University of Southern California Dissertations and Theses

Determinants of life satisfaction among the high-functioning elderly: Implications of effects of psychological factors and activity participation

(USC Thesis Other)

Determinants of life satisfaction among the high-functioning elderly: Implications of effects of psychological factors and activity participation

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Aiso, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back o f the book. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. UMI A Bell & Howell Information Company 300 North Zeeb Road, Ann Arbor MI 48106-1346 USA 313/761-4700 800/521-0600 with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. D ETER M IN A N TS OF LIFE SATISFACTION A M O N G TH E H IG H FU N C TIO N IN G ELDERLY: IM PLICA TIO N S OF EFFECTS OF PSY CH O LO G ICA L FACTORS AND A CTIV ITY PA R TIC IPA TIO N by Jung Ki Kim A Thesis Presented to the FA CU LTY OF TH E LEONARD D A V IS SCHOOL OF G ERO N TO LO G Y U N IV ER SITY OF SO U TH ERN CA LIFO RN IA In Partial Fulfillment o f the Requirem ents for the Degree M A STER OF SCIENCE IN G ERO N TO LO G Y A ugust 1997 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. UMI Number: 1387840 UMI Microform 1387840 Copyright 1998, by UMI Company. AH rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. UMI 300 North Zeeb Road Ann Arbor, MI 48103 s J R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. UNIVERSITY OF SOUTHERN CALIFORNIA LEONARD DAVIS SCHOOL OF GERONTOLOGY University Park Los Angeles, CA 90089 This thesis, written by Tung Ki jCm______________________________________ under the director of h e r Thesis Committee and approved by all its members, has been presented to and accepted by the Dean of the Leonard Davis School of Gerontology and the Dean o f : _________________________ in partial fulfillment of the requirements for the degree of j l i Dean Dean Date THESIS COMMITTEE Chairman R eproduced with perm ission of the copyright o w n e r Further reproduction prohibited without perm ission Acknowledgments I wish to thank several people for the thesis. I w ould like to thank Dr. Eileen Crimmins for her academic input, warm guidance, support and encouragement. Also I would like to thank my committee members, Dr. David Peterson and Dr. Teresa Seeman for their careful review and input. I also would like to give special thanks to my parents. They are always my supporters and 1 cannot overestimate their support for me. Lastly but m ost importantly, I would like to appreciate God for his endless and faithful love and care for me. S— if - I " R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. TA B LE OF C O N TE N T S PAGE IN T R O D U C T IO N .................................................................................................................................1 R ELEV A N T R E S E A R C H ............................................................................................................... 2 Life S atisfaction............................................................................................................................ 2 Definition o f Life Satisfaction Com parison with O ther Constructs M easures o f Life Satisfaction Factors Affecting Life S atisfaction..........................................................................................7 Socio-dem ographic Factors H ealth Factors Socioemotional Factors a. Netw orking/Social Support b. Activities Psychological Factors Objective Conditions and Perceived Conditions and Their Com parison C onclusion....................................................................................................................................21 M E T H O D S .......................................................................................................................................... 26 D ata and Subject S electio n ..................................................................................................... 26 M e a su re s.......................................................................................................................................26 A N A L Y S IS .......................................................................................................................................... 36 R E S U L T S .............................................................................................................................................37 D IS C U S S IO N ..................................................................................................................................... 58 Study Strengths and W eak n esses..........................................................................................62 IM P L IC A T IO N S ................................................................................................................................ 63 B IB L IO G R A P H Y ..............................................................................................................................66 A P P E N D IX .......................................................................................................................................... 69 * S i* R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. iv L IST OF FIG U RES AND TABLES PAGE Figure 1 M odel o f Life S atisfaction...........................................................................................................22 Figure 2 M odel o f Unidimensional Life Satisfaction in the S tu d y .................................................... 27 Figure 3 M odel o f M ultidimensional Life Satisfaction in the S tu d y .................................................28 Table 1 Characteristics o f the Study Population: M acA rthur Study o f Successful A g in g ...........................................................................................................................38 Table 2 M ean and Percentage o f Satisfaction Variables: M acA rthur Study o f Successful A g in g .................................................................................................................... 39 Table 3 a Objective Conditions and Subjective C onditions................................................................41 Table 3 b Unstandardized Param eter Estim ates o f Objective Conditions and Subjective Conditions in M odel o f Life Satisfaction..........................................................42 Table 3 c Standardized Param eter Estim ates o f Objective Conditions and Subjective Conditions in M odel o f Life Satisfaction..........................................................45 Table 3d U nstandardized Param eter Estim ates o f Objective Conditions and Subjective Conditions and Selected Variables in M odel o f Life S atisfaction............... 46 Table 4a Unstandardized Param eter Estim ates o f Psychological Factors and O ther Selected Variables in M odel o f Life S atisfaction.................................................... 48 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table 4b Standardized Param eter Estimates o f Psychological Factors and O ther Selected Variables in M odel o f Life Satisfaction..................................................... 49 Table 5a Unstandardized Param eter Estimates o f Activity Participation and O ther Selected Variables in Model o f Life Satisfaction by Employment S ta tu s ...............................................................................................................................................52 Table 5b Standardized Param eter Estim ates o f Activity Participation and O ther Selected Variables in M odel o f Life Satisfaction by Employment S ta tu s ..................... 53 Table 6 Life Satisfaction o f the Employed with Higher Job Satisfaction and Low er Job Satisfaction................................................................................................................57 Table 7 Unstandardized Param eter Estimates o f Selected Variables in Life Satisfaction by Job S atisfaction................................................................................................57 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 1 INTRODUCTION Increased life expectancy and the accompanying interests in quality o f life in the expanded lifetimes have caused a proliferation o f studies on life satisfaction. The focus o f these studies varies from the measurement o f life satisfaction to examination o f specific factors causing life satisfaction. The purpose o f this study is to examine factors related to life satisfaction among elderly people with a special focus on psychological factors and activity participation. The study o f life satisfaction is important because it promotes understanding o f what can improve life satisfaction and suggests practical strategies for enhancing adjustment to later life. The analysis is based upon data from the M acArthur Study o f Successful Aging, which contains information on a variety o f topics from socio-demographic and health to psychosocial factors. Since the sample is drawn from a community-dwelling population, it may be appropriate to make inferences about how retirement preparation programs could be developed to incorporate aspects that would increase life satisfaction. This paper clarifies how psychological factors are related to life satisfaction. It investigates whether and how much subjective assessment o f various aspects o f life is important relative to objective conditions o f life. Also, the effect on life satisfaction o f psychological well-being, as indicated by mastery, depression, efficacy, and happiness, is examined. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 2 Then, the paper examines the role o f activities and employment in life satisfaction among the elderly. Specifically, the effects o f formal and informal activity participation, and their interaction with employment on life satisfaction are examined. Finally, the effect on life satisfaction o f job satisfaction among the employed elderly is investigated. This is discussed in conjunction with activity participation and its interaction with employment. RELEVANT RESEARCH Life Satisfaction Definition o f Life Satisfaction Life satisfaction is a concept that has been frequently used in the social sciences as well as policy areas to measure the quality o f life. When life expectancy was lower, the major issue was how to promote longer life. That is, people focused on the quantity o f life. Since the life expectancy o f the population is increasing, interest has switched to the quality o f life and has produced a corresponding increase in interest in life satisfaction. According to M cConatha (1993), life satisfaction involves how individuals view their present lifestyle as well as future expectations. The term, life satisfaction, is somewhat self-suggesting: how content people are with their life. However, this vague conceptualization o f life satisfaction is not helpful in indicating how to measure life satisfaction. Because life satisfaction involves many dimensions, it is not a simple task to define life satisfaction. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 3 Life satisfaction is, as the term implies, ambiguous as well as comprehensive because it includes every possible dimension o f life conditions. The multi-dimensional aspect o f life satisfaction suggests different ways o f looking at the quality o f life. Some try to define quality o f life by assessing the intake o f nutritious food or comfortable housing, and others, by measuring physical conditions such as health problems and disease status. Thus, both defining and operationalizing life satisfaction are difficult. Comparison with Other Constructs There are several terms or constructs that have been used in quality o f life research: well-being, happiness, satisfaction, morale, affect, mood, etc. Since the focus is the elderly, most o f whom are retired, terms such as adaptation and adjustment to changes in later life (e.g. retirement, widowhood, health changes) are also frequently used. Similarities and differences in these approaches are identified by Kozma et al. (1991). Kozma et al. (1991) differentiate the above concepts by the degree o f inclusiveness. For example, domain satisfaction is inclusive o f happiness. While the former is a measure of satisfaction with specific domains in life, the latter is that o f global subjective well-being. W hen they talk about ‘satisfaction,’ they mean domain satisfaction as compared to life satisfaction which is overall satisfaction with life. Because Kozma et al.'s classification is based on a psychological point o f view, however, it does not comprehensively capture the issue o f life satisfaction for all o f the social sciences. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 4 M easures ofLife Satisfaction Researchers have employed a variety o f life satisfaction measures. In the initial approach to measuring life satisfaction, Neugarten et al. (1961) attempted to establish a refined and valid measure o f successful aging or well-being, which could be used with other social and psychological variables. They postulated five components for their Life Satisfaction Rating scales (LSR). Their index is one o f the first approaches to establishing a comprehensive and valid measure o f life satisfaction with which elderly people's well-being could be addressed. The five components in this approach are (1) zest vs. apathy, (2) resolution and fortitude, (3) congruence between desired and achieved goals, (4) self- concept, and (5) mood tone. Each o f these components is discussed below. First, zest vs. apathy is the extent to which respondents are enthusiastic about various activities, persons, or ideas. This rating considers the type o f activities o f each person as well as different personality types. That is, ‘satisfied’ people are not necessarily those who are involved in ‘active or outgoing’ activities. Those who like sitting home and knitting can be highly satisfied with their activities, and thus, with their life. Second, the approach to resolution and fortitude is to assess the extent to which respondents accept personal responsibility for their life. Third, congruence between desired and achieved goals assesses the extent to which respondents feel that they have achieved their goals in life, whatever those goals might be. What they mean by self-concept is physical as well as psychological and R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 5 social self-concepts. Lastly, those who express happy optimistic attitudes and mood are scored high in life satisfaction (mood tone). Since the LSR requires a long interview, Neugarten et al. developed self-report instruments which could be completed in a short time. One o f them, the Life Satisfaction Index A (LSIA), consists o f 25 attitude items for which only an agree or disagree response is required. Among the questions are ‘As I grow older, things seem better than I thought they would be,’ and ‘My life could be happier than it is now.’ Menec and Chipperfield (1997) used the LSIA to measure life satisfaction in their study. Campbell et al. (1976) specified the dimensions o f life satisfaction as satisfaction with health, activities, finances, social interaction, etc. This conceptualization o f life satisfaction is similar to that ofK ozm a et al. (1991) and Dorfman (1985). Ferris and Bramston (1994) measured elderly people's quality o f life in residential facilities with the Comprehensive Quality o f Life Scale (ComQol), which assesses both subjective and objective life quality. They specified seven life domains such as material possessions, health, productivity intimacy, safety, place in the community, and emotional well-being (happiness) in examining the elderly people's quality o f life in details. Higginbottom et al. (1993) defined retirement satisfaction with multifaceted constructs, consisting o f satisfaction with the activities engaged in, finances, health, and people the retirees associate with. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 6 M cIntosh and Danigelis (1995) used positive and negative affect as a measure o f well-being. Even though their way o f measuring life satisfaction is not comprehensive, it captures the psychological well-being o f elderly people specifically. Krause (1993), in his study o f race differences in life satisfaction, measured life satisfaction by assessing positive mood tone (e.g. optimistic attitudes and happy feelings) and negative mood tone (e.g. pessimistic attitudes and unhappy feelings). Interestingly, he measured positive mood tone and negative mood tone separately and found some factors influence positive mood tone but not negative mood tone, and vice versa. Lawton (1983) argues that life satisfaction is composed o f tw o parts. One is an internal well-being and the other is an external well-being. Internal well-being is, for example, attitude on aging, agitation, optimism, self-concept, resolution, etc. External well-being is, on the other hand, subjective satisfaction with financial satisfaction and subjective health, and so forth. Multi-item scales o f life satisfaction are good for measuring particular aspects o f life satisfaction. However, as Kozma et al. (1991) point out, because they contain certain preconceptions about the nature o f life satisfaction, the results may be biased. In addition, a measure with multi-items may contain different aspects o f quality o f life in a single scale. Thus, a multi-item life satisfaction scale needs to be confirmed as valid by factor analysis. As used in Palmore et al.'s (1985) study, global (unidimensional) life satisfaction measure is conducive to compensate for the weakness o f multidimensional life satisfaction measure. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 7 As noticed above, because researchers conceptualize life satisfaction in different ways, consensus on how to measure life satisfaction is hard to achieve. Factors Affecting Life Satisfaction There are potentially many factors that may affect elderly people's life satisfaction. Based on the literature review, four major factors that could affect life satisfaction are discussed below. Often, the literature produces conflicting results. In part, this may reflect differences in control for confounding variables. Socio-demographic Factors Researchers who examine life satisfaction among the elderly commonly use socio demographic variables as generic causes o f life satisfaction. Among the commonly used socio-demographic factors are age, gender, education, income, race, and marital status. Age is included as a direct, indirect, or possibly control variable (Aquino et al., 1996; Davis, 1991; Dorfman, 1995; Higginbottom et al., 1993; Kozma et al., 1991; Krause, 1993; M enec & Chipperfield, 1997; Palmore et al., 1985; Szinovacz & Washo, 1992). Age may directly affect life satisfaction or it may be spuriously related to life satisfaction through its relationship with another variable. Palmore et al. (1985) used two longitudinal data sets in their retirement satisfaction study: the National Longitudinal Survey o f Older Men (NLS) and the Retirement History - 1 ” R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Study (RHS). The sample o f the NLS is men between 45 and 59 at baseline while that o f the RHS is persons (both men and women) between 58 and 63 at baseline. The findings from these tw o data sets differ in the effect o f age on life satisfaction. Age was a significant predictor o f life satisfaction in the NLS. The older the subject, the greater the life satisfaction in the NLS. However, age was not a significant predictor o f life satisfaction among the RHS respondents. The latter result may be more applicable to elderly people on whom this paper focuses. The majority o f studies on life satisfaction have produced results similar to those o f the RHS, that is, no significant effect o f age on life satisfaction. There are a lot o f studies o f gender differences in life satisfaction or studies that include gender as a generic factor related to life satisfaction (Aquino et al., 1996; Davis, 1991; Dorfman, 1995; Kozma et al., 1991; Krause, 1993; M atthew & Brown, 1987; M cIntosh & Danigelis, 1995; Menec & Chipperfield, 1997; Palmore et al., 1985; Seccombe & Lee, 1986). Aquino et al. (1996) found no gender effect on life satisfaction. On the other hand, Szinovacz and Washo (1992) found gender differences in the effects o f life event exposure on retirement adaptation such that women tended to report more life events and their retirement adaptation was more affected than men's by the experience o f life events. Their explanation is that it may be due to women's greater involvement in social network crises as well as their dependence on social supports. Since men and women differ in numerous ways, gender difference in life satisfaction may represent differences in many other factors. I R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 9 Income and financial situation (Aquino et al., 1996; Davis, 1991; Higginbottom et al., 1993; Szinovacz & Washo, 1992), which is related to education (Aquino et al., 1996; Davis, 1991; Kozma et al., 1991; Krause, 1993; McIntosh & Danigelis, 1995; Menec & Chipperfield, 1997; Palmore et al., 1985; Szinovacz & Washo, 1992), can affect life satisfaction. Financial situation has been shown to be related to subjective well-being as well as resources available such as activities, health care purchase, etc. However, using path analysis, Aquino et al. (1996) found neither education nor adequate income is related to life satisfaction. On the contrary, the regression model with the RHS by Palmore et al. (1985) revealed education and income adequacy as predictors o f life satisfaction. The findings on the relationship between race and life satisfaction are equivocal. Krause (1993), similar to Palmore et al. (1985), reported lower levels o f life satisfaction among older blacks than among older whites, and he attributed this race difference mainly to the interplay among educational attainment, economic plans for retirement, current financial difficulties, and economic dependence on family members rather than race itself Marital status is also a commonly used variable that may affect life satisfaction (Aquino et al., 1996; Davis, 1991; Dorfman, 1995; Iwatsubo et al., 1996; M cIntosh & Danigelis, 1995; Palmore, et al., 1985; Szinovacz & Washo, 1992). However, the findings on the relationship between marital status and life satisfaction are also equivocal. Dorfman (1995) found no significant relationship o f marital status with retirement satisfaction in his R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 10 multiple regression model while Iwatsubo et al. (1996) found a significant effect o f marital status on life satisfaction: the married have higher life satisfaction than the unmarried. Among socio-demographic factors that are less frequently used in life satisfaction studies are the length o f retirement (Dorfman, 1995; Szinovacz & Washo, 1992), household extension (Szinovacz & Washo, 1992), employment (Aquino et al., 1996; Davis, 1991), spouse's employment status (Szinovacz & Washo, 1992), occupational prestige (Dorfman, 1995; Szinovacz & Washo, 1992), and geographic location (Aquino et al., 1996). Employment seems to be important in life satisfaction. However, because the focus in the study o f life satisfaction o f the elderly is on ‘elderly people,’ most o f whom are retired, employment is a less important factor than in the study o f life satisfaction among the general population. Only when the study contains a reasonable number o f the employed are comparisons o f the life satisfaction o f the retired and the employed or life satisfaction o f people before and after retirement studied among the elderly. Socio-demographic factors are included as background information on life satisfaction. However, most studies show indirect effects or no effect of those factors on life satisfaction. As George and Clipp (1991) say, it would be worthwhile to integrating more subjective or perceptual dimensions o f socio-demographic factors into the conceptual model, as will be discussed later in this section. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 11 Health Factors There seems to be no debate that health is an essential factor in life satisfaction (Aquino et al., 1996; Davis, 1991; Dorfman, 1995; Higginbottom et al., 1993; Iwatsubo et al., 1996; M cIntosh & Danigelis, 1995; Menec & Chipperfield, 1997; Palmore et al., 1985; Szinovacz & Washo, 1992). However, the issue o f what dimensions/types o f health are the best predictors o f life satisfaction and how strong the effect o f health on life satisfaction is, is a lot more debatable. The answers depend on how much focus researchers put on health variables. Some researchers find that perceived health rather than health conditions dominates life satisfaction. Some studies even do not include health factors in their analysis. For example, Krause (1993) focused on race differences in life satisfaction among elderly men and women and excluded possible health intervention in life satisfaction. Health factors, for elderly people in particular, can be measured with several dimensions: physical health conditions such as diseases or illnesses, psychological or mental health conditions like depressive symptoms, and functional conditions such as ADLs (Activities o f Daily Living) impairments, IADLs (Instrumental Activities o f Daily Living) impairments, and other activity limitations. Different studies operationalize health factors in different ways. M ence and Chipperfield (1997) used chronic health problems (presence o f specific diseases/conditions such as heart and circulation problems, hypertension, heart attack, stroke, arthritis, palsy, diabetes, cancer, amputation, and problems with eyes, ears, teeth, chest, stomach, kidneys, R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 12 feet, and skin) and found a negative relationship with life satisfaction. While they aggregated these conditions as health problems, Dorfman (1995) separated life-threatening health conditions from non-life-threatening conditions. The former include heart attack, stroke, cancer, diabetes, and pulmonary disease, and the latter, hearing impairment, vision impairment, urinary incontinence, and arthritis. H e divided the dimensions o f retirement satisfaction (satisfaction with activities, finances, health, and people) and assessed the effects o f single health conditions on each dimension. All o f the life-threatening health conditions except cancer and all o f the non-life-threatening health conditions except urinary incontinence w ere predictors o f dissatisfaction with health for one or both genders. However, he found relatively fewer significant effects o f non-life-threatening conditions than o f life-threatening conditions. For other dimensions o f satisfaction, none o f the health conditions, except stroke on activities in men and diabetes on finances in women, predicted satisfaction. It would be interesting to sum these four dimensions o f satisfaction into one life satisfaction scale and see how different types o f health conditions affect life satisfaction as a whole. While some research includes only current health problems (Szinovacz & Washo, 1992), others investigate the change o f health conditions. Palmore et al. (1985) examined whether health limitations or impairments have increased. Changes in health conditions give a better picture on the effect o f immediate health conditions on life satisfaction. Palmore et al. found a significant relationship between health limitations and life satisfaction: the lower the health limitations, the higher the life satisfaction. However, they did not report how health R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 13 changes were related to life satisfaction because their inclusion o f question on health changes was not, in fact, analyzed in a way to show the actual effect o f health changes on life satisfaction. They did not compare health conditions o f tw o times o f the survey. They used two models, prospective and concurrent models. B ut neither o f their models incorporated health conditions in both times o f survey: prospective models included baseline independent variables and a dependent measure (subjective well-being) from a later point in time; in the concurrent models, both the dependent and independent variables including health were measured at the final test date. Longitudinal studies would be more effectively used in life satisfaction study if health changes in different times could be compared. In addition, it would be interesting to compare the effect o f current health conditions and that o f changes in health conditions. Since most elderly people have chronic health conditions, their health status should not be measured solely based on diseases or illnesses. Physical ability to perform daily activities and/or other activities (Kozma et al., 1991; M cIntosh & Danigelis, 1995) should be considered as one o f the critical basis o f the elderly’s health. Functional health is usually measured with ADLs and/or IADLs. Among the ADLs and IADLs are the ability to engage in day-to-day activities like bathing, eating, taking medications, taking walks, etc. As shown above, some studies measure an overall health status by simply asking whether respondents currently have health problems or have acquired them since retirement or within the past year(s). On the other hand, other studies specify types o f health conditions * s i A : v * ■ R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 14 and see how specific health conditions aflfect life satisfaction. A decision on which measure is better should be based on the focus o f the study. However, a comparison between current health conditions and changes in health conditions, and o f severe health problems and activity-limiting conditions, in addition to the comparison between objective health conditions and perceived health conditions, would provide a more thorough examination o f the effect o f health factors on life satisfaction. Socioemotional Factors Since a person cannot live without a social context, contact or activity with or related to other people has a potential to affect life satisfaction. Palmore et al. (1985) said that Indicators o f social activities1 ... the generic class o f independent variables hypothesized to be the most proximate predictors o f subjective well-being.... (Palmore et al., 1985) a. Networking/Social Support Elderly people are related to other people in many ways. They may live with other people such as a spouse, children, siblings, other relatives, or friends. They may be members o f such organizations as a church, a health fitness club, or other activity groups. They may do 1 Activities here include not only activities such as meeting and phoning neighbors and friends, attending religious services, non-employment works, but solitary activities and self- care activities as well. ! R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 15 some volunteer works. They may work in various settings with or without getting paid. These human resources may provide physical as well as emotional help when needed. Having a strong network o f social relationships with family (Iwatsubo et al., 1996), friends (Larson et al, 1986), neighbors, and community/religious ties is widely believed to be a factor enhancing life satisfaction among the elderly. Socioemotional selectivity theory supports the idea that elderly people are adapted to aging by increasing emotional closeness to significant others (Carstensen, 1992). Having close friends (McConatha, 1993; O'Connor, 1995) and frequent visits with friends (Larson et al., 1986; Szinovacz & W asho, 1992) contribute to retirement adaptation. Larson et al. (1986), similar to O'Connor (1995) and Goudy & Goudeau (1981), in their analysis, showed that friendship is a more important predictor o f life satisfaction than family relationship. The explanation is partly related to frequency and quality o f active leisure activities with friends. Even though older adults get emotional support from family members as well as from their friends, the unique quality o f reciprocal support with friends is hard to get from family. It is related to the weak point o f exchange theory which overlooks the quality o f exchange between older people and other age groups. [Exchange theory defines] social interaction solely in terms o f the number o f interactions initiated. (Passuth & Bengtson, 1988) i— i - R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. r - 16 In addition to the benefits o f getting support, providing support to others (Krause et al., 1992) also appears to increase the feelings o f personal control in later life, and lower levels o f depression, in turn enhancing life satisfaction, b. Activities Although such activities as shopping, reading, involvement in hobbies and sports may be viewed as being closely related to levels o f physical functioning, they also reflect adequacy o f role performance, thus predicting life satisfaction. Activity theory argues that active elderly people are more likely to have greater life satisfaction than those who are inactive (Burgess, 1954). Even though this theory has been criticized because o f its overlook o f variations in the meaning o f particular activities in the lives o f older people2 (Passuth & Bengtson, 1988) and its unclear explanation on how different types o f activity, such as formal, informal, and solitary activities, would affect elderly people's life satisfaction, it seems that this theory deserves consideration because it initiated considering the potential benefits o f activities. Aquino et al. (1996) showed the effects o f paid and volunteer work on social support and life satisfaction by controlling various demographic and health characteristics and concluded that participation in paid and volunteer work may be an important determinant o f life satisfaction among the elderly. In particular, it was noted that volunteer positions increase 2 Activity in activity theory holds more interpersonal nature (informal activity) that offers role supports. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 17 life satisfaction through social support mechanisms while the relationship between working at a paying job and life satisfaction is not mediated by social support. Aquino et al. (1996) also showed that although whether a person participates in volunteer activities is associated with social support, the number o f hours volunteered is not. Thus, greater involvement in volunteer activities does not appear to lead to higher levels o f social support, in turn higher life satisfaction. Menec and Chipperfield (1997) examined the potential mediating role o f exercise and participation in nonphysical leisure activities such as attending cultural events, involvement in volunteer organization, etc. and found that exercising and nonphysical leisure activity participation are predictive o f positive perception on health and greater life satisfaction. While M enec and Chipperfield (1997) differentiated physical and nonphysical activities in their study, Iwatsubo et al. (1996) combined three types o f leisure activities together: physical activities, hobby or other leisure activities such as painting or reading, and social activities in association, trade, unions, or clubs. Their findings also confirm the positive relationship between leisure activities and life satisfaction. If personality difference and health conditions are considered, division o f physical and nonphysical leisure activities doesn't seem to be needed. Palmore et al. (1985) used several measures o f activities and social relationships: participation in formal organizations, time spent in extra-work interests (including home maintenance and other instrumental activities), interaction with family, interaction with R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 18 friends, and time spent in solitary activities. They found that increased interaction with friends and increased time in solitary leisure activities are positively related to life satisfaction. Particularly, increased participation in formal organizations and time spent in extra-work interests are predictors o f retirees' (but not nonretirees') increased life satisfaction. This finding is conceivable because retirees may feel deprived o f their identity as workers, so that those who pursue the "work-like" activities are more likely to have greater life satisfaction. This finding is supported by Jahoda's (1982) latent function model o f employment (deprivation theory) which identifies five latent functions o f employment such as time use, social contacts, self-identify, regular activities, and participation in collective purposes. McIntosh and Danigelis (1995) also found that both paid and unpaid productive activities outside the home such as paid work, formal religious/nonreligious participation, and informal volunteering are positively related to life satisfaction. Psychological Factors Psychological factors such as self-concept, self-efficacy, depression, self-esteem, and locus o f control (personal mastery) are likely to be involved in people's life satisfaction particularly at the advanced ages when numerous changes happen and thus corresponding adjustments are needed (McConatha, 1993). Neugarten et al. (1968) empirically tested an activity theory and a disengagement theory o f successful aging and found that neither theory is not supported by empirical R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. findings. They argue that it is not a social environment nor a some set o f intrinsic changes that elderly people cannot influence but personality that is the main key to predict the relationship between level o f social role activity and life satisfaction. According to McConatha (1993), self-concept is formed o f the self-perceptions individuals have. A positive self-concept is associated with coiif.dence, independence, assertiveness, and goal-seeking behavior. On the other hand, self-esteem, which is regarded as a part o f self-concept, is the emotional evaluation people have o f themselves. Self-concept or self-esteem is essential in elderly people's life satisfaction. Their roles have changed from a working role to a retired role (McConatha, 1993) and their health conditions have been deteriorated. In this situation, their adjustment to those changes is largely influenced by their self-concept. Locus o f control or personal mastery is the belief that events in life are the result o f personal actions rather than determined by external forces or fate (McConatha, 1993). The logic is that older people with more o f locus o f control are more likely to be adjusted to changes easily in their later life, thus having higher life satisfaction. Measuring depression is another way o f examining psychological well-being (Aquino et al., 1996; Higginbottom et al, 1993). In many studies, psychological well-being is used as one portion o f life satisfaction rather than an independent variable affecting life satisfaction. It seems more reasonable to include psychological well-being as a factor affecting life satisfaction rather than one , i * * i ~ R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 20 component o f life satisfaction. That is because regarding psychological factors as a portion o f life satisfaction measures ignores those factors’ possible interaction with other variables such as activity participation, perceived health conditions, and network, to name a few. Also psychological factors as a component o f life satisfaction may not leave a room for psychological intervention which is quite possible through counseling or education programs. Objective Conditions and Perceived Conditions and Their Comparison Davis (1991) reported a nationwide telephone-interview survey o f older people (ages 50-64) regarding their life satisfaction. Her study concluded that full-time working, the married, high income, and good health conditions are major predictors o f life satisfaction. H er study is a typical example that neglects the perception o f conditions o f life in studying life satisfaction. In addition, although she included a range o f socio-economic (age, race, gender, working status, marital status, education, and income) and health status, she did not include the likely effects o f other social and psychological factors discussed earlier in this paper. Already mentioned above, researchers have recently focused on the perceptual or perceived conditions rather than solely focusing on the objective conditions. Among them are perceived financial situations (Aquino et al., 1996; Higginbottom et al., 1993; Krause, 1993), perceived health conditions (Kozma et al., 1991), and perceived social support (Aquino et al., 1996). Kozma et al. (1991) argued that the independent contributions o f objective health to s r r 1 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 21 an explanation o f life satisfaction appear to be low compared to those o f subjective health. Krause (1987) argued that By merely assessing the frequency or amount o f social support that has been exchanged, researchers may be overlooking the potentially important role played by variations in individual need for support. In fact, a number o f researchers have argued recently that perceived adequacy o f support may be the best predictor o f well-being in the elderly. (Krause, 1987) It is unreasonable to solely focus on subjective conditions o f life, omitting baseline effects o f objective determinants just as it is to ignore perceived conditions o f life in a study o f life satisfaction (George & Clipp, 1991). Thus, both objective and subjective conditions o f life should be incorporated in study o f life satisfaction study with proper balance. Conclusion To conclude the literature review, a model o f life satisfaction among elderly people may be drawn from previous research findings, which will serve as the basis for this study. The model incorporates such dimensions as socio-demographic aspects, physical health, social support, network and activities, and psychological well-being and subjective conditions o f socio-demographic factors and support received (Figure 1). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. t F ig u re 1. M odel o f L ife Satisfaction S ocio-D em ographic A spects H ealth A ctivity Social S upport N etw o rk Psychological W ell-being P erceived C onditions 23 M ost previous researchers on the topic o f life satisfaction seem to exclusively focus on the aspects whose effects they want to show without including a variety o f possible intervening or interacting variables in their analysis. On the contrary, this paper intends to examine a full set o f factors affecting life satisfaction resulting in greater certainty that the effects identified are correctly isolated. It needs to be made clear that the reason for including this broad range o f factors in the analysis is not to identify a perfect explanation o f life satisfaction. It is rather to examine the isolated effect o f psychological factors and activity participation on life satisfaction. This study is similar to Palmore et al.'s (1985) study in the variety o f variables incorporated in the analysis. However, different from Palmore et al.'s study, this paper does not assume the hierarchical order in the effect o f variables on life satisfaction. Palmore et al. viewed demographic variables as the least proximate predictors o f adjustment to retirement, followed by socioeconomic status and health, and social activities and relationships as the most proximate predictors. However, this study is open to a strong and direct effect from any variables. This paper is unique in its focus on psychological factors and activity participation, and its implication. While many studies investigate these factors, no study has related the implications to potential development o f modules for retirement preparation programs. Four specific hypotheses are investigated in this paper. The first is that objective conditions o f health, income, marital status, availability o f children and close relatives/friends, R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 24 and employment are less likely than subjective assessment o f satisfaction with health, financial situation, marriage, relationship with children and relatives/friends, and employment to be related to life satisfaction. Even though more and more researchers are interested in examining the effect o f perceptions on outcomes and its comparison with the effect o f objective conditions on life satisfaction, there is no study done whose primary focus is to compare a series o f objective and subjective conditions. Thus, an episodic comparison o f the effect o f some subjective conditions with that o f objective conditions with factors such as health, social support, financial situation needs to be supported by a study which focuses on the comparison o f various objective and subjective conditions. The second hypothesis is in extension o f the first: psychological factors such as mastery, efficacy, depression, and happiness are likely to be highly related to life satisfaction and, when included, they cancel or weaken the effects o f other factors. I f psychological factors have a great influence on life satisfaction among the elderly, there should be some provisions to enhance psychological well-being in retirement preparation programs because a change in outcome can be made by modifying predictor characteristics.3 The third hypothesis is that leisure activity and group activity participation and volunteer w ork are likely to be positively related to life satisfaction. The effect o f these three activities is likely to be greater or less depending on employment status. For example, leisure 3 A bottom-up approach assumes that the only way to bring about a change in outcome is to modify predictor characteristics (Kozma et al., 1991). | R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 25 activity participation will be more positively related to employment among the employed than the unemployed because the employed tend to prefer informal activity. Volunteer w ork will be more positively related to life satisfaction among the unemployed than the employed because retired elderly tend to compensate for their role loss by participating in formal activity (Jahoda, 1982). The last hypothesis begins from the interest in the employed and their job satisfaction. The hypothesis is that job satisfaction is likely to be highly related to life satisfaction. Examination o f variation in job satisfaction seems important because the elderly who are working over 70 may have special reasons for working or attachment to their jobs. For example, those with higher job satisfaction may continue to w ork because they are satisfied with their jobs while those with lower job satisfaction may quit the job even with similar health problems to those o f a highly job satisfied group. Thus, those who (maybe, have to) stay in their job with lower job satisfaction would have lower life satisfaction. It would be also interesting to investigate the different effect o f various variables on life satisfaction when job satisfaction is included. The hypothesis is that high job satisfaction among the employed increases life satisfaction and may moderate the effect o f other variables on life satisfaction. While the effect o f satisfaction with employment as well as employment status is examined in the first hypothesis, job satisfaction in this hypothesis is examined to investigate the relationship between job satisfaction and life satisfaction among the employed elderly. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 26 M ETHODS Data and Subject Selection The analyses in this paper used the data from the M acArthur Research Network on Successful Aging Community Study, a longitudinal study o f three large, population-based persons age 70-79 in Durham, East Boston, and New Haven. Subjects were drawn from the Epidemiologic Study o f the Elderly (EPESE) at each site. In 1988, those between 70 and 79 (N=4,030) were screened on the basis o f six criteria o f physical and cognitive function to identify a high-functioning cohort. At screening, 1,313 subjects met the criteria for the higher functional status group and 1,192 (90.8%) among them agreed to participate in the study. Measures In this study, there are five major dimensions that may be related to life satisfaction among elderly people: socio-demographic factors, physical health status, psychological well being, socio-emotional factors, and subjective conditions. Each is discussed in detail below and more detailed models are presented in Figure 2 and Figure 3. Life Satisfaction: Two life satisfaction variables were used in this study. The first is a unidimensional indicator o f life satisfaction. The question is “how much satisfied are you with your life as a whole?” Respondents were asked to rate the degree o f their life satisfaction on “ T ' R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. I . 4 Figure 2. Model o f Unidimensional Life Satisfaction S ocio-D em o g rap hic A spects A ge E ducation incom e M arital Status E m ploym ent N u m b er o f F am ily /F rien d s Subjective Conditions S atisfaction w ith H ealth Satisfaction w ith Finances S atisfaction w ith M arriage S atisfaction w ith K ids R elationship S atisfaction w ith R elatives/FriV nds Ri S atisfaction w ith Job PhysicaL H ealjh / ^ H e a l t h change f S elf-rated H ealth I C h ro n ic C onditions \ P h y s i c a l Performance lionship Life Satisfaction Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Figure 3. Model of Multidimensional Life Satisfaction S o cio JS em o g rap h ic A spects X A g e N. W ell-being G en d er Psycl M astery E fficacy H appiness D epression E ducation Income R ace M arital Status E m ploym ent L ife Satisfaction P h y sicaL H eallh S tatu s ^—fiealtli C h a n g e '^ \ N Self-rated H ealth C hronic C onditions Physical P e rfo rm a n c e u m b er o f F am ily /F rien d s S upport T ak ing/G ivin 'Activity Participattoi Eormatxx^tnfornial A ctivity X A ctivity a scale o f one to five, with one being “completely satisfied” and five being “not at all satisfied.” Scoring was reversed, thus the higher the score, the greater the satisfaction. The second measure o f life satisfaction is a combined score based on multifaceted life satisfaction, which was measured by sum o f satisfaction with each specific domain such as health, marriage, financial situation, relationship with kids, and relationship with relatives and friends (M acArthur calculated scale o f life satisfaction). After scoring on items o f the five satisfaction variables was reversed, sum o f item scores was divided by number o f items answered and then multiplied by five. Thus, the higher the score, the greater the satisfaction. In order to check whether the combination o f these five variables into a single scale is proper, a Cronbach alpha coefficient was calculated. The alpha is .59 proving this life satisfaction scale to be reliable. The score ranges from 5 to 25. The higher the score, the greater the satisfaction. In order to test the first hypothesis, the effect o f each six satisfaction variable is compared with that o f the corresponding objective condition. That is, with the unidimensional life satisfaction as a dependent variable, the effect o f objective health (self- rated health, the number o f chronic conditions, health change, and physical performance score) and that o f satisfaction with health are compared. The same comparison is done between income and satisfaction with financial situation, employment and satisfaction with employment, marital status and satisfaction with marriage, the number o f kids and satisfaction with relationship with kids, and the number o f close relatives/friends and satisfaction with R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 30 relationship with them. Because o f the linear relationship o f marital status and satisfaction with marriage, and employment and satisfaction with employment, a reformulation o f these variables were made. Satisfaction with job and satisfaction with marriage originally had a five point scale, with one being completely satisfied and five being not at all satisfied. Since 52.5% o f the employed and 59.8% o f the married answered that they are completely satisfied with their employment and their marriages respecting, these variables are divided into two groups: the completely satisfied group indicating a more satisfied group and others a less satisfied group. Then, three dummy variables were created: the unmarried or the unemployed, the more satisfied with marriage or employment, and the less satisfied. In this analysis, the lower satisfaction group was used as a reference. Socio-demographic Factors: Age, gender, education, income, race, employment (in paid work), and marital status are measured in the analysis. Age in 1988 interview is used. Race is defined as white or black. Years o f education completed is used for education. Marital status is measured by whether a respondent is currently married or not. Even though the separated, the widowed, and the never-married may differ in some ways, this dichotomization o f marital status may indicate a lot more about other aspects such as family relationship, support available, emotional well-being, etc. in a more clear way. Income consists o f incomes from all sources such as wages, salaries, social security or retirement benefits, help from relatives, rent from property, and so on. It was measured by asking respondents to pick the closest category o f income among 13 income groups shown them. The categories are R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. presented in the appendix. Employment is measured by simply asking whether a respondent did any work for pay at the interview time. Considering respondents' age, those who are not employed at the present time due to temporary sickness or layoff were regarded as non employees (in an expanded sense, retirees) even though they may not be viewed as retirees in other analyses. Health Factors-. Variables used to measure physical health status are self-rated health, the number o f chronic conditions, health change, and physical performance score. Self-rated health is scored as a five point scale with values from bad to excellent. The number o f chronic conditions is used as another measure o f objective physical health status. This is a continuous variable ranging from 0 to 5. The included conditions are myocardial infarction, stroke, cancer, diabetes, hip fracture, broken bones, and high blood pressure. Health status change is measured by two dummy variables o f improved health and worse health, with no change as a reference. It was mentioned earlier in this paper that the use o f indicators o f functional status rather than number and/or severity o f specific diagnosed conditions has gained favor with researchers dealing with the health status o f the elderly. Functional health is measured by using a physical performance scale. MacArthur data set has a physical performance scale which is sum o f the five standardized domains: standardized balance score, score for chair stands, score for signature, score for timed taps, and score for measured walk. Given the fact that the sample are high functioning elderly people, physical performance score seems more R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. appropriate than indicators o f basic functioning measures such as the Katz' Index o f Activities o f Daily Living.4 The score range o f physical performance is from 0 to 5. The higher the score, the better performance. Psychological Factors-. Variables such as efficacy, personal mastery, depression, and happiness are used to measure psychological well-being o f elderly people. Even though these psychological factors may be related to each other and measure overlapping aspects o f psychological well-being, they may also capture different dimensions o f psychological well being. The M acArthur calculated variable o f efficacy is used to measure efficacy. Nine items were summed to make this index. Among them are: “keeping healthy depends on what I do”; and “it’s up to me to arrange transportation” (see appendix for all o f the nine items). The score ranges from 9 to 36 and the higher the score, the higher the efficacy. Five items about positive affect were summed to represent happiness (MacArthur calculated scale o f happiness): “life could be happier than now”; “these are the best years o f my life” (see appendix for all o f the five items). The range o f score is 3 to 20. The higher the score, the higher the happiness. 4 Katz' Index o f Activities o f Daily Living (Katz et al., 1963) provides easily rated criteria o f functional capacity by scaling the degree o f assistance needed in eating, bathing, dressing, toileting, transferring, and incontinence. In fact, 100% o f the sample eligible for this study have no difficulty in the Katz measure, thus Katz score is not usable in this study. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Personal mastery is measured with the sum o f seven items (MacArthur calculated scale o f mastery) such as: “I have little control over what happens to me”; and “what happens in the future depends on me” (see appendix for all of the seven items). The score range is 7 to 28. Tw o o f the items (“what happens depends on me” and “I can do anything I decide”) were reversed in their scoring and so the higher the score, the greater the personal mastery. Even though some researchers include depression as an indicator o f mental "health," depression in this study is used to represent psychological well-being. M acArthur calculated scale o f depression is used, which consists o f 11 items: “how much distressed by loss o f sex”; “how much distress by thoughts o f suicide”; and “how much distress by poor appetite” (see appendix for all o f the 11 items). The score range is 11 to 44. Higher score means higher depression. Socio-emotioncil Factors: Participation in leisure activity, volunteer activity, and group activity is measured to see the effect o f activities on life satisfaction among elderly people. Leisure activity, representing informal activity, was measured by asking whether respondents ever do leisure activities like taking walks, going swimming or golfing, or others such as these. It was coded as a dichotomous variable (yes/no). This question only asks about physical leisure activities, not non-physical leisure activities. Volunteer activity was assessed by asking whether respondents did volunteer work for a church or other religious organization, for a political group, a senior citizens' group, or for any other type o f organization during the last 12 months. It was also answered yes or no. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Group activity was measured by asking the frequency o f attending club meetings, religious service, and other activities with religious group. Six-point values were used in each item (from more than once a week (1) to never (6)). After the value o f each item was reverse, a calculated mean o f three items was multiplied by three. In tiiis analysis, only those who answered all o f the three items were used. The higher the score, the more frequently involved in group activity. The number o f kids, o f other close relatives, and of close friends, the level o f instrumental/emotional support, help provided to others, and attitude on networking are used to assess aspects o f social support and networking. The number o f kids, other relatives and friends to whom respondents feel close were asked directly and was coded as continuous variables ranging from 0 to 21, 0 to 60, and 0 to 96, respectively. Social support is measured as emotional and instrumental support received. Networking and social support are important not only because they make instrumental support available but because they are related to emotional support as well. Thus, in this study the sum o f instrumental support and emotional support was used to measure total social support. The M acArthur calculated variable o f total social support summed nine items: “how often spouse listens”; “how often you count on spouse” ; “how often spouse gives advice”; “how often kids listen your worries”; “how often you count on kids”; “how often kids give you advice”; “how often friends/relatives willing to listen”; “how often you count on R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 35 friends/relatives to shop”; and “how often friends/relatives give you advice.” Each item was asked on a five point scale o f never (1), rarely (2), sometimes (3), frequently (4), and no need (5). The MacArthur calculated variable o f total social support recoded values o f never and no need to 0, rarely as 1, sometimes as 2, and frequently as 3. Then, the sum o f these values was divided by number o f items answered. Score ranges from 0 to 3. The higher the score, the greater support received. To measure how often respondents help family or friends, five variables were combined (see appendix). Each question has a five point value like in social support questions. Recoding was done in the same as the social support variable. The score was calculated only for those who answered more than three items. A higher score means that respondents provide more support to their family or friends. For networking, six items on networking were combined: “how often you feel useful to your family/friends”; “how often you feel you have a definite role (place) in your family and among your friends”; “when you are with your family and friends how often you feel lonely”; “when you are with your family and friends, how often you feel like you don't belong”; “how often you wish you knew more people you could rely on to help you”; and “how often you wish you knew more people you could talk with about personal matters and problems, people who would make you feel that they care about you”. The values o f each item were never, rarely, sometimes, frequently, and no need. Recoding was done in the same way as for the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 36 above two variables o f social support received and giving. The values o f items were summed. The score was calculated only with those who answered more than four items. ANALYSIS The methods of analysis used for this study were multiple regression, t-test, chi- square, frequency and percent distribution, mean score, and Pearson's correlation coefficients. Multiple regression models were used to identify the effect o f the proposed variables on life satisfaction. First, OLS regression analysis was conducted to examine the relative effects o f subjective and objective conditions. Second, variables in each factor such as socio demographic, health, socio-emotional, and psychological indicators were regressed on life satisfaction, and some combinations o f factors were also regressed to see the effect o f psychological factors on life satisfaction. The suspected high correlations between some o f the socio-demographic factors and psychological factors were tested using Pearson's correlation coefficient and t-test. Then, another multiple regression was used to compare the effect o f different types o f activities and their interaction with employment on life satisfaction. Then, t-test and chi- square analyses were done to examine the mechanism which made leisure activity negatively affect life satisfaction, the potential outlier effect on the relationship between leisure activity participation and life satisfaction, and the reason for disappearance o f the effect o f volunteer work in the fully adjusted model. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 37 Finally, the effect o f job satisfaction on life satisfaction was examined with two different techniques: one a t-test and the other a regression analysis. RESULTS5 The baseline characteristics o f the subjects are presented in Table 1 and Table 2. Age ranged from 70 to 80 and the average age was 74.27 with a standard deviation o f 2.72. There were a few more female respondents (55.5%) than male ones. Whites (81.2%) were dominant in this sample. Only 18.8% o f the respondents were employed in paid work and the majority were not employed. A little less than half o f the subjects were married at the time of interview. The health status o f the sample shows that they are healthy. Over 70% o f the respondents rated their health excellent or good. Twenty six percent o f respondents did not report any chronic conditions and 65% o f the respondents reported one or two.6 The mean score o f physical performance (ranging from 0.485 to 3.861) is 2.76 with a standard deviation o f 0.51. About seventy eight percent o f the sample did not experience a health change in the 5 In this study, the sample w as not limited to those w ho answered all o f the questions used. D oing this could cause a significant reduction o f sample size and a possible systematic elimination o f certain kinds o f people (for example, those without a child). On the other hand, som e results also could be biased because they might reflect characteristics o f specific persons who answered certain questions. Since the results o f re-analysis with the sample w ho answered all the questions w ere the same as those in the original analysis, the original results are reported here (sample o f 1192). 6 In the MacArthur data set, which is high functioning sample based data, three hundred ten subjects (26.14%) did not have any chronic condition and nobody had more than five chronic conditions. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 38 Table 1. Characteristics of the Study Population: MacArthur Study of Successful Aging (1988) Variable Mean SD Percentage Range Socio-demographic factors Age 74.27 2.72 70-80 Gender (% female) 55.5 Education (years completed) 10.61 3.32 0-17 Income ($ annual) 15,312.85 R ace(%) White 81.2 Black 18.8 Employment (% in paid work) 18.8 Marital status (% currently married) 47.4 Health factors Self-rated health (%) Excellent 18.8 Good 54.4 Fair 25.7 Poor 1.1 Bad 0 Number of chronic conditions 1.18 0.96 0-5 Health change Improved 9.9 Worse 12.5 No change 77.7 Physical performance score 2.76 0.51 0.485-3.861 Psychological factors Personal mastery 18.95 2.29 11-28 Happiness 14.17 2.42 6-20 Depression score 14.17 3.24 11-31.90 Efficacy 26.29 2.45 18-36 Socio-emotional factors Number of children 2.38 2.17 0-21 Number of other close relatives 3.98 4.91 0-60 Number of close friends 5.79 8.88 0-96 Instrumental/Emotional support 2.04 0.53 0.5-3 Providing help to others 7.38 3.24 0-15 Network 20.98 2.82 7-24 Leisure activity participation (% yes) 81.8 Volunteer activity participation (% yes) 26.9 Group activity 8.87 3.97 3-18 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table 2. Mean and Percentage of Satisfaction Variables: MacArthur Study of Successful Aging (1988) Variable Mean SD Percentage Range Life satisfaction as a whole (%) Completely satisfied 30.2 Very satisfied 35.2 Somewhat satisfied 30.6 Not very satisfied 3.7 Not at all satisfied 0.3 Multifaceted life satisfaction 20.09 2.87 8.33 -25 Satisfaction with financial situation (%) Completely satisfied 24.1 Very satisfied 21.9 Somewhat satisfied 37.5 Not very satisfied 12.6 Not at all satisfied 3.9 Satisfaction with health (%) Completely satisfied 25.6 Very satisfied 33.1 Somewhat satisfied 36.5 Not very satisfied 3.8 Not at all satisfied 0.9 Satisfaction with marriage (%) Completely satisfied 59.8 Very satisfied 26.4 Somewhat satisfied 11.3 Not very satisfied 1.7 Not at all satisfied 0.9 Satisfaction with relationship with kids (%) Completely satisfied 59.4 Very satisfied 26.9 Somewhat satisfied 11.0 Not very satisfied 1.9 Not at all satisfied 0.7 Satisfaction with relationship with close relatives/friends (%) Completely satisfied 46.8 Very satisfied 38.0 Somewhat satisfied 14.4 Not very satisfied 0.7 Not at all satisfied 0.1 Satisfaction with employment (%) High satisfaction 52.2 Low satisfaction 47.8 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 40 past year while about 10% o f them improved their health and about 13% deteriorated. While the majority (81.8%) did leisure activity, only a quarter o f the respondents participated in volunteer activity during the last 12 months. The subjects had, on average, two to three children, four other close relatives, and six close friends. Total social support (instrumental and emotional support) and positive attitude on network have high mean scores. Particularly, the network score is concentrated in the range from 20 to 24 (72.4%). Table 2 describes the domain specific satisfactions and life satisfaction o f the study population. For the unidimensional life satisfaction question, at least 30 percent o f the sample perceived their life as completely, very, and somewhat satisfied, respectively and very few were less satisfied. The mean score o f the multifaceted life satisfaction scale was 20.09 with the maximum score o f 25. Specifically, satisfaction with financial situation showed a relatively smaller proportion o f subjects answering completely satisfied or very satisfied. On the other hand, satisfaction with relationships with others such as satisfaction with marriage and satisfaction with relationship with kids and relatives/friends showed greater proportion in "completely satisfied" category. In terms o f employment satisfaction, approximately half o f the respondents were completely (highly) satisfied with their employment. Over ninety percent o f respondents showed some degrees o f satisfaction in health. Table 3 a outlines what variables were used as objective and subjective conditions. Table 3 b shows the unstandardized parameter estimates o f life satisfaction models which compare the effects o f objective conditions and subjective conditions. In general, the R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 41 results confirmed the hypothesis that the effects o f the subjective assessment are much more important than those o f the objective conditions in predicting life satisfaction. In model 1, all o f the objective health variables except self-rated health did not show a significant effect on life satisfaction. When satisfaction with health was held constant in Table 3a. Objective Conditions and Subjective Conditions Health Employment Finances Marriage Children Objective Conditions Subjective Conditions Self-rated Health Number of Chronic Conditions Health Change (Worse, Improve, No Change) Physical Performance Employment Status (Employed or Unemployed) Income Marital Status (Married or Unmarried) Number of Children Satisfaction with Health Relative/Friends Number of Close Relatives Number of Close Friends Satisfaction with Employment Satisfaction with Financial Situation Satisfaction with Marriage Satisfaction with Relationship with Children Satisfaction with Relationship with Close Relatives/Friends ii'.i i ■ ■ ^ 1 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. : L _ „ Table 3b. Unslan dardized Parameter Estimates of Objective Conditions anil Subjective Conditions in Model of Lire Satisfaction Health Sell-rated health Number ordironic conditions Health change (no change as rererence) Improved Worse Physical peifomiance Salisraclion with health Model 1 .30** -.02 -.02 -.12 -.02 Model2 .10* -.00 .02 .07 -.03 .3 7 " Model3 Model4 Models Modelfi Model7 Model8 Model9 Model 10 Model 11 Model 12 Model 13 .081 .00 .04 .13 -.08 .29” Employment (lower salisraclion as rcrercuce) Employment (Unemployed) High Satisradion with employment -.18” .07 .47” .01 .08 rinances Income Satisfaction with financial situation .04” .00 .26” .00 .18” Marriage (lower satisradion as reference) Marital status (Unmarried) High satisfaction with marriage -.17” .10 .46” .07 Children Number ofkids Satisfadion with relationship with kids .01 .02 .36” .01 .21” Relatives/Friends Number o f close relatives Number o f close friends Satisradion with relationship with relalives/rriends .02” .01* .01” .00 .30” .01* .00 .08* R1 .07 .17 .01 .02 .01 .11 .01 .04 .00 .10 .02 .08 .34 +p<.10, *p< OJ, ” p<OI 43 model 2, the effect o f self-rated health on life satisfaction decreased. That is, satisfaction with health transmitted 70% o f the effect o f self-rated health on life satisfaction probably indicating that they are measuring the same concept, which is discussed later in this paper. In model 3, being unemployed was negatively related to life satisfaction. However, the effect o f employment status became insignificant by taking into account satisfaction with employment in model 4. Model 5 and model 6 show that what strongly determined life satisfaction was not income per se, but rather satisfaction with financial situation.7 Model 7 and model 8 show that when satisfaction with marriage was added in the model, whether a respondent is currently married did not predict his/her life satisfaction any more. Model 9 shows that the number o f children did not affect life satisfaction. Neither did it when satisfaction with relationship with them being controlled. Satisfaction with relationship with children was a significant predictor o f life satisfaction (model 10). A somewhat unexpected outcome is shown in model 11 and model 12. The number o f other close relatives and close friends had a significant effect on life satisfaction. When satisfaction with relationship with relatives/friends was considered, however, the effect o f the 7 However, correlation between income and satisfaction with financial situation is higher (.35) compared to that o f other pairs o f perceived and objective conditions. It implies that income may explain a large proportion o f satisfaction with financial situation compared to the relationships o f other pairs. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. s i - 44 number o f other close relatives remained significant while the significant effect o f the number o f close friends on life satisfaction was gone. That is, having more relatives was related to greater life satisfaction holding satisfaction with relationship with them constant in model 12 and model 13 as well. M odel 13 includes all the variables o f the objective and subjective conditions o f the previous models in Table 3b. The effects o f satisfaction with health remained significant while the effect o f satisfaction with employment disappeared. Also, the effects o f satisfaction with financial situation, satisfaction with marriage, satisfaction with relationship with children, and satisfaction with relationship with relatives/friends were weakened. Standardized parameter estimates o f the variables (Table 3c) reveal that the strongest effects on life satisfaction are, in order, from satisfaction with health (b=30), satisfaction with financial situation (b=.23), satisfaction with relationship with kids (b=.19), higher satisfaction with marriage (b=. 14), the number o f close relatives and self-rated health (b=.07), and satisfaction with relationship with relatives/friends (b=.06). In addition to these models, a fuller model is contained in Table 3d. They controlled for the other selected variables in life satisfaction model. While the effect on life satisfaction o f health satisfaction, satisfaction with financial situation, higher satisfaction with marriage, and satisfaction with relationship with children were decreased, they remained significant. The number o f close relatives and satisfaction with relationship with close relatives/friends, however, lost their significance. i I R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. L Table 3c. Standardized Parameter Estimates o f Objective Conditions and Subjective Conditions in Model o f Life Salisraclion Model 1 Model2 Model3 Model4 Models Model6 Model7 Models Model9 Model 10 Model 11 Model 12 Model 13 I leallh Self-rated health Number orclironic conditions Health change (no change as reference) Improved Worse Physical performance Satisfaction with health .24* -.03 -.01 -.05 -.01 .08* -.00 .01 .03 -.02 .38** Employment (lower satisfaction _____________as reference)_________ Employment (Unemployed) High Satisfaction with employment Finances -.08** .03 .16” Income Satisfaction with financial situation Marriage (lower satisfaction as _________ reference)__________________ Marital status (Unmarried) High satisfaction with marriage_______ Children___________________________ Number o f kids Satisfaction with relationship with kids Relatives/Friends .ll” .00 .33* -.10** .06 .24” .02 Number o f close relatives Number o f close friends Satisfaction with relationship with relatives/friends R .07 .17 .0 1 .02 .0 1 .11 .01 .04 .00 .03 .32” .10 . 11” .06* .02 .07 .00 .0 1 .05 -.05 .30” .01 .03 .01 .23** .04 .14” .02 .19** .08” .07* .04 .02 .26” .06* .08 .34 +p<.!0, *p<,05, ” p<.01 46 Table 3d. Unstandardized Parameter Estimates of Objective and Subjective Conditions and Selected _________ Variables in Model of Life Satisfaction________________________________________ Model 1 _________Model 2 Health Self-rated health .08+ .06 Number of chronic conditions .00 .01 Health change (no change as reference) Improved .04 .05 Worse .13 .17* Physical performance -.08 -.07 Satisfaction with health .29** .23** Employment Employment (Unemployed) .01 .08 (less satisfaction High satisfaction with employment .08 .09 as reference) Finances Income .00 .01 Satisfaction with Financial situation .18** .16** Marriage Marital status (Unmarried) .07 .07 (less satisfaction High satisfaction with marriage .27** .21** as reference) Children Number of kids .01 -.00 Satisfaction with relationship with kids .21** .18** Relatives/Friends Number of close relatives .01* .01 Number of close friends .00 -.00 Satisfaction with relationship with .08* .03 relatives/friends Age .00 Education -.00 Race (White) -.08 Gender (Female) -.00 Leisure activity -.08 Group activity .01 Volunteer work -.05 Providing help to others .00 Network .01 Instrumental/Emotional support .04 Mastery -.01 Happiness .06** Depression -.05** Efficacy .01 JT_________________________________________________________ .34_____________ .39 + p<.10, *p<.05, **p<.01 "T “ R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 47 In order to examine the effect o f psychological factors, Table 4a and Table 4b present the unstandardized and standardized parameter estimates o f the life satisfaction models including the effects o f the socio-demographic, health, and socio-emotional factors. Model 1 shows the effect o f socio-demographic factors on life satisfaction. The effect o f income was distinguishable and there was a moderate effect o f employment on life satisfaction. Self-rated health and health change (worse) were the only variables that significantly affected life satisfaction among the health variables (model 2). The number o f kids and friends did not influence life satisfaction while the number o f other close relatives did (model 3). Giving and taking support, and positive feelings about network had a significant effect on life satisfaction. While doing volunteer work was related to life satisfaction, group activity and leisure activity participation were not. Model 4 shows that all o f the four psychological factors-personal mastery, efficacy, happiness, and depression-w ere highly related to life satisfaction. Before developing confidence in each psychological factor's individual effect on life satisfaction, multicollinearity o f these four psychological factors was suspected. Thus, their tolerances were checked.8 The tolerance o f each psychological variable was as follows: mastery (.76), efficacy (.81), 8 Tolerance is a statistic to check multicollinearity o f variables in a model (tol=l-R 2 ). The logic is that mastery is regressed as a dependent variable predicted by efficacy, happiness, and depression. The same procedures are done for efficacy, happiness, and depression. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4a. Unstandardized Parameter Estimates of Psychological Fuctors and Oilier Selected Variables in Model of Life Satisfaction Model 1 Mode!2 Model3 Model4 ModelS Model6 Mode|7 Model8 Age .04 .04 .04 .06* .05 + Education -.03 -.05 + -.02 -.07** -.06+ Ruce (While) .31 .22 .52+ .41* .47+ Income .19+* .16+* .16+* .13** .12++ Marital status (Married) .32 .26 .14 .25 .11 Gender (Female) -.12 -.06 -.50+ -.14 -.37+ Employment (Employed) .40 + .16 .24 .13 -.02 Self-rated health I.25+* 1.21++ .77+* Number of chronic conditions -.09 -.08 -.07 Health change (no change as reference) Improved -.41 -.14 -.17 Worse -1.05++ -1.I5+* -.67** Physical performance .23 .08 .10 Number o f kids .02 .06 + .03 Number of other relatives .03 + .02 .01 Number of friends .00 -.00 -.00 Providing help to others -.06+ -.07+ -.06* Network .33++ .29*+ .16++ Instrumental/Emotional support .77*+ .84** .55++ Leisure activity .29 .02 -.15 Group activity .01 .04 .02 Volunteer work .60*+ .37 + .16 Mustery .16*+ .12++ .05 Efficacy .17+* .17*+ .13+* Happiness .28** .27+* .24++ Depression -.19++ -.18++ -.11++ R2 .05 .14 .15 .26 .18 .18 .28 .36 + p < .1 0 , +p<.05, ++p<.01 Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4b. Standardized Parameter Estimates of Psychological Factors and Qlher Selected Variables in Model of Life Salisfaction Model! Model2 Model3 Mode!4 Model5 Model6 Model7 Model8 Age .04 .04 .04 .05* .05 + Education -.04 -.06 + -.03 -.09++ -.07* Race (White) .04 .03 .07+ .06+ .07+ Income .18** .16+* .16*+ .12** .11*+ Marital status (Married) .06 .05 .02 .04 .02 Gender (Female) -.02 -.01 -.09* -.02 -.07+ Employment (Employed) .06 + * .02 .03 .02 -.00 Self-rated health .30** .30++ .19++ Number of chronic conditions -.03 -.03 -.03 Health change (no change as reference) Improved -.40 -.01 -.02 Worse -. 12*+ -.13++ -.08++ Physical performance .04 .02 .02 Number of kids .02 .05 + .03 Number of other relatives .05 + .04 .02 Number of friends .00 -.01 -.01 Providing help to others -.07* -.07+ -.06* Network .31** .28+* .16++ Instrumental/Emotional support .14++ .16++ .10+* Leisure activity .04 .00 -.02 Group activity .01 .05 .03 Volunteer work .09+ .06 + .03 Mastery .13++ .10+* .04 Efficacy .15*+ .15++ .11++ Happiness .23++ .23+* .20++ Depression -.21++ -.21++ -.12*+ R3 .05 .14 .15 .26 .18 .18 .28 .36 + p < . 10, +p< .05, ++p<.01 50 happiness (.77), and depression (.85). The results show that only 24% o f the variation in mastery was explained by other three variables. Nineteen percent, 23%, and 15% o f the variation o f efficacy, happiness, and depression, in respect, were explained by three other factors. Thus, the hypothesis o f multicollinearity o f these four psychological variables was rejected. Model 8 shows interesting results. With all variables together, age, education, race, and gender became significant in their effect on life satisfaction. In order to examine what factors drove this change, some combinations o f factors were regressed in model 5, model 6, and model 7. In model 5 in which health factors were added to socio-demographic factors, age, education, race, and gender were constantly insignificant or became moderately significant in their effect on life satisfaction. By adding socio-emotional factors in the model (model 6), the effect o f race and gender became significant. In model 7, psychological factors were added into model 1. The effect o f age, education, and race became significant. Therefore, age, education, and race were suspected o f being related to psychological factors while gender (also race) seemed to be related to socio-emotional factors. The simple correlation and t-test o f these variables with each psychological factor showed that they were closely related: age was related to efficacy and happiness; education w as to efficacy, happiness, and mastery; race was to mastery and depression. Close associations confirmed by t-test between volunteer work and mastery, happiness, and efficacy, and also between volunteer work and some o f the socio-demographic w _Jl R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 51 factors explain why the significant effect o f volunteer work on life satisfaction disappeared in model 8. The overall fit o f the model, in terms o f explained variance, is fine. The total R2 for the fully adjusted model (model 8) is 36%. Preceding the examination o f the effect o f activity and its interaction with employment on life satisfaction is a brief description o f a group o f the employed and the non employed. Simple means and percent distributions were calculated first and t-tests and chi- square analyses were followed to see a statistical difference, if any, in some o f the socio demographic factors among these two groups o f respondents. As expected, the employed group was younger, received more education, and had more income. The mean age o f the non-employed people was 74.37 (SD=2.72) while that o f the employed was 73.82 (SD=2.66). While the income difference between these two groups was statistically significant, the level o f education was not significantly different. About 50% o f non-employed had annual incomes less than $10,000 and the counterpart in the employed group was only 30%. On the other hand, about 19% o f the employed had annual incomes over $30,000 whereas only 10% o f the non-employed earned more than $30,000 a year. Race difference between the two groups was statistically significant as well. About 83% o f white people were currently unemployed while about 72% o f black people were currently unemployed. Gender difference was also noticed. M ore females than males were unemployed. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 52 Table 5a. Unstandardized Parameter Estimates o f Activity Participation and Other Selected Model 1 Model 2 Model 3 Model 4 Age .05+ Education -.06* Race (White) .48* Income .12** Marital status (Married) .11 Gender (Female) -.38* Self-rated health .75** Number of chronic conditions -.07 Health change (no change as reference) Improved -.18 Worse -.65** Physical performance -.10 Number of kids .03 Number of other close relatives .01 Number of close friends -.00 Providing help to others -.06* Network .16** Instrumental/Emotional support .55** Master}' .05 Efficacy .13** Happiness .24** Depression -.11** Employment 2.34* 1.51 Leisure activity .14 .12 .31 -.01 Leisure activity*Emp!oyment .17 -.72 -.67 Group activity .03 .03 .03 .02 Group activity*Employment .00 -.10 .00 Volunteer work .52* .48* .55* .21 Volunteer work*Employmcnl .13 -.21 -.28 R- .01 .02 .02 .37 +p< 10, *p<.05, **p<.01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 53 Table 5b. Standardized Parameter Estimates o f Activity Participation and Other Selected Model 1 Model 2 Model 3 Model 4 Age .05+ Education -.07* Race (White) .07* Income .11** Marital status (Married) .02 Gender (Female) -.07* Self-rated health .18** Number of chronic conditions -.03 Health change (no change as reference) Improved -.02 Worse -.08** Physical performance -.02 Number of kids .03 Number of other close relatives .02 Number of close friends -.01 Providing help to others -.06* Network .16** Instrumental/Emotional support .10** Mastery .04 Efficacy .11** Happiness .20** Depression -.12** Employment .32* .21 Leisure activity .02 .02 .04 -.00 Leisure aclivity*Employmcnt .04 -.18 -.17 Group activity .04 .04 .05 .03 Group activily*Employincnt .01 -.03 .01 Volunteer work .08* .07* .08* .03 Volunteer work*Employmcnt .02 -.04 -.05 R2 .01 .02 .02 .37 +p<.10, *p<05, **p<.01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. r- 54 In Table 5a and Table 5b, the effects o f activity participation and its interaction with employment on life satisfaction were examined. The hypothesis that activity is likely to be positively related to life satisfaction and the effect o f leisure, group, and volunteer activity is likely to be greater or less depending on employment status was not confirmed. Model 4 in Table 5a shows that there was no significant effect o f activity participation or its interaction with employment on life satisfaction, no matter what the type o f activity was. Volunteer work, assumed to be formal activity, had some effect on life satisfaction in model 1 to model 3. Given the majority o f the respondents are unemployed, the significant positive relationship between formal activity participation and life satisfaction was expected because the non-employed elderly tend to pursue a "work-type" formal activity. In terms o f informal type o f activity participation, on the other hand, leisure activity participation was not a predictor o f life satisfaction and had no interaction effect with employment on life satisfaction. However, the relationship o f volunteer work with life satisfaction disappeared when other factors were considered, and leisure activity was always insignificant in its effect on life satisfaction (model 4 in Table 5a). Thus, neither formal nor informal activity participation influenced life satisfaction. In addition, the similar parameter estimates among activity variables (model 4 in Table 5b) may imply that the type o f activity is not important in its effect on life satisfaction o f the elderly. N or did activity participation have an effect on life satisfaction via its interaction with employment. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 55 The parameter estimates o f the three activity variables in model 4 o f Table 5a are similar to those in model 8 o f Table 4a. Also the disappearance o f the significant effect of volunteer work from model 1, model 2, and model 3 to model 4 in Table 5a is similar to the effect disappearance from model 3 to model 6 to model 8 in Table 4a. Leisure activity in the full model showed the negative relationship with life satisfaction, which was unexpected (model 8 in Table 4a, model 4 in Table 5a, and the model in Table 7).9 Particularly this negative impact o f leisure activity participation on life satisfaction was significant in Table 7. So the possible association between leisure activity and other variables was examined by chi- square and t-test. The findings showed that gender, race, worse health, self-rated health, income, education, giving help to others, network, mastery, and happiness were all highly associated with leisure activity. Whites and males were more likely to participate in leisure activity than blacks and females. Higher education, related to more income, was highly associated with leisure activity participation. Thus, when these variables were taken into account in the model, the effect o f leisure activity became negative even though it continued to be insignificant in its effect on life satisfaction through most o f the models. In addition, interestingly, whether elderly people are employed or not was not related to leisure activity participation, but related to volunteer work, which will be mentioned below. 9 By checking residual chi-square, five outliers were found (reschi <.015). However, the frill model examined without the outliers showed even stronger negative relationship between life satisfaction and leisure activity participation. Thus, outliers do not explain why participation in leisure activity was related to lower life satisfaction. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 56 The hypothesis o f relatively greater importance o f formal activity (volunteer work) and its interaction with employment than informal activity (leisure activity) and its interaction with employment was not confirmed (Table 5b). The standardized estimate o f leisure activity on life satisfaction was -.00 while that o f volunteer w ork was .03 in model 4, both being insignificant. While parameter estimates o f group activity were somewhat similar across models, the significance level o f volunteer work, even though the direction o f effect was not changed as in the case o f leisure activity, was weakened and disappeared from model 3 to model 6 and from model 6 to model 8 in Table 4. The same suspected association between volunteer work and other variables led to chi-square and t-test analysis to examine it. Among those variables related to volunteer work are race, employment status, self-rated health, income, education, physical performance score, the number o f close friends, providing help to other, mastery, efficacy, and happiness. Gender was not associated with volunteer work. Whereas leisure activity was not related to employment status, employed people were more likely to participate in volunteer work than the unemployed, which is somewhat surprising. It may be possible if volunteering is through job-related organizations like unions. The last analysis is to see whether life satisfaction differs among older workers with higher and lower job satisfaction. Table 6 proves difference o f life satisfaction between the more satisfied group and the less satisfied group with employment. The t value was -4.10 (p<.01), which means that there is a significant difference between these tw o groups in life R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table 6. Life Satisfaction o f the Employed with Higher Job Satisfaction and Lower Job Satisfaction Employed with high satisfaction with job Employed with low satisfaction with job Number Mean SD T 115 109 21.33 19.83 2.35 -4.10** 3.07 **p<.01 Table 7. Unstandardized Parameter Estimates of Selected Variables in Life Satisfaction by Job Satisfaction (n=224) Model 1 Model 2 Age .10 .10 Education -.10+ -.10+ Race (White) .58 .49 Income .18* .21** Marital status (Currently married) .16 .04 Gender (Female) -.48 -.57 Self-rated health .42 .26 Number of chronic conditions -.11 -.11 Health change (no change as reference) Improved -.85 -1.07+ Worse -1.07+ -1.18* Physical performance -.68+ -.63 Number of kids -.03 -.01 Number of other close relatives .04 .05 Number of close friends .01 -.00 Providing help to others -.17** . 19** Network .13+ .11 Instrumental/Emotional support .70+ .77* Mastciy .07 .06 Efficacy .16* .17* Happiness .29** .27** Depression -.16* -.15* Leisure activity -.65 -.74+ Group activity .02 .02 Volunteer work -.04 -.11 Job Satisfaction (Higher Satisfaction) 1.21** R3 .46 .50 + p <.10, *p<05, **p<.01 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 58 satisfaction. The mean life satisfaction score o f higher job satisfaction group was 21.33, with a standard deviation o f 2.35 while the mean o f lower job satisfaction group was 19.83, with a standard deviation o f 3.07. Thus it was confirmed that those employed with higher job satisfaction tend to have greater life satisfaction than those with lower job satisfaction. Table 7 presents the effect o f job satisfaction on life satisfaction. It also proved that higher job satisfaction predicted greater life satisfaction. However, the hypothesis that job satisfaction would moderate the effect o f other variables on life satisfaction was rejected. When job satisfaction was added in the model 2, several variables such as income, worse health, instrumental/emotional support, and leisure activity rather increased their effect on life satisfaction. In particular, the fact that the degree o f the effect o f income and worse health on life satisfaction became stronger when job satisfaction was added in the model led to investigate the relationship o f income and worse health with job satisfaction. Simple correlation and chi-square test revealed that job satisfaction was not significantly associated with either income or worse health, however. DISCUSSION The intent o f this study was to explore what aspects of elderly people and their life affect their life satisfaction in order to draw some implications for retirement education programs which try to help adaptation to changes after retirement and thus to promote well being o f the elderly people. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Some o f the findings in this study deserve in-depth discussion. First, the analysis showed greater influence o f subjective conditions on life satisfaction compared to that of objective conditions in general. Thus, most of the effects o f objective conditions may be somewhat "indirect" through subjective conditions. However, the objective condition like worse health was also an important predictor o f life satisfaction (Table 3d). As discussed earlier in this paper, it was confirmed in this study that health status change is more important in life satisfaction o f the elderly than chronic diseases or illnesses. This paper shows that worse health had a significant impact on life satisfaction (Table 3c), which may be partly because that health change in this analysis was based on self-report, which is related to perception on health. Second discussion draws heavily from Table 4a. Self-rated health and worsening health significantly affected life satisfaction. Related to the first hypothesis, it may be partly due to the fact that self-rated health and self-rated health change are probably related to the perceptual aspect o f health status more than to other health variables which are more technical and objective health conditions. Third discussion is regarding the second hypothesis which tries to examine the effect o f psychological factors on life satisfaction. All o f the psychological factors were significantly related to life satisfaction in model 4 and model 8 (in Table 4a) except mastery, proving the significant relationship between psychological factors and life satisfaction. However, the hypothesis that the effect o f other factors such as, for example, providing help to others, R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 60 attitude on network, and instrumental/emotional support, would not be, in fact, real ones, was not confirmed. On the other hand, volunteer work became insignificant in its effect on life satisfaction when psychological factors were considered. However, because its effect was already weakened by socio-demographic factors (model 6), the second hypothesis that the effect o f psychological factors would be so dominant as to cancel out the effects o f other factors was not confirmed at all in this study. Thus, just as retirement preparation programs should not solely focus on financial and health management, psychological factors should be incorporated in the curriculum o f the retirement education in balance with other aspects o f life. Fourth, the hypothesis o f the positive effect o f activity participation and its interaction with employment on life satisfaction was not confirmed, either. Leisure activity, which is regarded as informal activity, was not significant at all in its effect on life satisfaction even in junction with employment status while group activity and volunteer work participation had some degrees o f bivariate relationship with life satisfaction. In particular, participation in volunteer work was differentiated by employment status. However, different from the hypothesis, the effect o f volunteer work as formal activity did not differ by employment status in the fully adjusted model. N or was informal activity (leisure activity) differentiated in its effect on life satisfaction by employment status. It seems to be mainly due to the preponderance o f the non-employed people compared to the employed people in the sample. However, it may also show that types o f activity and employment status are not associated R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. I 61 with each other. Furthermore, some elderly people’s tendency to pursue formal activity and others' tendency to pursue non-work-like activity after a long years o f working life may cancel out each other's effect. The above findings seem unsupportive o f the activity theory which argues that greater involvement in activities in later life increases well-being o f elderly people. The findings may partly support continuity theory, which implies that what makes elderly people satisfied with their life is not a sudden activity participation or an increase in level o f participation, but a continuation o f activity participation and the participation level. However, this interpretation should be made with caution because there is no information on how long respondents have participated in activities, nor how satisfied they are with activities they participate in. Fifth, job satisfaction was an important predictor o f life satisfaction among the employed elderly. It was expected because employment over 70 years old would be either exhaustively "by choice" or "by necessity." Thus, employment would determine a large part o f their life and correspondingly, their life satisfaction. Finally, in model 13 o f Table 3b, the number o f children was constantly insignificant in effect on life satisfaction. This was true whether satisfaction with relationship with them was considered or not. The number o f close relatives was still significant even when satisfaction with them was controlled in effect on life satisfaction. On the other hand, the effect o f the number o f close friends became insignificant when satisfaction with them was taken into account. Also, the standardized parameter estimate o f satisfaction with kids R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 62 relationship was greater than that o f satisfaction with relatives/friends relationship. Standardized estimate o f the number o f close relatives was greater than that o f the number o f close friends. These results did not confirm previous findings that friendship is a more important predictor o f life satisfaction than family relationships (Goudy & Goudeau, 1981; Larson et al., 1986; O'Connor, 1995). It should be noticed, however, that this study did not consider the effect o f co-residence with children or relatives on life satisfaction. Therefore, the finding about the effect o f the number o f children and relatives has a limitation to compare with previous studies on relationship o f life satisfaction with family relationship. Study Strengths and Weaknesses The strengths o f this study are: (1) a sample drawn from a community-based population, thus more generalizable, (2) age range o f the sample (70-80) which minimizes the effect o f some dramatic changes, which usually take place in the 60's, in interpreting the results, and (3) the availability o f a diverse set o f psycho-social factors. However, there are several limitations in the study findings. First, the generalization o f the findings to the general elderly population including those with lower-functioning population is questionable. For the group o f people who are relatively unhealthy or have lower functioning, health factors may have more important influences on life satisfaction, and the relationship between psychological well-being and life satisfaction and between activity R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 63 participation and life satisfaction may differ for this group. However, this problem was somewhat compensated for by including a measure o f health change in the study. Second, questions on activity are somewhat ambiguous. The questions, “Do you ‘ever do’ physical leisure activities such as...,” “During the last 12 months, ‘did you do’ volunteer work for...,” and the sum o f “how ‘often do’ you attend such group activities,” make the comparison o f different activities hard. Also, it is not certain whether respondents participate in activities on a regular basis or in an episodic way. Regular activity participation and sporadic participation would definitely make a lot o f differences in effect on life satisfaction. IM PLICATIONS There are several suggestions for developing retirement preparation programs that can be based on the study findings. First, retirement financial planning should not be overemphasized in the curriculum o f retirement preparation programs if life satisfaction is a desirable outcome. M ost existing retirement planning programs focus exclusively on financial planning. However, the findings show relatively less importance o f income in life satisfaction compared to that o f satisfaction with financial situation. Thus, exclusive focus on financial management after retirement in the program would be only partially helpful for retirees’ adjustment to retirement. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. r- 64 Second, a somewhat harder task to be accomplished by retirement education program is to improve psychological well-being among potential retirees. Even though psychological factors such as personal mastery, efficacy, depression, and happiness are much more difficult to change compared to, for example, habits o f food intake, incorporation o f a psychological component in programs is highly recommended given the dominant influence o f psychological factors on life satisfaction. Third, relationships with other people, particularly with family members seem important in elderly persons’ high levels o f life satisfaction. Counseling on relationships with people could be included in retirement preparation programs. Also it could be helpful to include family members and possibly friends in some sessions o f the program. There are also four suggestions for future research on this topic. First, questions on participation in leisure activity and volunteer w ork should have more detail and depth to have practical implications for developing retirement education programs. Future study on the impact o f activity participation on life satisfaction should include the effect o f the frequency o f activity participation on life satisfaction as well. Second, a qualitative study on why elderly people do not do various activities would be able to give some guidelines on why and how activities should be incorporated in retirement education programs. Third, consideration o f personality would be worthwhile, particularly in clarifying the role o f psychological variables. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Fourth, as health change was found to be an important predictor o f life satisfaction among the elderly, it is recommended that a future study with a similar topic be done with longitudinal data. Longitudinal studies to assess the effect o f actual changes (rather than subjective evaluation o f health change) in health conditions and other transitional conditions would capture more explanatoiy power in a life satisfaction study. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 66 BIBLIOGRAPHY Aquino, J. et al. (1996). Employment status, social support, and life satisfaction among the elderly. Journal o f Counseling Psychology. 43(4), 480-489. Atchley, R. (1989). A continuity theory o f normal aging. Gerontologist. 29(2). 183-190. Burgess E. W. (1954). Social relations, activities, and personal adjustment. American Journal o f Sociology. 59, 352-360. Campbell, A. et al., (19761. The Quality o f American Life: Perceptions. Evaluations, and Satisfactions. New York: Russel Sage Foundation. Carstensen, L. (1992). Social and emotional patterns in adulthood: Support for socioemotional selectivity theory. Psychology and Aging. 7(3). 331-338. Davis, K. (1991). Life satisfaction and Older Adults. The Commonwealth Fund: Americans over 55 at W ork Program. Dorfman, L. (1995). Health conditions and perceived quality o f life in retirement. Health & Social W ork. 20(3). 192-199. Ferris, C. & Bramston, P. (1994). Quality o f life in the elderly: A contribution to its understanding. Australian Journal on Ageing. 13(3). 120-123. George, L. & Clipp, E. (1991). Subjective components o f aging well. Generations. 15, 57-60. Floyd, F. et al. (1992). Assessing retirement satisfaction and perceptions o f retirement experiences. Psychology and Aging. 7(4). 609-621. Goudy, W. & Goudeau, J. (1981). Social ties and life satisfaction o f older persons: Another evaluation. Journal o f Gerontological Social Work. 4(1). 35-49. Higginbottom, S. et al. (1993). Linking retirement experiences and marital satisfaction: A mediational model. Psychology and Aging. 8(4). 508-516. Iwatsubo, U. et al. (1996). Predictors o f life satisfaction amongst retired people in Paris. International Journal o f Epidemiology. 25(1). 160-170. I R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 67 Jahoda, M. (1982). Employment and Unemployment: A Social-psychological Analysis. New York: Cambridge University Press. Katz, S. et al. (1963). Studies o f illness in the aged: The index o f ADL, standardized measure o f psychological and psychosocial functions. Journal o f American Medical Association. 185.914-919. Kozma, A. et al. (1991). Psychological Well-being in Later Life. Toronto and Vancouver: Buutterworths. Krause, N. (1987). Satisfaction with social support and self-rated health in older adults. Gerontologist. 27(3). 301-308. Krause, N. (1993). Race differences in life satisfaction among aged men and women. Journal o f Gerontology. 48(5). S235-S244. Krause, N. et al. (1992). Providing support to others and well-being in later life. Journal o f Gerontology. 47(5). 300-311. M cConatha, J. (1993). Maintaining well-being after retirement. InN . Julius & H. Krauss (Eds.), The Aging W ork Force: A Guide for Higher Education Administrators. Washington, DC: College and University Personnel Association. McIntosh, B. & Danigelis, N. (1995). Race, gender, and the relevance o f productive activity for elders' affect. Journal o f Gerontology. 50B(4). S229-S239. M enec, V. & Chipperfield, J. (1997). Remaining active in later life. Journal o f Aging and Health. 9(1). 105-125. Neugarten, B. et al. (1961). The measurement o f life satisfaction. Journal o f Gerontology. 16(2). 134-143. Neugarten, B. et al. (1968). Personality and patterns o f aging. Middle Age and Aging: A Reader in Social Psychology. Chicago: University o f Chicago Press. O'Connor, B. (1995). Family and friend relationships among older and younger adults: Interaction motivation, mood, and quality. International Journal o f Aging and Human Development. 40(1). 9-25. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 68 Palmore, E. et al. (1985). Retirement: Causes and Consequences. New York: Springer Publishing Company. Passuth, P. & Bengtson, V. (1988). Sociological theories o f aging: Current perspectives and future directions. In J. Birren & B. Bengtson (Eds.), Emergent Theories o f Aging. New York: Springer Publishing Company. Reis, M & Gold, D. (1993). Retirement, personality, and life satisfaction: A review and two models. Journal o f Applied Gerontology. 12(2). 261-282. Szinovacz, M. & Washo, C. (1992). Gender differences in exposure to life events and adaptation to retirement. Journal o f Gerontology. 47f4). S191-S196. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 69 APPENDIX Variables Socio-Demographic Factors Age: Age in years Education: Years completed Race: White Black Income: Total income o f a respondent (and his/her spouse) for the past month/year from all sources such as wages, salaries, social security or retirement benefits, help from relatives, rent from property, and so forth. ( Monthly Yearly A. 0 - $166 A. 0-$ 1,999 B. $167-$416 B. $2,000-$4,999 C. $417-5583 C. $5,000-56,999 D. $584-$833 D. $7,000-59,999 E. $834-51,249 E. $10,000-514,999 F .$ 1,250-$ 1,666 F .$ 15,000-$ 19,999 G. $1,667-52,083 G. $20,000-524,999 H. $2,084-$2,499 H. $25,000-29,999 I. $2,500-52,916 I. $30,000-534,999 J. $2,917-53,333 J. $35,000-539,999 K. $3,334-$3,749 K. $40,000-$44,999 L. $3,750-54,166 L. $45,000-$49,999 M. $4,167 or more M. $50,000 or more ) Marital Status: Married Separated Widowed Never married Gender: Male Female R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 70 Paid Employment: Any work for pay at the present time (Yes/No) Health Factors Self-rated Health: Respondent's health rating at the present time ( Excellent Good Fair Poor Bad ) Chronic Conditions: Number o f chronic conditions such as myocardial infarction, stroke, cancer, diabetes, hip fracture, broken bones, and high blood pressure. Health Change: A change in health over the past year ( Improved, gotten better Gotten, worse Stayed the same ) Physical Performance: 1. Balance score 2. Score for chair stands 3. Score for signature 4. Score for timed laps 5. Score for measured walk Socio-Eniotionnl Factors Children (Kids): Number o f presently living children (including adopted children or children you have raised) Relatives: Number o f other close relatives (people a respondent feels at ease with, can talk to about private matters, and can call on for help) Friends: Number o f close friends (people a respondent feels at ease with, can talk to about private matters, and can call on for help). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 71 Help Providing: During the past year, how frequently have you done each o f the following things for family or friends? 1. H ow often have you helped with minor household tasks or repairs? 2. H ow often have you given them advice or information? 3. H ow often have you loaned them money? 4. H ow often have you discussed their worries about their personal problems (for example, about their family, work, or health)? 5. H ow often have you tried to make them feel good about themselves (cheered them up, praised them)? ( Never Rarely Sometimes Frequently N o need ) Network: 1. H ow often do you feel useful to your family and friends? 2. H ow often do you feel you have a definite role (place) in your family and among your friends? 3. When you are with your family and friends how often do you feel lonely? 4. When you are with your family and friends, how often do you feel like you don't belong? 5. Thinking about all the people you know, how often do you wish you knew more people you could rely on to help you with things like work around the house, giving you a ride or other daily tasks? 6. Thinking about all the people you know, how often do you wish you knew more people you could talk with about personal matters and problems, people who would make you feel that they care about you? ( Never Rarely Sometimes Frequently ) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 72 Instrumental/Emotional support: 1. H ow often is your (husband/wife) willing to listen when you need to talk about your worries or problems? 2. H ow often can you count on your spouse to help with daily tasks like shopping, giving you a ride or helping you with household tasks? 3. H ow often does your spouse give you advice or information about medical, financial, or family problems? 4. H ow often (is/are) your child(ren) willing to listen when you need to talk about your worries or problems? 5. H ow often can you count on your child(ren) to help with daily tasks like shopping, giving you a ride or helping you with household tasks? 6. H ow often (does/do) your child(ren) give you advice or information about medical, financial, or family problems? 7. H ow often are your close friends and relatives willing to listen when you need to talk about your worries or problems? 8. H ow often can you count on your close friends and relatives to help with daily tasks like shopping, giving you a ride or helping you with household tasks? 9. H ow often do your close friends and relatives give you advice or information about medical, financial, or family problems? ( Never Rarely Sometimes Frequently ) Leisure Activities: Do you ever do any o f physical activities (like taking walks, going swimming or golfing) or others such as these? (Yes/No) Group Activities: 1. H ow often do you attend meetings o f clubs or organizations such as school groups, unions, fraternal organizations, athletic groups or the like? 2. H ow often do you usually attend religious services? 3. Besides regular services, how often do you take part in other activities with your religious group? R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 73 ( M ore than a week Once a week two to three times a month About once a month Less than once a month Never ) Volunteer Works: During the last 12 months, did you do volunteer work for a church or other religious organization, for a political group, a senior citizens' group, or for any other type o f organization? (Yes/No) Psychological Factors Efficacy: Think o f how things have been going for you and tell me how strongly you agree or disagree with each o f the statements. 1. Keeping healthy depends on things that I can do. How strongly do you agree or disagree with this statement? 2. It's up to me to arrange transportation when I want it. 3 .1 am able to get what I want from my relationships with my family. I am able to make sure that my relationships with my family are as satisfying and rewarding as I would like. 4 . 1 could make my financial situation better if I wanted to. 5. There are things I could do to make myself feel safer. 6 . 1 am able to get what I want from my relationships with my friends. I am able to make sure that my relationships with my friends are as satisfying and rewarding as I would like. 7 .1 do not have enough control over how good my living arrangements are. 8 .1 cannot be as productive as I want to. 9. (ask only if married) I have been able to get what I want from my relationship with my husband/wife. I am able to make sure that my relationships with my husband/wife are as satisfying and rewarding as I would like. ( Strongly agree Agree Disagree Strongly disagree ) R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Personal Mastery: 1 .1 have little control over the things that happen to me. 2. What happens to me in the future mostly depends on me. 3. There is really no way I can solve some o f the problems I have. 4. There is little I can do to change many o f the important things in my life. 5 . 1 can do just about anything I really set my mind to do. 6 . 1 often feel helpless in dealing with the problems in life. 7. Sometimes I feel that I'm being pushed around in life. ( Strongly agree Agree Disagree Strongly disagree ) Depression: During the past week, how much have you been distressed by: 1. Loss o f sexual interest or pleasure 2. thought o f ending your life 3. poor appetite 4. crying easily 5. feeling o f being rapped or caught 6. blaming yourself for things 7. feeling lonely 8. feeling blue 9. worrying or stewing about things 10. feeling no interest in things 11. feeling hopeless about the future ( N ot at all A little Quite a bit Extremely ) i R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 75 Happiness: 1. M y life could be happier than it is now. 2. These are the best years o f my life. ( Strongly agree Agree Disagree Strongly disagree ) 3. During the past few weeks how often have you felt pleased about having accomplished something? 4. H ow often have you felt particularly interested in something? 5. H ow often have you felt on top o f the world? ( Never Rarely Sometimes Frequently ) Life Satisfaction Life Satisfaction (Multifaceted Scale): 1. Financial situation - How satisfied are you with (your/your family's) present financial situation? 2. Health - In general how satisfied are you with your health? 3. M arriage - How satisfied are you with your (marriage/relationship)? 4. Children (Kids) - At this point in your life, how satisfied are you with your relationship with your children? 5. Relatives/friends - H ow satisfied are you with the kinds o f relationships you have with your relatives and friends? Life Satisfaction as a Whole: Please think about your life as a whole. How satisfied are you with it? ( Completely satisfied Very satisfied Somewhat satisfied N ot very satisfied N ot at all satisfied ) i R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Intergenerational social support and the psychological well-being of older parents in China

PDF

Alcoholism among the elderly: Fact or fallacy

PDF

Evaluation of Emeritus College

PDF

Comparison of gerontology-specific organizations to the business sector in hiring personnel: An analysis of qualifications and characteristics

PDF

Survey of registered nurses' attitudes toward elderly patients

PDF

Elder abuse crisis intervention techniques and their effectiveness

PDF

Marriage and health: The effect of marital status on health and the mechanisms by which marital status affects health among elderly people

PDF

The effectiveness of nutritional counseling in nutritional status and behavior in the elderly

PDF

Communication patterns between case management and contracted medical providers within the Social HMO

PDF

Children's attitudes toward old people and aging

PDF

Assisted living. Oregon and California: Two models compared

PDF

Heart disease among middle -aged and elderly persons in the United States: Trends and a multistate model

PDF

Dual eligible beneficiaries and managed care: Ensuring consumer protection through state Medicaid contracts

PDF

California's In-Home Supportive Services debate: An application and critique of the governmental politics model of the decision-making process

PDF

Alcohol expectancies and consumption: Age and sex differences

PDF

A comparative study of older and younger adults who have received job-specific classroom training to determine if there are significant differences as they relate to job placement

PDF

Aging in a continuing care retirement environment: A case study of Villa Gardens retirement community

PDF

Effects of migration and modernization on perceptions of role significance and status among elderly Samoans in the Los Angeles area

PDF

A statistical analysis of the formation and location factors of high -tech centers in the United States, 1950--1997: An evaluation using quasi -experimental control group methods

PDF

Endocytosis and transcytosis of diferric transferrin and apotransferrin in Caco-2 cells

Asset Metadata

Creator Kim, Jung Ki (author)

Core Title Determinants of life satisfaction among the high-functioning elderly: Implications of effects of psychological factors and activity participation

School Leonard Davis School of Gerontology

Degree Master of Science

Degree Program Gerontology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Gerontology,OAI-PMH Harvest,Psychology, clinical,psychology, developmental

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Crimmins, Eileen M. (committee chair), Peterson, David (committee member), Seeman, Teresa (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-17261

Unique identifier UC11336956

Identifier 1387840.pdf (filename),usctheses-c16-17261 (legacy record id)

Legacy Identifier 1387840.pdf

Dmrecord 17261

Document Type Thesis

Rights Kim, Jung Ki

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA