Taylor and Francis shall not be liable for any
losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities
whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or
arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial
or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or
distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use
can be found at http://www.tandfonline.com/page/terms-and-conditions
 Aging & Mental Health 2002; 6(2):153–160

ORIGINAL ARTICLE

Measurement and correlates of family caregiver self-efficacy for

managing dementia

R. H. FORTINSKY1, K. KERCHER2 & C. J. BURANT 3

1Division of Geriatrics and Center on Aging, University of Connecticut Health Center, Farmington, CT, USA;
2Department of Sociology & 3Department of Bioethics, Case Western Reserve University, Cleveland, OH, USA

Abstract
Researchers in the aging field are paying increasing attention to the importance of perceived self-efficacy in understanding
experiences and health-related outcomes of family caregivers. This paper details the strategy we used to measure family
Downloaded by [Erciyes University] at 21:24 19 December 2014

caregiver self-efficacy for managing dementia, and reports on observed associations between the resulting self-efficacy
measures, caregiver depressive symptoms, and caregiver physical health symptoms. Family caregivers (n = 197) were inter-
viewed after calling a local Alzheimer’s Association chapter in the mid-western USA. Nine items inquiring about
caregivers’ certainty that they could carry out specific behaviors related to dementia care clustered into two distinct self-
efficacy factors: symptom management self-efficacy (4 items) and community support service use self-efficacy (5 items).
Internal consistency reliability for both factors was high (Cronbach’s a = 0.77 and 0.78, respectively). Symptom manage-
ment self-efficacy demonstrated a much stronger correlation with a published global caregiver competence measure than
did service use self-efficacy (r = 0.49 and 0.27, respectively). In a multivariate regression model predicting caregiver
depression symptoms, higher symptom management self-efficacy scores were associated with fewer depressive symptoms
(b = -0.17, p < 0.05). In a separate model, higher service use self-efficacy scores (b = -0.20, p < 0.01) and higher symptom
management self-efficacy scores (b = -0.16, p < 0.05) were associated with fewer physical health symptoms. These new
measures of dementia management self-efficacy hold promise for use in future studies.

Introduction efficacy measures have been correlated with

performing self-care behaviors, and with reductions
Perceived self-efficacy is the belief that one can in depressive symptoms, in chronic arthritis patients
perform a specific behavior or task. Stemming from (Holman & Lorig, 1992; Horan et al., 1998; Lorig
social learning theory, the premise of perceived self- et al., 1989; Lorig et al., 1993).
efficacy is that individuals are more likely to perform This paper applies the principles of perceived
specific behaviors if they have greater certainty that self-efficacy to family caregivers of persons with
they can perform those behaviors (Bandura, 1991). Alzheimer’s disease and other types of irreversible
Moreover, perceived self-efficacy is a behavior- dementia. It is widely known that families provide
specific psychological attribute that can be learned the vast majority of health and personal care to
and enhanced (Bandura, 1986; Lorig et al., 1993). persons with dementia, often at great expense to
In the context of health promotion and chronic the physical and mental health of family caregivers
illness, perceived self-efficacy refers to the belief that (Ory et al., 1999; Schulz & Beach, 1999). More-
one can assume personal control over health over, a great deal is known about the stress process
problems by learning about key aspects of care underlying family care for relatives with dementia
(Bandura, 1991; Holman & Lorig, 1992; Lorig (Pearlin et al., 1990). Yet little is known about
et al., 1989). Higher levels of self-efficacy are levels of family caregivers’ self-efficacy for
expected to lead to health-promoting behaviors and managing dementia, where self-efficacy is defined
improved measures of physical and mental health. in terms of specific behaviors or tasks that can be
Measures of self-efficacy have been devised for learned, and that are highly relevant to daily
persons with musculoskeletal disorders, encom- management of the disease process. Several studies
passing domains such as pain control, diet, and have found that family caregivers are greatly inter-
exercise (Horan et al., 1998; Lorig et al., 1989). Self- ested in learning how to effectively manage their

Correspondence to: Richard H. Fortinsky, PhD, University of Connecticut Center on Aging, 263 Farmington Ave., Farmington,
Connecticut, USA, 06030–5215. Tel: (860) 679 8069. Fax: (860) 679 1307. E-mail: [email protected].

Received for publication 18th December 2000. Accepted 5th September 2001.
ISSN 1360–7863 print/ISSN 1364–6915 online/02/020153–08 © Taylor & Francis Ltd
DOI: 10.1080/1360786022012676 3
 154 R. H. Fortinsky et al.

relatives’ symptoms such as wandering, behavior personal or telephone contact with a physician since
problems, and incontinence, and in learning how to the dementia diagnosis were eligible for this study.
find and use community support services, such as
paid in-home care, adult day care, and respite care
(Fortinsky & Hathaway, 1990; Teri et al., 1992; Consent and data procedures
Brody et al., 1989; Deimling, 1991; MaloneBeach
et al., 1992). These domains of interest to family Eligible family caregivers received a letter from the
caregivers have not been developed into measures Association that described the larger study and
of perceived self-efficacy. encouraged their participation. Contact information
Therefore, the primary objective of this paper is about eligible family caregivers was forwarded from
to describe the development of a strategy to the Help Line staff the to the study team. A member
measure family caregiver self-efficacy for managing of the study team verified eligibility, and assigned
dementia, incorporating domains of symptom verified caregivers to study interviewers. Interviewers
management and community support service use. were instructed to wait at least one week after the
The intent of the measurement approach reported Association’s letter was mailed before initiating
here is to objectively distinguish family caregivers telephone contact with eligible caregivers. When
according to their degree of certainty about interviewers contacted eligible caregivers, they
knowing how to manage dementia symptoms and referred to the Help Line call and the letter from the
use community support services. A long-range goal Association, described the study in greater depth,
of this research is to determine if enhancement of and requested participation via verbal consent.
Downloaded by [Erciyes University] at 21:24 19 December 2014

self-efficacy in specific behavioral domains leads to Eligible caregivers were considered enrolled in the
improved dementia care management behaviors, study when they provided verbal consent. These
and improved mental and physical health, among consent procedures were approved by the human
family caregivers of persons with dementia. subjects review boards at University Hospitals of
The secondary objective of this paper is to Cleveland and at the Association. Two experienced
examine how self-efficacy for managing dementia is interviewers conducted all telephone interviews. Pre-
associated with: (a) family caregivers’ depressive testing and initial interviewer training took place in
symptoms; and (b) family caregivers’ physical health March 1996, and a refresher training session was
symptoms. Following the logic of self-efficacy held in June 1996 to coincide with the beginning of
theory, it is hypothesized that family caregivers sample enrollment. All interview data were entered
with higher levels of self-efficacy for managing and managed via SAS software on a mainframe plat-
dementia will report fewer depressive symptoms form, and were converted to SPSS system files for
and fewer physical health symptoms. analyses on the PC platform.

Methods Measures

Sample eligibility criteria Study measures in this section are described in the
following order: (1) family caregivers’ self-efficacy
As part of a larger study on dementia care for managing dementia; (2) caregivers’ physical and
experiences with physicians, family caregivers were mental health symptoms; (3) other caregiver charac-
prospectively recruited from July 1996–June 1997 teristics; and (4) severity of relatives’ dementia
from the population of callers to the telephone Help symptoms.
Line of the Cleveland Area Alzheimer’s Association
(the Association). Help Line staff at the Association Family caregivers’ self-efficacy for managing dementia.
incorporated three study eligibility criteria into their As discussed in the introduction section, two
telephone conversations with all callers throughout conceptual domains guided the development of
the 12-month enrollment period. First, callers were questions to measure caregiver self-efficacy for
required to be family caregivers, defined as managing dementia: symptom management and
individuals with primary or secondary responsibility community support service use. Questions were
for daily care of a relative with dementia. Second, designed for telephone interview format, and a limit
callers’ relatives with dementia were required to be of 10 self-efficacy questions was established at the
living in a home setting at the time of the Help Line outset to minimize respondent burden. Responses
call. Third, callers’ relatives were required to have a for all questions ranged from one (not at all certain)
physician’s diagnosis of either Alzheimer’s disease or to 10 (very certain), and caregivers were asked to
another type of irreversible dementia. This study place themselves on the 10-point scale for each
relied on verbal confirmation by callers that a question. All questions began with the phrase: ‘How
physician gave their relatives a formal diagnosis. As certain are you right now that you can…’, following
further verification of physician involvement in the self-efficacy measurement conventions developed in
diagnostic process, only callers who reported the arthritis literature (Lorig et al., 1989; 1993).
 Family caregiver self-efficacy 155

Items are listed below according to the conceptual this study primarily as external correlates to help
domains as proposed prior to any data analyses. validate the self-efficacy measurement strategy.
Self-efficacy for symptom management: five Type of family caregiver was categorized as either
items were originally designed to determine how primary (solo or primary) or secondary (equal or
certain family caregivers were that they could supplemental) responsibility for care. Relationship
manage their relatives’ dementia symptoms. The to patient was categorized as spouse, adult child,
items were: and other relative. Ethnicity was ascertained by
asking caregivers to classify themselves into an
Handle any problems your relative has, like
ethnic or racial group (White or African-American
memory loss, wandering, or behavior problems.
in this sample). Financial resources were deter-
Handle any problems that might come up in the mined by asking caregivers to report their
future with your relative’s care. perception of their financial resources. Caregivers
Deal with the frustrations of caring for your were asked to strongly agree, agree, disagree, or
relative. strongly disagree with seven statements about their
Do something to keep your relative as inde- money (e.g., we have enough money to buy
pendent as possible. groceries; we have enough money to manage major
Get answers to all your questions about your emergencies). This Perceived Adequacy of Income
relative’s problems. (PAI) scale was designed so that higher scores on
Self-efficacy for community support service use: five the 0–28 scale indicate greater perceived adequacy
of income (Noelker & Bass, 1994).
Downloaded by [Erciyes University] at 21:24 19 December 2014

items were originally designed to determine how

certain caregivers were that they could find and Care-giving competence represents a global
arrange support services in the community to measure of ‘intrapsychic strain’ as discussed by
supplement their own care. Items were: Pearlin and colleagues (1990) in their seminal
article on care giving and the stress process. Theo-
Care for your relative without help from organi- retically, it is expected that perceived self-efficacy,
zations or agencies that provide services. as a psychological construct related to intrapersonal
Find organizations or agencies in the community certainty and confidence, would be correlated with
that provide services to help you care for your global competence. In the present study, compe-
relative. tence is included for two reasons: first, to
Get answers to all of your questions about these determine if the two proposed domains of self-
services. efficacy are associated with competence at the same
Arrange for these services yourself. magnitude (i.e., an external correlate); and second,
Find ways to pay for these services. to determine if self-efficacy is associated with
depressive symptoms and physical health symp-
Caregivers’ mental and physical health symptoms. toms after controlling for competence. Within the
Depressive symptoms were measured using the 10- care giving competence scale, caregivers were first
item short form of the Center for Epidemiological asked to rate themselves on a four-point scale in
Studies-Depression inventory (Kohout et al., 1993). response to two items: ‘How much do you: (1)
Caregivers were asked if they felt any item much of believe that you’ve learned how to deal with a very
the time in the previous week (e.g., I felt depressed; I difficult situation; (2) feel that all in all, you’re a
could not get going). Each item was presented in a good caregiver’. Responses were very much, some-
yes/no format. Responses to each ‘yes’ item were what, just a little, and not at all. Then caregivers
scored as ‘1’; higher scores on the 0–10 scale were asked to rate themselves as ‘competent’ and as
represented more depressive symptoms. Internal ‘confident’ on a four-point scale. Responses were
consistency reliability was 0.77 in the study sample. very, fairly, just a little, and not at all (Pearlin et al.,
Physical health symptoms were measured using 1990).
the 12-item physical signs and symptoms subscale
from the Hopkins Symptoms Checklist (Derogatis Severity of relatives’ dementia symptoms. Four domains
et al., 1974). Caregivers were asked if they were of dementia symptom severity were measured:
bothered by each symptom during the previous cognitive status, behavioral problems, ability to
week (e.g., headaches; faintness or dizziness; pains conduct personal activities of daily living (ADL) and
in the heart or chest). Higher scores on the 0–12 instrumental ADLs, and bowel and bladder
scale represented a greater number of symptoms. continence.
The internal consistency reliability of this scale was Cognitive status was measured using the Cogni-
0.82 in this study. tive Status Scale developed and tested by Pearlin
and colleagues (Pearlin et al., 1990). This scale
Other caregiver characteristics. Several family caregiver contains eight items such as remembering recent
characteristics found to be significantly associated events, understanding simple instructions, and
with measures of family caregiver health and well- recognizing familiar people. Caregivers are asked to
being in the published literature were introduced in rate their relative on a four-point scale ranging
 156 R. H. Fortinsky et al.

from ‘can’t do at all’ to ‘not at all difficult’. the underlying factor structure of the 10 items.
Summed scale scores range from 0 (least impaired) Criteria used to determine the number of factors into
to 32 (most impaired). Pearlin and colleagues which the 10 items clustered were: factor eigen-values
(1990) reported that this scale had high internal ³ 1.0; and examination of distinct elbows in scree
consistency reliability in their San Francisco area plots of eigen-values. Items were included within a
sample (a = 0.86); reliability was nearly identical in specific factor if their primary loading was > 0.40, and
the present study (0.84). if their loading on a second factor did not approximate
Behavioral problems were measured using the the value on the primary factor.
Problematic Behavior Scale, also developed by Following the EFA, each of the resulting self-
Pearlin and colleagues (Pearlin et al., 1990). This efficacy factors was correlated with the same series
scale contains 14 items ranging from suspicious- of external correlates to determine whether they had
ness, agitation, night wandering, and anger; similar or different patterns of association with
responses are in terms of the number of days in the these external correlates. All bivariate relationships
previous week the caregiver had to deal with these were tested for non-linearity and influential cases.
behaviors. Scale scores range from 14 (no problems Scatter plots fitted with linear, quadratic and cubic
in previous week) to 56 (five or more days in past regression lines as well as the R2 associated with
week for all problems). Reliability was reported as each regression line did not indicate any substantial
a = 0.79 by Pearlin and colleagues; reliability was differences among the lines. Therefore linearity was
nearly identical in this study (0.77). not violated in these correlations. Additionally,
Independence in personal and instrumental Cook’s D was used to test for influential cases in
Downloaded by [Erciyes University] at 21:24 19 December 2014

activities of daily living (IADLs) was measured each of the bivariate relationships and no influential
according to the caregiver’s report about the cases were found.
patient’s ability to perform personal care and The second study objective was to determine
instrumental ADLs. Activities of daily living were how self-efficacy was associated with depression
based on a modified scoring of the Katz measure and physical health symptoms, after controlling for
(Katz et al., 1963); IADLs were based on a other important covariates. To achieve this objec-
modified version of the Duke Older Americans tive, multivariate linear regression models were
Resources and Services (OARS) measure (Duke, constructed in which caregiver depression and
1978). The ADL scale contained five items; caregiver physical symptoms measures were
bathing, dressing, eating, transferring from bed to employed as dependent variables. Independent
chair, and toileting. Caregivers were asked whether variables in these regression models included the
or not their relative could perform activities caregiver self-efficacy measures, global caregiver
without help; responses were in a yes/no format competence, caregiver socio-demographic charac-
(score range: 0–5). Internal consistency reliability teristics, and severity of patient symptoms. All
for the ADL scale in the study sample was 0.80. analyses were cross-sectional. In the multivariate
The IADL scale contained seven items ranging regression models tested there were no influential
from telephone use to handling finances; responses cases identified by the Cook’s D statistic. Addition-
were unable, some help, or no help (score range: 0– ally, testing for non-linearity by examining the
14). Reliability in the study sample for the IADL scatter plots of the partial correlations of the
scale was 0.75. Both scales were scored such that independent variables did not show any violation of
higher values indicated more activities performed linearity between each independent variable and
independently. the dependent variable.
Continence was measured according to caregiver
reports about whether their relatives sometimes
had accidents with bowels, and whether their rela- Results
tives sometimes wet themselves, during the day or
night. ‘No’ responses were coded as 0; therefore, Sample accrual and response rate
higher scores on the 0–2 scale indicated more
incontinence of bowel and bladder. A total of 257 individuals who called the Associa-
tion’s Help Line during the enrollment period met
study eligibility criteria and were contacted by study
Analyses interviewers. Of these, a total of 197 interviews were
completed, for a 77% response rate. The most
To achieve the first study objective, factor analytic common reasons for non-response were: relative
techniques were used to determine whether and how with dementia was already in a nursing home by the
the 10 self-efficacy items measured the two proposed time an interviewer reached a caregiver (n = 18);
conceptual domains of dementia symptom manage- interviewer was unable to locate caregiver by tele-
ment and community support service use. This study phone (n = 11); caregiver refused (n = 8); and
used exploratory factor analyses (EFA) based on prin- relative with dementia had died by the time the
cipal axis factoring with varimax rotation to identify caregiver was reached (n = 4).
 Family caregiver self-efficacy 157

Characteristics of family caregivers behavioral problem measures were in the midrange

of possible values (15.3 of a possible 32 for cognitive
Socio-demographic and health status characteristics status, and 29.1 of a possible 56 for behavioral
of the 197 caregivers are summarized in Table 1. problems). The mean number of personal ADLs
Caregivers in the study sample ranged in age from performed independently was 2.9 of five activities,
29–81 years old (mean = 56 years old). More than and the mean IADL score was 9.0 on a scale where
80% were female, and nearly 20% classified them- 14 indicated maximum independence. The mean
selves as African-American or Black. Adult children continence score of 1.2 indicates that the average
and children-in-law made up 62% of the sample patient in this group was incontinent of either bowel
members, while spouses accounted for 28% of the or bladder.
sample. Nearly two-thirds of sample members were
primary caregivers. Among caregivers who were
gainfully employed (46%), one-third worked full- Factor analysis of family caregiver self-efficacy for
time in addition to their care giving responsibilities. managing dementia
Results of the perceived adequacy of income score
suggest that most sample caregivers reported Exploratory factor analysis results of the 10 self-
moderate levels of financial resources (mean = 12; efficacy items are summarized in Table 3. Nine
SD = 2 on a scale ranging from 0–28). Caregivers items loaded into two distinct factors, while one item
also reported mild to moderate degrees of depres- (find ways to pay for services) did not load into
sive symptoms, with a mean of nearly three either factor. Both factors together explained 54% of
Downloaded by [Erciyes University] at 21:24 19 December 2014

symptoms (of a possible 10) reported in the week the total variance in the set of items. The content of
prior to the T1 interview. They also reported mild each factor was consistent with the two proposed
degrees of physical symptoms, with a mean of two conceptual domains: symptom management self-
symptoms (of a possible 12) during the previous efficacy (SXEFF) and community support service
week. use self-efficacy (SERVEFF), although specific
items loaded into these two factors with slightly
different groupings than initially conceptualized (see
Characteristics of relatives with dementia measures section). The internal consistency of the
two factors was found to be very adequate
Table 2 profiles socio-demographic and health (Cronbach’s a = 0.77 for SXEFF and 0.78 for
status characteristics of sample caregivers’ relatives SERVEFF). Descriptive statistics for the two self-
with dementia, as reported by the caregivers. efficacy factors were: for the 5-item SXEFF, the
Persons with dementia ranged in age from 56–96 possible range was 5–50; the mean score in the study
years old (mean = 78 years old), 60% were female, sample was 28.9 (SD.= 12.1). For the four-item
and nearly 20% were reported as African-American SERVEFF, the possible range was 4–40; the mean
or Black. Mean scores on cognitive status and score was 29.7 (SD.= 9.7).

TABLE 1. Characteristics of family caregivers

Characteristic Mean (SD) or percentage Sample size
Age 56 (14) years 197
Female 82% 197
African-American 19% 196
Spouse 28% 197
Adult child 62% 197
Primary caregiver 64% 197
Employed 46% 196
Adequacy of income 12 (2) 196
Depressive symptoms 3 (3) 196
Physical health symptoms 2 (3) 195

TABLE 2. Characteristics of relatives with diagnosed dementia

Characteristic Mean (SD) or percentage Sample size
Age 78 (7.2) years 197
Female 59.9% 197
African-American 18.8% 194
Cognitive status score 15.3 (7.5) 195
Problematic behavior score 29.1 (8.0) 195
Personal activities of daily living score 2.9 (1.8) 196
Instrumental activities of daily living score 9.0 (3.2) 196
Continence score 1.2 (0.9) 194
 158 R. H. Fortinsky et al.

TABLE 3. Factor analysis of 10 self-efficacy items (n = 195)a

Symptom management Community support service
Self -efficacy item self-efficacy (SXEFF) use self-efficacy (SERVEFF)
Handle any problems like memory loss 0.88 0.16
Deal with frustrations of caring 0.70 0.27
Handle problems that come up in future 0.69 0.09
Do something to keep relative independent 0.47 0.19
Care for relative without help from organizations or agencies 0.41 0.19
Find ways to pay for services 0.35 0.31
Get answers to all questions about services 0.11 0.85
Find organizations or agencies that provide services 0.15 0.70
Arrange for services yourself 0.32 0.61
Get answers to all your questions about your relative’s care 0.21 0.47
aBoldface coefficients reflect primary factor loadings at > 0.40 for each factor

TABLE 4. External correlates of symptom management self-efficacy (SXEFF) and community support
service use self-efficacy (SERVEFF)a
External Correlate Correlation with SXEFF Correlation with SERVEFF
Caregivers’ characteristics
Age -0.08 -0.12
Downloaded by [Erciyes University] at 21:24 19 December 2014

Female -0.14* -0.11

Spouse -0.06 -0.04
African-American -0.10 -0.06
Perceived adequacy of income 0.19** 0.32**
Global caregiver competence 0.49** 0.27**
Relatives’ characteristics
ADL independenceb 0.03 0.03
IADL independenceb 0.02 0.06
Cognitive statusc -0.11 -0.15*
Behavior problemsc -0.22** -0.17*
* p £ 0.05; **p £ 0.01; a n = 191 for all correlational analyses; bHigher score = greater independence;
cHigher score = more severe symptoms

External correlates of family caregiver self-efficacy global caregiver competence and other important
measures covariates. Results of multiple regression analyses
conducted to test this hypothesis are shown in
As an additional check to determine whether Table 5. The left-hand column in Table 5 shows
SXEFF and SERVEFF represented meaningfully results of the multivariate model in which depressive
distinct domains of caregiver self-efficacy for symptoms was the dependent variable, and the right-
managing dementia, each measure was correlated hand column shows the results of the physical health
with a series of external correlates. Results of the symptoms model. In the depressive symptoms model,
external correlate analysis are summarized in Table the statistically significant negative b coefficient for
4. Findings revealed that SXEFF and SERVEFF SXEFF (-0.17) indicates that sample caregivers who
showed a similar strength of association (or lack of reported a higher level of dementia symptom manage-
association) with most external correlates, but there ment self-efficacy tended to report fewer depressive
were two notable exceptions. First, caregivers’ symptoms. Although a trend toward a negative asso-
perceived adequacy of income was more strongly ciation was found between SERVEFF and depressive
correlated with SERVEFF (r = 0.32) than with symptoms, this association was not statistically signif-
SXEFF (r = 0.19). Second, SXEFF was much more icant. In the physical health symptoms model in
strongly correlated with global caregiver competence Table 5, SERVEFF showed a highly statistically
(r = 0.49) than was SERVEFF (r = 0.27). significant and strong negative association (b = -0.20,
p < 0.01). In other words, caregivers who reported a
higher level of self-efficacy for community support
Associations between family caregiver self-efficacy service use reported fewer physical health problems.
measures and caregivers’ depressive and physical health A weaker negative association was found between
symptoms SXEFF and physical health symptoms.

As discussed in the introduction section, it was

hypothesized that higher levels of both SXEFF and Discussion
SERVEFF among sample caregivers would be
associated with fewer depressive symptoms and with The objectives of this paper were to: (1) report
fewer physical health symptoms, after controlling for the development of a measurement strategy to
 Family caregiver self-efficacy 159

TABLE 5. Multiple regression models predicting family caregivers’ depressive symptoms and physical health
symptoms (n = 191)
Depressive symptoms Physical health symptoms
Independent variable Beta Beta
SXEFF -0.17* -0.16*
SERVEFF -0.11 -0.20**
Global caregiver competence -0.25** -0.17*
Female 0.03 0.05
African-American -0.07 -0.02
Primary caregiver 0.18** 0.18**
Adequacy of income -0.12 -0.15*
Relative’s cognitive status -0.13 0.001
Relative’s behavioral problems 0.18** 0.09
Relative’s ADL independence -0.34** -0.12
Relative’s continence 0.01 0.01
Model R2 0.362 0.300
* p £ 0.05; **p £ 0.01

operationalize the concept of family caregiver self- caregivers reported ways of dealing with stressors
efficacy for managing dementia; and (2) examine (Gignac & Gottlieb, 1996). Haley and colleagues
Downloaded by [Erciyes University] at 21:24 19 December 2014

how caregiver self-efficacy for managing dementia is (1996) briefly described a measure of family
associated with measures of caregiver health and caregiver self-efficacy in coping with dementia-
well-being. The intent of the measurement approach related problems. The present study builds on
was to incorporate items that reflected caregiver self- these published efforts by applying self-efficacy
efficacy for managing dementia symptoms and for theory to specific domains known to be important
using community support services. These domains of to family caregivers—dementia symptom manage-
behavior have been found to be critical in the lives of ment and community support service use—and
family members caring for relatives with dementia. developing measures that are consistent with these
Three major findings are emphasized. First, it was self-efficacy domains.
found that nine items concerning family caregivers’ Several study limitations are acknowledged.
perceived self-efficacy formed two distinct factors First, family caregivers were recruited and enrolled
reflecting the domains of dementia symptom after they called the telephone Help Line of a local
management and community support service use. chapter of the Alzheimer’s Association for informa-
Each of these factors was shown to have high internal tion or guidance. Compared to 1996 national
consistency reliability. Second, higher levels of figures on dementia caregiver characteristics (Ory
dementia symptom management self-efficacy were et al., 1999), this study sample was older (mean age
strongly associated with lower levels of caregiver of 56 years vs. 46 years in the national sample),
depressive symptoms, after controlling for numerous much more likely to be spouses (28% vs. 7%) or
covariates. Third, a negative association between adult children (62% vs. 49%), and much more
community support service use self-efficacy and likely to self-identify as White or Caucasian (81%
caregivers’ physical health symptoms also was found vs. 43%). Moreover, this study sample may be
after controlling for other factors. Therefore, findings inclined to report higher levels of self-efficacy for
indicated support for the hypothesized negative asso- community support service use given the fact that
ciation between self-efficacy and both mental and they had called a local chapter of the Alzheimer’s
physical health in sample family caregivers. Association. Therefore, the measurement strategy
Putting these findings in the context of published reported here should be replicated in other
literature, a limited number of efforts have been caregiver samples, to provide evidence of external
made to measure self-efficacy and related concepts validity of these measures. Second, because this
in family caregivers of older adults. Most of these study employed only cross-sectional analyses, it
efforts have not been domain-specific, or they have was not possible to determine causal pathways
viewed self-efficacy primarily as a coping mecha- between self-efficacy and either depression or
nism in response to the stress of caring for a relative physical health symptoms in sample caregivers.
with dementia. For example, measures of family Longitudinal studies would help clarify the direc-
caregiver ‘mastery’ and ‘competence’ have been tion of influence between self-efficacy and mental
advanced as important global measures, but and physical health measures. Third, observed
without reference to specific domains of care giving differences in the magnitude of correlation coeffi-
behaviors or tasks (Lawton et al., 1989; Pearlin cients between perceived adequacy of income and
et al., 1990). A measurement approach for ‘coping SXEFF vs. perceived adequacy of income and
efficacy’ among caregivers of persons with SERVEFF, and between global caregiver compe-
dementia has been reported, focusing on how often tence and SXEFF vs. global caregiver competence
 160 R. H. Fortinsky et al.

and SERVEFF, might have been due to chance FORTINSKY , R.H. (2001). Health care triads and dementia
variation. Therefore, conclusions based on these care: integrative framework and future diretions. Aging
findings should be made with caution. Finally, no & Mental Health, 5 (Suppl.1), S35–S48.
FORTINSKY , R.H. & HATHAWAY, T.J. (1990). Information
attempt was made in this paper to correlate and service needs among active and former family
reported self-efficacy with behaviors that reflect caregivers of persons with Alzheimer’s disease. The
actual dementia symptom management or commu- Gerontologist, 30, 604–609.
nity service use. This is an important avenue for GIGNAC, M.A.M. & GOTTLIEB , B.H. (1996). Caregivers’
future research. appraisals of efficacy in coping with dementia.
Psychology and Aging, 11, 214–225.
Despite these limitations, however, it is believed HALEY, W.E., R OTH , D., COLETON, M.I., et al. (1996).
that these new self-efficacy measures demonstrate Appraisal, coping, and social support as mediators of
useful properties for future studies that endeavor to well-being in Black and White family caregivers of
understand or improve family caregivers’ capacity patients with Alzheimer’s disease. Journal of Consulting
and Clinical Psychology, 64, 121–129.
to manage their relatives’ dementia symptoms and
HOLMAN, H.R. & LORIG, K. (1992). Perceived self-
find helpful community services to support their efficacy in self-management of chronic disease. In: R.
care (Fortinsky, 2001). This deconstruction of SCHWARTZER (Ed.), Self-efficacy: thought control of action.
caregiver self-efficacy into measurement domains New York: Hemisphere Publications.
that are meaningful to family caregivers, and that HORAN, M.L., KIM, K.K., GENDLER, P., FROMAN, R.D. &
PATEL, M.D. (1998). Development and evaluation of
are associated with their mental and physical
the osteoporosis self-efficacy scale. Research in Nursing
health, is an important advance at a time when the & Health, 21, 395–403.
Downloaded by [Erciyes University] at 21:24 19 December 2014

projected number of families caring for relatives KATZ, S., FORD, A.B., M OSKOWITZ, R.W., JACKSON, B.A.
with dementia is expected to rise dramatically over & JAFFE , M.W. (1963). Studies of illness in the aged.
the next several decades. The Index of ADL: a standardized measure of biological
and psychosocial function. Journal of the American
Medical Association, 185, 914–919.
KOHOUT, F.J., BERKMAN, L.F., EVANS, D.A. & CORNONI-
Acknowledgements HUNTLEY, J. (1993). Two shorter forms of the CES-D
Depression Symptoms Index. Journal of Aging and
The study on which this paper is based was funded Health, 5, 179–193.
LAWTON, M.P., KLEBAN, M.H., M OSS, M., ROVINE, M. &
by the AARP Andrus Foundation. We gratefully
GLICKSMAN, A. (1989). Measuring caregiving appraisal.
acknowledge the project management and data Journals of Gerontology: Psychological Sciences, 44(3),
management expertise provided during the course of P61–P71.
this study by Karen F. Bowman. PhD, and Miriam LORIG, K.R., CHASTAIN , R.L., UNG, E., SHOOR, S. &
Rosenblatt, MS, respectively. Portions of this paper HOLMAN, H.R. (1989). Development and evaluation of
a scale to measure perceived self-efficacy in people with
were presented at the 51st Annual Scientific Meeting
arthritis. Arthritis & Rheumatism, 32, 37–44.
of the Gerontological Society of America, Philadel- LORIG, K.R., MAZONSON, P.D. & HOLMAN, H.R. (1993).
phia, PA, USA, November 22, 1998. Evidence suggesting that health education for self-
management in patients with chronic arthritis has
sustained health benefits while reducing health care
costs. Arthritis & Rheumatism, 36, 439–446.
References MALONE-BEACH, E.E., ZARIT, S.H. & SPORE, D.L.
(1992). Caregivers’ perceptions of case management
BANDURA, A. (1986). Social foundations of thought and and community-based services: barriers to service use.
action. Englewood Cliffs, NJ: Prentice-Hall. Journal of Applied Gerontology, 11, 146–159.
BANDURA, A. (1991). Self-efficacy mechanism in physio- NOELKER, L.S. & BASS, D.M. (1994). Relationships
logical activation and health-promoting behavior. In J. between the frail elderly, informal and formal helpers.
MADDEN IV (Ed.), Neurobiology of learning, emotion and In: E. KAHANA, D. BIEGEL & M. WYKLE (Eds.), Family
affect (pp. 229–269). New York: Raven Press. caregiving across the life span. Thousand Oaks, CA: Sage.
BRODY, E.M., SAPERSTEIN, A.R. & L AWTON, M.P. (1989). ORY, M.G., H OFFMAN, R.R., YEE, J.L., TENNSTEDT ,
A multi-service respite program for caregivers of Alzhe- S. & SCHULZ, R. (1999). Prevalence and impact of
imer’s patients. Journal of Gerontological Social Work, 14, care giving: a detailed comparison between dementia
41–74. and non-dementia caregivers. The Gerontologist, 39,
DEIMLING, G.T. (1991). Respite use and caregiver well- 177–185.
being in families caring for stable and declining AD PEARLIN, L.I., MULAN, J.T., SEMPLE, S.J. & SKAFF, M.M.
patients. Journal of Gerontological Social Work, 18, (1990). Care giving and the stress process: an overview
117–134. of concepts and their measures. The Gerontologist, 30,
DEROGATIS, L.R., LIPMAN, R.S., R ICKLES, S.P., 583–594.
UHLENHUTH, U.H. & COVI, L. (1974). The Hopkins SCHULZ, R. & BEACH, S.R. (1999). Care giving as a risk
symptom checklist (HSCL): a self-report symptoms factor for mortality. Journal of the American Medical
inventory. Behavioral Sciences, 19, 1–5. Association, 282, 2215–2219.
DUKE U NIVERSITY CENTER FOR THE STUDY OF AGING TERI, L., RABINS, P., WHITEHOUSE, P.W. et al. (1992).
AND HUMAN DEVELOPMENT. (1978). Multidimensional Management of behavior disturbance in Alzheimer
functional assessment: the OARS methodology. Durham, disease: current knowledge and future directions.
NC: Duke University Press. Alzheimer Disease & Associated Disorders, 6, 77–88.