| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
a Graduate Psychology Program, Trinity Western University, Langley, British Columbia, Canada
Correspondence: P. S. Fry, PhD, Graduate Psychology Program, Trinity Western University, 7600 Glover Road, Langley, British Columbia, V2Y 1Y1, Canada. E-mail: Fry{at}twu.ca.
Decision Editor: Laurence G. Branch, PhD
 |
Abstract |
|---|
 TOP Abstract The Underlying Theoretical Model...Widowhood, Self-Efficacies, and...Overall Scope of the...HypothesesMethodsResultsDiscussionAppendixReferences |
|---|
Key Words: Predictors of perceived quality of life • Self-efficacy beliefs of widowed elderly • Predictors of well-being following spousal loss
With more and more elderly persons living considerably longer after the death of a spouse (Bengtson, Rosenthal, and Burton 1990
; Krause 1986; Mendes de Leon, Kasl, and Jacobs 1994), understanding the self-efficacy beliefs that are linked to perceived diminishment or maintenance of quality of life and life satisfactions following a period of spousal loss is becoming a critically important matter to gerontologists. Identifying social–cognitive beliefs of self-efficacy and sociodemographic factors that are predictive of men and women's heterogeneous perspectives of the quality of their lives and their life satisfactions following spousal loss is of increasing interest to clinicians, practitioners, and families concerned with assisting elderly individuals in maintaining a reasonable quality of life following widowhood (see discussions by Campbell and Silverman 1999; Lieberman 1996). Also of interest to practitioners is the way in which widows and widowers differ in self-efficacy beliefs in various domains of their interpersonal, social, emotional, and physical functioning, and how these differences may be related to their perceived quality of life, life satisfactions, and self-esteem in the period following spousal loss (see Hansson, Remondet, and Galusha 1993; Lieberman 1996, for full discussion). In keeping with this line of thinking, the purpose of the study was to examine the extent to which widowed individuals' preexisting self-efficacy beliefs (both global and domain specific) are predictive of their perceived quality of life, life satisfactions, and self-esteem in the period following loss and to establish whether assessments of self-efficacy domains of men and women should be usefully included in any existing programs of self-care and coping with spousal loss (Pearlman and Uhlmann 1991).
|
The Underlying Theoretical Model and Framework for the Present Study |
|---|
|
TOPAbstractThe Underlying Theoretical Model...Widowhood, Self-Efficacies, and...Overall Scope of the...HypothesesMethodsResultsDiscussionAppendixReferences |
|---|
. Bandura 1993 proposed that underlying self-efficacy beliefs of individuals afford an improved understanding of the cognitive processes and mechanisms that buffer a person's sense of control against the limitations of the loss of physical, social, and temporal resources. Within the context of spousal loss, our theoretical assumptions concerning the effects of widowhood derive from the premise that the extent to which elderly widows and widowers perceive a decline in quality of life or life satisfactions depends on their prewidowhood self-efficacy beliefs in specific domains of functioning and the personal significance of those domains to their previous functioning. Following Bandura's social–cognitive model (Bandura 1993), we posit that widows' and widowers' global sense of efficacy and domain-specific efficacies assists them in preserving positive perceptions of quality of life and positive perceptions of self-esteem and life satisfactions in the long term following spousal loss. Thus, social-learning theorists (for example, Bandura 1993; Berry and West 1993; Smith, Kohn, Savage-Stevens, Finch, et al. 2000) speculate that preexisting self-efficacy beliefs or expectancies of elderly individuals influence the level of effort they expend to preserve their quality of life and the level of perseverance they may exhibit in the face of threats to their perceived quality of life. Social–cognitive theorists would posit that widows and widowers with weaker self-efficacy beliefs, at the outset, may put forth less effort than individuals with stronger preexisting self-efficacy beliefs to combat diminishment and deterioration. |
Widowhood, Self-Efficacies, and Self-Esteem |
|---|
|
TOPAbstractThe Underlying Theoretical Model...Widowhood, Self-Efficacies, and...Overall Scope of the...HypothesesMethodsResultsDiscussionAppendixReferences |
|---|
; Lund, Caserta, and Dimond 1993; Quandt, McDonald, Arcury, Bell, and Vitolis 2000) has implicated decline in self-esteem following spousal loss as a function of lowered emotional efficacy. Following spousal loss, further risks to self-esteem occur through a variety of pathways: Loss of meaningful spousal role and loss of significant relationships with other associates of the deceased spouse may erode self-esteem. Increased physical seclusion, social isolation, and reduction in social activities following spousal death may severely distort perceptions of self-worth and esteem (Arens 1982; Ferraro 1984). Finally, frequently diminished economic resources, instrumental dependency, and interpersonal dependency of the widowed spouse may increase vulnerability and lead to decline in self-esteem. The indications are (e.g., Bianchi's 1995, meta-analysis review; Bengtson et al. 1990) that individuals with lower levels of self-efficacy are at greater risk for loss of self-esteem. No previous study, however, has been identified that examines the relationship between spousal loss, self-efficacy, and self-esteem. |
Overall Scope of the Study |
|---|
|
TOPAbstractThe Underlying Theoretical Model...Widowhood, Self-Efficacies, and...Overall Scope of the...HypothesesMethodsResultsDiscussionAppendixReferences |
|---|
), instrumental efficacy (Berry and West 1993), physical-health efficacy (Bausell 1986), social-support efficacy (Rowe and Kahn 1997), financial efficacy (Bryant and Morgan 1989), emotional efficacy (see Beckingham and Watt 1995), nutritional efficacy (Quandt et al. 2000), and spiritual-health efficacy (Fry 2001a & 2001b) are individually related to the psychological well-being of elders. Certainly, no previous study has examined the predictive value of the full core of domain-specific self-efficacy of elderly persons in the context of spousal loss. |
Hypotheses |
|---|
|
TOPAbstractThe Underlying Theoretical Model...Widowhood, Self-Efficacies, and...Overall Scope of the...HypothesesMethodsResultsDiscussionAppendixReferences |
|---|
First, we hypothesized that global self-efficacy beliefs and domain-specific self-efficacy beliefs are both important to the prediction of widowed elderly individuals' perceptions of quality of life, life satisfactions, and self-esteem following spousal loss. The logic of this hypothesis is consistent with Bandura 1993 proposal that self-efficacy is a global construct for some individuals but a domain-specific construct for others. Thus the two constructs need to be assessed separately. Second, we hypothesized that widows, compared with widowers, will show stronger baseline self-efficacy beliefs in the interpersonal, social-support adequacy, and spiritual domains, whereas widowers will show stronger self-efficacy beliefs in the financial, instrumental, and physical-health domains. These hypotheses of gender differences in domain-specific self-efficacy beliefs are substantiated by the findings of earlier studies (e.g., Antonucci 1990; Carr et al. 2000; Fry 1998, Fry 2001a, Fry 2001b; M. S. Stroebe and Stroebe 1983; N. Stroebe, Stroebe, and Schut 1997) showing that women, compared with men, have different domains of strength. Third, we hypothesized that widows' stronger self-efficacies in the domains of social-support adequacy, interpersonal efficacy, and spiritual-health efficacy and widowers' stronger self-efficacies in the financial, instrumental, and physical-health domains would be predictive of higher levels of perceived health-related quality of life, life satisfaction, and self-esteem in the longer term following spousal loss. Fourth, on the basis of Bandura 1993 self-efficacy model of growth, we hypothesized that widowed individuals' self-efficacy beliefs in various domains of functioning would be more significant predictors of their perceived quality of life, life satisfactions, and self-esteem than would the sociodemographic factors of income, education, social support, physical health, and age.
|
Methods |
|---|
|
TOPAbstractThe Underlying Theoretical Model...Widowhood, Self-Efficacies, and...Overall Scope of the...HypothesesMethodsResultsDiscussionAppendixReferences |
|---|
, Fry 2001a, Fry 2001b). Community-residing widows and widowers for the present study were obtained mainly from among 284 attendees at social support groups for recently bereaved elders, organized by a number of seniors' centers and community agencies in three midsized cities in Southern Alberta. Participants were also recruited through community newsletters and posters placed in prime social services agencies and by word of mouth with caregivers of elderly persons. Eligibility criteria included spousal loss approximately 4 to 6 months ago. Other criteria were that participants must be English speaking, must reside in the community (not in institutions), and must not have physical or cognitive impairments that would preclude taking paper-and-pencil tests of psychological well-being. As a result of preliminary inquiry, 242 individuals (132 widows, 110 widowers) agreed to take baseline measures that were proposed for this study at Time 1 and agreed also to be followed up 12 months later at Time 2. Eleven volunteer participants (7 women, 4 men) did not meet the criterion of adequate physical and cognitive functioning and were excluded. Some further attrition in the sample (n = 20) occurred also at Time 2, partly because of relocation, disapproval by another family member, or from other functional difficulties experienced by the individual in the course of the 12 months preceding the follow-up. It should be noted that the final attrition rate at Time 2 was approximately 9%. The baselines of the 20 individuals who were absent at Time 2 assessments were compared with baselines of the 211 individuals who participated in Time 2 assessments. No significant differences were found between baselines of these individuals. Thus, analyses of the data that followed from Time 1 and Time 2 assessments are reported for a final sample of 211 widowed elderly persons (118 widows, 93 widowers).
Description of Participants
As seen in Table 1 , participants were predominantly of Caucasian background; the majority of participants reported their health to be in the "good" category (categories were fair, good, excellent). Most participants had the equivalent of a high school education, and most reported that their income was adequate for making ends meet or for a comfortable living. The majority of participants reported that they had been married for over 20 years.
|
General Design
The present study, designed to be an 18-month follow-up study, reports on a subset of 14 variables: sociodemographics (5), global self-efficacy measures (1), and domain-specific self-efficacy measures (8), which were hypothesized to be predictors of perceived health-related quality of life, life satisfactions, and self-esteem in both samples of widows and widowers. At Time 1 (4 to 6 months after spousal loss), all participants responded to baseline measures of sociodemographics and self-efficacy reportings. A period of 4 to 6 months after spousal loss was chosen on the basis of Parkes 1995 guidelines and Parkes and Weiss 1983 comments, noting that it is ethically not appropriate to intrude on bereaved individuals' privacy before this time. At Time 2 (i.e., 18-months follow-up of respondents) participants responded to various psychosocial measures including the Health-Related Quality of Life Perceptions Scale (HRQOL), Life Satisfaction Index (LSIA), and Self-Esteem Index (SEI). An 18-month period for follow-up of participants was proposed for both theoretical and practical considerations. A number of studies on widowhood (e.g., Fry 1998, Fry 2001a, Fry 2001b; Lieberman 1996; Lopata 1973; Parkes and Weiss 1983) have implicated 2 years following spousal loss to be a critical period of adjustment to spousal loss and a changed quality of life. Practical considerations were that individuals involved in the study were uncertain about future plans and therefore unwilling to commit to a longer term follow-up beyond the 2 years following spousal loss.
Measures
Adopting constructs of quality of life developed by earlier researchers (e.g., Diener 1984; Diener and Suh 1997; Lawton 1991) and recent researchers who have investigated quality-of-life components (e.g., Allen, Becker, and Swank 1990; Oka, Gortner, Stotts, and Haskell 1996), we selected both objective indicators of quality of life (e.g., income, gender, education, social support, and health status) and subjective measures of quality of life such as perceived health-related quality of life, life satisfactions, and self-esteem. Although some concern has been expressed in the gerontological literature (e.g., Lohmann 1977) in regard to life satisfaction and quality of life being highly correlated measures, Lawton, Moss, Hoffman, and colleagues (1999) and Hoyt and Creech 1983 confirmed that these measures are independent. Preliminary analyses in our study also confirmed that correlations among the dependent measures of health-related quality of life, life satisfaction, and self-esteem were only minimally significant, with coefficients ranging between .11 and .16.
Baseline Measures
Sociodemographic Characteristics.
Sociodemographic characteristics included in these analyses were age, gender, income, education, physical health status, and social support status. Education was measured as the highest grade completed. Final analyses used a dichotomous classification of less than 12 years completed versus 12 or more years completed. Annual income was measured in $10,000 increments to $40,000. Final analyses used a dichotomous classification of $15,000 to $20,000 versus $20,000 to $35,000. In the present study, we decided to use both objective measures of social-support availability and physical health and subjective measures of perceived social support efficacy and perceived physical health efficacy.
Physical Health Status Objective Indicator.
Physical health was assessed by a 26-item physical health checklist of illnesses (Older Americans Resources and Services; Duke University Center for the Study of Aging and Human Development 1978) most commonly encountered by older adults. This is a measure that has been widely used in gerontological studies to assess physical-health status (e.g., Fry 2001a; Lawton et al. 1999). Participants responded yes or no to the question, "Do you have any of the following illnesses at the present time?" A high score of yes responses was assumed to reflect poor health status. Alpha coefficients were .69 to .75 for the present sample of widows and widowers, respectively.
Social-Support Resources.
These resources were measured by a shorter modified version of the Multidimensional Scale of Social Support (Zimet, Dahlem, Zimet, and Farley 1988). This scale has been used with nonclinical samples of elderly individuals (e.g., Fry 1998, Fry 2001b). Fry 2001b, using the short form, reported excellent psychometric properties, reliability, and validity for an elderly sample of widowed individuals, especially considering the limited number of items in the scale. Internal reliability (alpha coefficient) was .86. Three sources of available support were measured by asking individuals to report the number of persons among family, friends, and significant others regularly available to give social support. A composite higher score was construed to mean higher social-support availability from the three sources. Cronbach's coefficients were .76 and .86 for widows and widowers, respectively.
Self-Efficacy Measures.
1. Subscales of Domain-Specific Efficacy. The baseline measure of domain-specific efficacies in this study was composed of 25 items (each arranged on a 4-point Likert scale) that were developed and validated on two pilot-study samples drawn from my previous studies (Fry 1998, Fry 2001a, Fry 2001b). Participants for the pilot studies were selected on the basis of sociodemographic criteria similar to those used in the present study. Self-efficacy perceptions identified in eight life domains (interpersonal, instrumental, emotional, social support, financial, physical health, nutritional, and spiritual health) were found to be of particular relevance to the day-to-day functioning of older adults. Eight subscales, with 35 items, were created on the basis of a priori selection of items, item analysis, factor loadings, and a rational approach to scaling that argues for a strategy of constructing summary measures from moderately rather than highly correlated items (Edwards 1970; Goldberg 1972). With further refinement of items and elimination of 10 items with low reliability, the inter-item correlations were improved for the remaining 25 items, and ranged from .69 to .73. Three-month test–retest reliability data drawn from a previous pilot study of 87 individuals showed subscale correlations ranging from .50 to .67 (see Bandura's 1977, Bandura's 1993, discussion in support of moderate correlations among subscales of domain-specific efficacies). Cronbach's alphas for the eight subscales ranged from .70 to .78 and from .77 to .81 for widows and widowers, respectively. (Details of the eight subscales of domain-specific efficacies, along with specimen items and coding procedures are provided in the A.)
2. Global Efficacy Index. This baseline measure consisted of one item: "Taking into account various areas of my day-to-day functioning (for example, personal life, social life, family life, religious life, physical health, and emotional health), I feel I am handling myself competently." This item was scored on a 4-point Likert-type scale (ranging from strongly agree [4] to strongly disagree [1]). The higher the score obtained, the stronger was the individual's perceived global self-efficacy.
Dependent Measures
An abbreviated measure of the HRQOL (Stewart, Hays, and Ware 1988) was developed using three subscales of the SF-36. The HRQOL is a widely used and validated measure for collecting elderly individuals' health status information, and short forms and longer forms of the HRQOL have been used in several recent studies of elderly adults (e.g., Lyons, Perry, and Littlepage 1994; Leon and Moyer 1999). The SF-36 assessment provides eight multi-item subscales derived from using a 3-point to 6-point Likert-type measure for collecting physical- and mental-health-related information. Our review of the findings of the various studies in which the SF-36 item survey was used revealed that the three subscales of general health perceptions, vitality, and mental health perceptions had been found to reflect the best summary measures of mental and physical health-related quality-of-life assessments. They also conveyed the information most relevant to clinicians and those in outcome research (see discussion by Arnold 1991; Leon and Moyer 1999). They are recognized to be the best summary measures of health-related quality of life (see Chapter 10 in Ware, Kosinski, and Keller 1994), and have high reliability and greatest relevance to elders' health status. These three subscale measures are (a) Subscale for General Health Perceptions, which evaluates current general health as well as whether or not the person believes his or her health will deteriorate (five questions), (b) the Vitality Subscale, which determines the degree to which the person feels full of energy and "pep" as well as how often he or she is tired or worn out (four questions), and (c) Mental Health Subscale, which examines nervousness and feelings of depression, sadness, peacefulness, and happiness (five questions). Responses to the 14 questions included in the composite measure of the three subscales of the HRQOL were arranged on a 5-point Likert type scale ranging from strongly disagree (1) to strongly agree (5). The higher the cumulative score, the more positive are the HRQOL perceptions. Range of scores on this measure is 14 to 70. Following procedures approved by Ware and colleagues 1994, we derived an aggregate HRQOL score for each individual based on the subscale scores of general health perceptions, vitality, and mental health. Cronbach's alphas for this survey were .75 and .82 for widows and widowers, respectively.
LSIA; Neugarten, Havighurst, & Tobin, 1961.
This is a 20-item scale designed to measure subjective feelings of well-being and satisfaction among older adults. Life satisfaction is viewed as a multifaceted concept incorporating domains of well-being, finding life meaningful, a feeling of success in achieving goals, a positive self-image, and an optimistic attitude (McDowell and Newell 1987; Neugarten et al. 1961). Validity studies compared the LSIA scale with the Philadelphia Geriatric Center Morale Scale (Lohmann 1977) and life satisfaction ratings by clinical psychologists (Neugarten et al. 1961), with correlations of .74 and .64, respectively. Independent factor analysis studies (Hoyt and Creech 1983) have reported consistent findings of a three-factor solution of congruence, mood tone, and optimism. The LSIA is one of the more widely used scales of life satisfaction for older adults, and has several strengths of reliability, strong correlations with other scales, and consistent validity findings (McDowell and Newell 1987). The present study used the original "agree/disagree" response format. Scores range from 0 to 20. Higher scores represent higher levels of life satisfaction. Cronbach's alphas for this inventory were .76 and .74 for widows and widowers, respectively.
SEI.
The Rosenberg Self-Esteem Inventory (Rosenberg 1979) was used to measure self-esteem as a global and stable disposition. This measure is widely known and the most frequently used measure for different age populations. The inventory has 10 items, 5 positively keyed and 5 negatively keyed, and all items used a 4-point scale ranging from strongly disagree (1) to strongly agree (4). The negatively keyed items were recoded, and the 10 items were summed to produce a total score showing that higher scores indicate higher levels of self-esteem. Scores range from 10 to 40 on this self-esteem measure. Cronbach's alphas for this inventory were .80 and .72 for widows and widowers, respectively.
All paper-and-pencil self-report measures used in this study were previously piloted on an independent volunteer sample of 40 older adults who met all of the same eligibility criteria used for participants in this present study.
Statistical Design and Analyses
A hierarchical multivariate linear regression analysis was conducted on the data derived from the 14 predictor variables: 5 sociodemographic variables, 1 global self-efficacy index, and 8 subscales of domain-specific self-efficacy. The order of entry of the predictor variables into the linear regression model was based on methodological, historical, and theoretical considerations in the gerontological literature on spousal bereavement and widowhood (e.g., Arens 1982; Fry 1998, Fry 2000, Fry 2001b; Siegel and Kuykendall 1990; Wortman and Silver 1990; Yalom and Lieberman 1991) and on psychosocial development and self-efficacy development in late life (Rodin and McAvay 1992; Seeman, Unger, McAvay, and Mendes de Leon 1999). Consistent with order of entry of variables proposed by these earlier studies, the block of sociodemographic variables was entered first in the regression model, followed by global self-efficacy (Step 2) and the block of eight subscale efficacies (Step 3). Instances of missing data were rare (<1%), therefore presenting little concern about Type I and related Type II errors occurring (Schumacker and Lomax 1996). All analyses were performed using SAS 6.12 software (SAS Institute 1990).
|
Results |
|---|
|
TOPAbstractThe Underlying Theoretical Model...Widowhood, Self-Efficacies, and...Overall Scope of the...HypothesesMethodsResultsDiscussionAppendixReferences |
|---|
) indicated no significant multicollinearity. Intercorrelations were modest, ranging from .14 to .27, median = .08. Thus multicollinearity did not present a concern. Independent sample t tests (df = 209) were conducted on all measures used, and as they were all significantly different at the p < .05 level, they were subsequently entered in the two separate regression analyses for widows and widowers (see Table 4 and Table 5 ). It should be noted that the potential for Type 1 errors was closely monitored in the series of hierarchical regression analyses conducted separately for widows' and widowers' samples. It was judged to be minimal and did not present a concern.
|
|
|
|
LSIA.
Sociodemographic variables entered in Step 1 of the regression accounted for 8% of the variance in LSIA, which is quite significant. Global self-efficacy and subscale efficacies, when entered in Steps 2 and 3, accounted for 18% and 31%, respectively, of the explained variance in LSIA, which is also very significant (see Table 4 ). Subscale efficacy variables that made a unique and salient contribution to the explained variance include interpersonal efficacy, emotional efficacy, social-support efficacy, and spiritual-health efficacy (see significant beta values in Table 4 , under LSIA).
SEI.
With the exception of social support, sociodemographic variables failed to contribute significantly to the variance in SEI of elderly widows. However, when global self-efficacy and subscale efficacies were entered in Steps 2 and 3 of the regression, they accounted for 22% and 31%, respectively, of the explained variance in the SEI of elderly widows. As indicated by their significant beta values, interpersonal efficacy, instrumental efficacy, emotional efficacy, social-support efficacy, and spiritual-health efficacy individually made a unique and salient contribution to the variance in SEI of elderly widows (see significant beta values in Table 4 , under SEI).
Widowers' Sample
HRQOL.
The results of linear regression analysis performed on the widowers' sample show that the sociodemographic variables collectively accounted for a significant percentage of the variance in HRQOL (i.e., 8%). Similarly, when global self-efficacy and subscale efficacies were entered in Steps 2 and 3 of the regression, they accounted for 17% and 23%, respectively, of the explained variance in HRQOL of elderly widowers. As indicated by their significant beta values, instrumental efficacy, financial efficacy, physical-health efficacy, and nutritional efficacy individually made a salient contribution to the explained variance in HRQOL of widowers (see significant beta values in Table 5 , under HRQOL).
LSIA.
The regression analyses show that sociodemographic variables, entered together in Step 1, accounted for a significant percentage of the variance in LSIA (i.e., 7%). Similarly, when global self-efficacy and subscale efficacies were entered in Steps 2 and 3 of the regression, they accounted for 22% and 26%, respectively, of explained variance in LSIA. As indicated by their significant beta values, instrumental efficacy, emotional efficacy, financial efficacy, and physical-health efficacy individually made a significant contribution to the explained variance in LSIA of widowers (see Table 5 for significant beta values, under LSIA).
SEI.
The regression analyses show that sociodemographic variables, entered in Step 1, accounted for a significant percentage of the variance in SEI (i.e., 11%). Global self-efficacy and subscale efficacies, entered in Steps 2 and 3, accounted for 26% and 25%, respectively, of the explained variance in SEI of the widowers. As indicated by their significant beta values, instrumental efficacy, emotional efficacy, financial efficacy, and spiritual-health efficacy individually accounted for a significant level of variance in SEI of widowers (see significant beta values in Table 5 , under SEI).
As evident in both sets of regression analyses for widows and widowers, global self-efficacy and subscale efficacies accounted for almost double the percentage of variance that was accounted for by the sociodemographic variables toward predicting HRQOL, LSIA, and SEI (see Table 4 and Table 5 ).
Salient Gender Differences
A comparison of the two sets of hierarchical regression analysis for the widows' and widowers' samples revealed a number of salient gender differences. First, education and income were significantly more potent predictors of quality of life, life satisfaction, and self-esteem for widowers, compared with widows. Second, the subscale efficacies that contributed to the predictive value were different for the two groups. Instrumental efficacy, financial efficacy, and physical-health efficacy uniformly revealed higher beta weights for widowers, compared with widows. Conversely, interpersonal efficacy, emotional efficacy, social-support efficacy, and spiritual-health efficacy uniformly revealed higher beta weights for widows, compared with widowers.
|
Discussion |
|---|
|
TOPAbstractThe Underlying Theoretical Model...Widowhood, Self-Efficacies, and...Overall Scope of the...HypothesesMethodsResultsDiscussionAppendixReferences |
|---|
, Bandura 1993 theoretical formulation that self-efficacy beliefs of individuals in various domains of functioning contribute to their physical and mental health-related quality of life. As hypothesized, weaker levels of perceived self-efficacy in various domains such as interpersonal, instrumental, emotional, social support, financial, physical health, nutritional, and spiritual health at Time 1 (baselines) were associated with lowered perceptions of health-related quality of life, life satisfactions, and self-esteem for both widows and widowers at Time 2 (follow-up). On the basis of our results, which show significant associations between baseline self-efficacy ratings and ratings of health-related quality of life, life satisfactions, and self-esteem at follow-up, it seems reasonable to conclude that preexisting weak and strong domains of self-efficacy of elderly spouses are indeed reliable predictors of their later adjustments to spousal loss and widowhood.
The finding that widows, compared with widowers, had stronger levels of perceived efficacy in the interpersonal, social-support, and spiritual-health domains and that widowers had stronger levels of perceived efficacy in the financial, instrumental, and physical-health domains is consistent with the hypothesis of the present study, and supports the findings implicit in previous studies (see Arens 1982; Bianchi's 1995, review; Carr et al. 2000; Ferraro, Mutran, and Barresi 1984; Lopata 1973) that have examined gender differences in widowed persons. Several earlier researchers have speculated that women's confidence in beliefs about their ability to manage interpersonal activities, to generate large social networks, and to sustain inner strength through spiritual beliefs helped them adjust more easily than men have to various losses, including loss of a primary relationship. Our findings support this speculation. The findings tell an interesting story about gender differences in interpersonal dependence, instrumental dependence, financial dependence, and spiritual dependence that has important implications for practitioners. They further substantiate Bandura's social–cognitive position, which argues that strong perceptions of self-efficacy in various domains such as interpersonal, social, and instrumental may be more conducive to higher levels of life satisfactions and self-esteem, whereas weak perceptions of self-efficacy in various domains may imply that the individual is at risk for lowered self-esteem following a crisis or a critical loss.
The overall finding that self-efficacy beliefs in several domains were more important predictors of perceived quality of life and life satisfactions for both widows and widowers than were sociodemographic factors is also consistent with the hypothesis of the present study.
Implications of the Findings for Practitioners
The transcending theme of perceived self-efficacy in various domains as predicting the quality of life and life satisfactions for both widows and widowers has important implications for practitioners. The fact that self-efficacy beliefs are modifiable (see Bandura 1993; Meichenbaum 1994) makes them an excellent target for interventions with widowed individuals who may be struggling with weak self-efficacy beliefs in several domains of independence, autonomy, and control. While surveying the history of widowed individuals, practitioners need to take greater account of domains of weak and strong self-efficacy beliefs to plan programs of self-care and self-management consistent with their efficacy needs. Specific domains in which perceived self-efficacy is low should be identified through discussion with the widowed individuals, and practitioners may offer suggestions for enhancing self-efficacy. Because spiritual health-efficacy emerged as a strong predictor of life satisfactions and self-esteem of widowed elderly people, its importance in intervention programs should be more widely recognized by practitioners. Currently, intervention programs to assist widowed individuals in adjusting to spousal loss have focused exclusively on teaching new social skills and new instrumental tasks (see Caserta, Lund, and Rice 1999; Stroebe and Stroebe 1987). Recently however, several researchers (e.g., Campbell and Silverman 1999; Fry 2001b; Thorson 1999; Yalom and Lieberman 1991) are urging geriatric practitioners to include components for strengthening self-efficacy, especially spiritual self-efficacy, in intervention programs for adjusting to spousal loss.
Limitations of the Study
Several limitations are acknowledged. First, as with any study of such a sensitive nature as spousal loss, the use of volunteer participants, although unavoidable, presents an important limitation to the sampling framework. The sample, although generally representative of community elders, is skewed toward individuals who were White, in fair to very good health, and living in the southern region of Alberta. Consequently, the generalizability of this study is limited by the characteristics of the sample. A second limitation is that the number of years for which individuals were married was not included as a predictive variable, and we did not consider potential interactions among variables in our exploratory study. Consequently, future follow-up studies based on ours may find our initial model to be limited in these respects. In spite of these shortcomings, it should be noted that we expanded on prior research in a number of ways. First, we used prospective data to examine the possible contributions of self-efficacy beliefs of widowed elderly (at baseline) to their perceived quality of life and life satisfactions at 18-month follow-up. On the basis of our preliminary findings, we concluded that there is a strong predictive link between self-efficacy beliefs of bereaved spouses and their future resiliency. Second, we focused on men and women separately and were able to identify different sets of self-efficacy domains that characterize current cohorts of elderly widows and widowers. Both these features are of considerable relevance to the work of geriatric clinicians and counselors and until recently have been neglected in studies of widowhood and overlooked in intervention programs for widowed elderly people.
Practice Concepts
The Forum
Book Reviews
|
Acknowledgments |
|---|
We thank two anonymous reviewers for their insightful comments on the manuscript and also many older adults living in Calgary and in outlying areas of Airdrie, Cochrane, Crossfield, High River, Lethbridge, and Red Deer, in Southern Alberta, who responded to the call for participants. The help of Evelyn Doyle and Anne Humanek in recruitment of participants and data collection is especially appreciated. Diane Hussein, Patricia Parham, and Charlene Batlowski assisted in tabulation and organization of data files, and Mark Kolodziej helped with the statistical procedures and interpretations of the statistical analysis at various stages. Consistent with understandings established with the elderly participants in this study, all raw data files, consent forms, and list of addresses for participants were destroyed on the completion of the follow-up study.
The results of the follow-up analyses were presented at a symposium organized by P.S. Fry at the 108th Annual Convention of the American Psychological Association, Washington, DC, August 2000.
Received for publication November 10, 2000. Accepted for publication July 9, 2001.
|
Appendix |
|---|
|
TOPAbstractThe Underlying Theoretical Model...Widowhood, Self-Efficacies, and...Overall Scope of the...HypothesesMethodsResultsDiscussionAppendixReferences |
|---|
Coding procedures: All 25 items were scored on a 4-point scale (strongly agree [4], strongly disagree [1]). Scores for the total scale range from 25 to 100. Respondents are encouraged to respond carefully to each item in order to obviate missing scores in any subscale items. Summary scores for the subscales reflect the average of item responses, weighted by the number of items in the subscale.
|
References |
|---|
|
TOPAbstractThe Underlying Theoretical Model...Widowhood, Self-Efficacies, and...Overall Scope of the...HypothesesMethodsResultsDiscussionAppendixReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M.-A. Lee and D. Carr Does the Context of Spousal Loss Affect the Physical Functioning of Older Widowed Persons? A Longitudinal Analysis Research on Aging, September 1, 2007; 29(5): 457 - 487. [Abstract] [PDF]
|
|
|
|
 |
|
 M. Berg-Weger, S. M. Rauch, D. M. Rubio, and S. S. Tebb Assessing the health of adult daughter former caregivers for elders with Alzheimer's disease American Journal of Alzheimer's Disease and Other Dementias, July 1, 2003; 18(4): 231 - 239. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
