| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
RESEARCH ARTICLE |
Department of Sociology, and Department of Behavioral Sciences and Health Education, Rollins School of Public Health, Emory University, Atlanta, Georgia.
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
The purpose of this study is to investigate whether socioemotional selectivity predicts the exchange of emotional support and its linkage to emotional well-being throughout adulthood. Emotional well-being is individuals' avowed feelings toward, and emotional reactions to, their lives. It is often measured as the evaluation of happiness and satisfaction with life or as the subjective report of the frequency of positive and negative affects over a time period (Diener & Larsen, 1993
). Studies have indicated that levels of avowed happiness and life satisfaction remain constant, or increase slightly, between the ages of 20 and 70 (Diener & Suh, 1997; Keyes & Waterman, 2003). The frequency of positive affect shows evidence of minor declines in some studies and gains in others, whereas the frequency of negative affect generally declines with age (Charles, Reynolds, & Gatz, 2001; Diener & Suh, 1997; Mroczek & Kolarz, 1998; Shmotkin, 1990). Moreover, the duration of positive emotions appears to increase, and the duration of negative emotions appears to decrease, with age (Carstensen, Pasupathi, Mayr, & Nesselroade, 2000; see Appendix, Note 1).
Social support is salubrious, and research has shown this to be a primary function of emotional support, which entails expressions, gestures, and behaviors that are interpersonally intimate and emotionally positive (Stroebe & Stroebe, 1995, 1996). Examples of emotional support range from a smile, a hug, holding another's hand, and gently patting or rubbing another's arm or back to giving advice, encouragement, and consolation (Mancini & Blieszner, 1992). The perception of the availability of more emotional social support (Krause, 1997; Wethington & Kessler, 1986) and its receipt even under low-stress conditions has been linked to positive outcomes such as higher subjective well-being and fewer symptoms of mental illness (Stroebe & Stroebe, 1995, 1996).
However, the exchange of support depends on social ties and contact that dwindle with age (Cumming & Henry, 1961; Lang, 2001; Lang & Carstensen, 1994; Palmore, 1981). Contact diminishes such that only 10% of an older adult's day consists of direct social contact (Baltes, Wahl, & Schmid-Furstoss, 1990). Nonetheless, quantity of social contact may decrease with age while its quality may increase. In fact, the reduction in social contact may be a deliberate attempt to improve the quality of social contact, and, as the quality of contact increases, well-being should also increase.
According to goal theories, subjective well-being will increase as individuals attain or come closer to achieving their needs and wants (Campbell, Converse, & Rodgers, 1976; Michalos, 1985). Socioemotional selectivity theory (Carstensen, 1992, 1993, 1995) suggests that the key to emotional well-being is prioritizing the goal of emotionally satisfying social contact. However, individuals have two goals that motivate social contact and vary as a function of age. Individuals may seek social contact to regulate knowledge (i.e., information) about self and culture, or they may seek social contact to regulate their emotions.
When time is subjectively short or compressed (e.g., by a terminal illness), the goal of emotion regulation becomes salient and the goal of information regulation becomes unimportant. Thus, when there is little time left to live or little time left to spend in a context, individuals become motivated for contact that is emotionally satisfying. Individuals will therefore forgo contact with casual acquaintances or developing new ones in favor of contact with emotionally intimate persons. When, however, there is a lot of time left to live and abundant time to spend in a particular context, individuals seek numerous and diverse social contacts from which they can learn about themselves and society. A large and diverse pool of social contact ensures a rich source of information to satisfy the goal of information regulation (Carstensen, 1995; Carstensen, Gross, & Fung, 1997; Carstensen, Isaacowitz, & Charles, 1999).
A byproduct of growing older is a shift from a bountiful to a compressed subjective timeframe. Research has shown that the number and frequency of social ties decrease linearly with age, but relationships are perceived as more intimate and satisfying (Lang & Carstensen, 1994; Lansford, Sherman, & Antonucci, 1998). With age, adults are more likely to categorize and perceive others on the basis of whether they can provide emotionally close and satisfying contact (Fredrickson & Carstensen, 1990). There is also evidence that adults are better able to prevent unpleasant interpersonal exchanges as they age. With age, couples are able to discuss sensitive topics in ways that prevent the instigation and display of negative feelings (Carstensen, Gottman, & Levenson, 1995).
Besides the latter study by Carstensen and colleagues (1995), there has been little research on the dynamics of interpersonal exchanges with age. Do interpersonal exchanges become more emotionally satisfying or conducive to satisfaction with age? In addition to investigating whether the provision and receipt of emotional support decrease with age, this study therefore investigates whether the exchange of emotional support becomes more balanced with age.
Emotional support is a desirable quality that can produce in recipients and providers a sense of attachment (e.g., comforting and intimacy), belonging and alliance with others, guidance and reassurance of one's worth (Mancini & Blieszner, 1992; Weiss, 1969, 1974), and purpose and social contribution (Cutrona & Russell, 1987; Keyes & Ryff, 1998; Rook & Ituarte, 1999). Low levels of emotional support may create loneliness, anxiety, uncertainty, a sense that life is meaningless (Rook, 1987; Weiss, 1974), and vulnerability to stress (Stroebe & Stroebe, 1995).
Social exchange theorists (Homans, 1974) have asserted that individuals are motivated to maximize gains and minimize costs in social contacts. With age and the hypothesized shift toward the goal of emotion regulation (Carstensen et al., 1997), the exchange principle is translated into the goal of maximizing positive affect and minimizing negative affect. Social exchange theories of relationships may therefore help to partially explain how aging adults regulate emotion. The exchange of valued resources like emotional support may become more balanced and equitable with age (see, e.g., Antonucci & Jackson, 1990; Traupmann & Hatfield, 1983).
Exchange-based theories of relationships suggest that intimacy will increase as the equality of exchange increases (Deutsch, 1985; Rook, 1987; Sprecher & Schwartz, 1994; Walster, Walster, & Berscheid, 1978). Intimacy increases in equal-exchange relationships because equitable relationships incite positive emotions and minimize negative emotions. Imbalanced exchanges, where an individual may get less than she or he gives, may incite negative emotions or at least quash positive emotions. According to equity theory (Walster et al., 1978), perceived inequity generates negative affect. When individuals receive more than they put into a relationship (i.e., overbenefited), they feel guilt or shame. When individuals receive less than they put into a relationship (i.e., underbenefited), they become distressed or angry. Equitable exchanges of a resource like emotional support should create positive emotion and intimate social contacts; imbalanced exchanges of emotional support should create negative emotion or quell positive feelings.
The perspectives of socioemotional selectivity and social exchange theories converge into three hypotheses tested in this study. First, the quantity hypothesis predicts that the hours of emotional support given and received should decrease as chronological age increases because adults become selective about social contacts with age. Second, the exchange hypothesis predicts that the discrepancy between the hours of emotional support given and received should decrease with age because more balanced social exchanges are conducive to greater intimacy and satisfaction, and the goal of emotional regulation becomes more salient with age. Third, the goal hypothesis predicts that, with age, balanced exchanges of emotional support will predict higher levels of positive affect and lower levels of negative affect compared with unbalanced exchanges of (i.e., giving more or receiving more) emotional support. That is, the goal of emotional regulation becomes more salient with age; therefore, unbalanced exchanges of emotional support will fail to meet individuals' social contact goal with age.
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Field procedures were initiated in January 1995 and lasted approximately 13 months. The respondents were contacted by professional personnel; those who agreed to participate in the entire study took part in a computer-assisted telephone interview lasting 30 min on average. Respondents were then mailed two questionnaire booklets requiring about 1.5 hr on average to complete. As incentives for participation in the complete study, each respondent was offered $20, a commemorative pen, periodic reports of study findings, and a copy of a monograph on the study. The sample consisted of 3,032 adults, with a 70% response rate for the telephone phase and an 87% response rate for the self-administered questionnaire phase, or a combined response rate of 61% (.70 x .87 = .61).
Measures
Emotional support exchanges.
Respondents were asked to think about the emotional support they give and get. Emotional support was defined for respondents as "giving or getting comfort," "listening" or "having someone listen to you," and "giving or getting advice." Respondents then estimated the number of hours of emotional support they gave and received during an average month. Respondents first indicated the hours of emotional support given and subsequently estimated the hours of emotional support received in six sets of relationships.
The recipients (and then sources) of emotional support were (in the following order): "your spouse or partner," "your parents or the people who raised you," "your in-laws," "your children or grandchildren," "any other family members or close friends," and "anyone else (such as neighbors or people at church)" (see Appendix, Note 2). The hours of emotional support were recoded into ranges, and each range was coded to the midpoints as follows: none = 0, 1 to 4 hr in a month = 2.5, 5 to 8 hr = 6.5, 9 to 16 hr = 12.5, 17 to 24 hr = 20.5, 25 to 32 hr = 28.5, 33 to 40 hr = 36.5, and 41 or more hr = 44.5 (i.e., all variables were top-coded to reflect the equivalent of a 40-hr "work week"). The hours of emotional support over the six relationships were summed to form separate scales of the total monthly hours of support received and given.
Emotional well-being.
Respondents indicated how much of the time during the past 30 days—all, most, some, a little, or none of the time—they felt six negative and six positive indicators of affect. The negative indicators were feeling (a) so sad nothing could cheer you up, (b) nervous, (c) restless or fidgety, (d) hopeless, (e) that everything was an effort, and (f) worthless. The positive indicators were feeling (a) cheerful, (b) in good spirits, (c) extremely happy, (d) calm and peaceful, (e) satisfied, and (f) full of life. Each set of items was summed and divided by the number of constituent items so that the final scale ranged from 1 to 5, with a higher score indicating more negative affect and more positive affect. The internal reliability of the Positive Affect scale was .91; for the Negative Affect scale, it was .87. Details about the sources of the affect items used in the MIDUS can be found in Mroczek and Kolarz (1998)(p. 1337).
Social structural variables and controls.
Chronological age was the central independent variable, which was coded into age cohorts as follows: ages 25 to 34, ages 35 to 44, ages 45 to 54, ages 55 to 64, and ages 65 to 74. Several sociodemographic variables that are possible predictors of the receipt or provision of emotional support as well as emotional well-being were also measured. Dummy variables indicating respondents' gender, race (White vs minority) marital status (currently married vs all other), employment status (part time or full time vs all other) were used as control variables. In addition, a dummy variable for parental status and a continuous variable reflecting the number of biological and adopted children were used as controls. Respondents indicated the highest grade or year of schooling they had completed at the time of the interview. Education was coded as 1 (less than high school), 2 (high school), 3 (some college; vocational), 4 (bachelor's degree), or 5 (master's or doctoral degree). Income was measured as the sum of respondents' self-reported (a) own personal income, (b) income from a spouse or partner, and (c) any additional household income from all other sources during the past year. All sources of income were summed to form a continuous measure of total household income.
All multivariate analyses adjusted for respondents' physical health, which may affect the individual's need for support, ability to provide emotional support, and emotional well-being. Respondents indicated their current health status by judging how much their health limited several daily activities a lot, some, a little, or not at all. The activities included (a) lifting or carrying groceries; (b) bathing or dressing oneself; (c) climbing several flights of stairs; (d) bending, kneeling, or stooping; (e) walking more than a mile; (f) walking several blocks; (g) walking one block; (h) vigorous activity (e.g., running, lifting heavy objects); and (i) moderate activity (e.g., bowling, vacuuming). The internal alpha reliability of the limitations of daily activities scale was .93.
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
Next, multivariate regression was used to test the quantity hypothesis by regressing the hours of emotional support given and the hours of emotional support received onto the age and control variables. Because the conclusions for these and all subsequent analyses were unchanged by sample weighting, only the unweighted sample estimates are shown (see Winship & Radbil's 1994 rationale for using unweighted data in this case). Focusing on the age differences in the provision and receipt of emotional support in Table 2, one sees that the results support the quantity hypothesis. That is, the provision and receipt of emotional support decreased linearly with age.
|
|
|
The goal hypothesis predicted that unbalanced exchanges of support would affect emotional well-being with age because the social contact goal of emotional regulation only becomes salient with age. Table 4 reports the regressions of positive and negative affect onto the control variables, total hours of emotional support given and received, and variables reflecting the balance and imbalance of exchanges. Specifically, adults whose discrepancy of emotional support equaled 0 were used as the reference group; two dummy variables were created to reflect adults who gave more support than they received and adults who received more support than they gave. The regressions of positive and negative affect were performed separately by each age group to determine whether the relationship of exchanges of emotional support with positive and negative affect varied by age cohort.
|
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Findings unequivocally supported the quantity hypothesis and the exchange hypothesis. With age, adults spent less time engaging in acts of emotional support and less time receiving emotional support. Moreover, the discrepancy in the hours of emotional support given and received diminished with age. Not only did they give and receive the most hours of emotional support, the youngest adults (i.e., 25 to 34) gave an average of nearly 19 more hours of support than they received. In contrast, the oldest adults (i.e., 65 to 74) gave and received the fewest hours of emotional support, and they gave about 7 more hours of emotional support than they received.
In turn, this study inquired whether the relationship of the imbalance of exchanges of emotional support with emotional well-being varied by age. Imbalanced exchanges were not associated with decreased positive affect and elevated negative affect among younger adults between the ages of 25 and 54, although these adults had more unequal exchanges of support than the older adults. Although the older adults between the ages of 55 and 74 had relatively more equal exchanges, those who had imbalanced exchanges reported less positive and more negative affect than those who had equal exchanges. These findings are consistent with the theoretical proposition that the goal of emotion regulation may be achieved, in part, through principles of social exchange as adults age.
However, the goal hypothesis was supported primarily when older adults received more emotional support. With one exception, adults between the ages of 55 and 74 who gave more emotional support had similar levels of emotional well-being as those who had equal exchanges. The exception to this rule occurred among the oldest adults, aged 65 to 74, who felt more negative affect when they gave more emotional support compared with the adults with perfectly balanced exchanges. In addition to violating the goal of emotion regulation, receiving more emotional support as an older adult may predict lower emotional well-being because it reflects an undesirable sense of dependency (see Magai et al., 2001), or it could reflect a undesirable situation in which an older adult is receiving unsolicited or unsuitable support (see Rook, 1984). Even if giving more support violates the goal of emotional regulation, studies have shown that it can be a source of purpose and contribution (see, e.g., Krause, Herzog, & Baker, 1992) that may offset the goal violation.
Findings from this study challenge the convoy model of social support (see Antonucci & Akiyama, 1987; Antonucci & Jackson, 1990). Antonucci and colleagues argued that the norm of reciprocity can operate more long term over the life span because individuals keep a mental account of the total support put into and taken out of their "support bank." With age, individuals who may require more emotional support than they can give may not feel distress if they had provided more support to others when they were younger. Because younger adults are likely to give more support than they require—a finding consistent with the results of this cross-sectional study—they can draw on this debt as they age and receive more support without feelings of guilt. However, adults aged 55 to 74 who received more emotional support felt more negative affect than the same-aged adults with equal exchanges, which suggests that receiving more emotional support as an older adult violates some standard of interpersonal exchange. The correlations of support given and received were high in each age cohort (all rs = .80 or higher), suggesting that reciprocity was operative in each age cohort. Thus, if individuals view life in terms of time left to live as they age, the receipt of more emotional support could generate an unpleasant state of urgency over the time one has left to repay others' kindness and support. Future research may benefit from measurement of adults' subjective urgency and time frame for reciprocating the receipt of emotional support.
Strengths of the present study for testing selectivity theory include its emphasis on behavior in relationships (see also Carstensen et al., 1995). In addition, respondents in the MIDUS were not directly asked to rate the satisfaction of their social contacts, nor were they directly asked to rate the discrepancies in emotional support exchanges. Prior studies have focused on the frequency with which people have social contact and subjective reports of their satisfaction with social contact (see Carstensen et al.'s 1997 review). The present study used an indirect method of measuring social contact by asking respondents to report the hours of emotional support given to others in a typical month and the hours of emotional support received from the same set of people in a typical month. Also, this study used a representative national probability sample of U.S. adults. The multivariate analyses adjusted for relevant demographic variables, which permit generalization to the average adult in the United States.
Limitations of this study include the cross-sectional design of the MIDUS. Longitudinal data are needed to begin to investigate whether and how changes in the exchange of emotional support predict changes in emotional well-being. The measures of emotional support were perceived rather than actual exchanges of support. Although the social support literature has shown stronger linkages of perceived than received support with health outcomes (Stroebe & Stroebe, 1996), future tests of the exchange of emotional support would benefit from behavioral and experience sampling measurements of support (see, e.g., Carstensen et al., 2000). Research may also benefit from the investigation of the various kinds of exchanges that can occur throughout adulthood. Rook and Ituarte (1999)(see also Rook, 1987) argued that social contacts can be characterized in terms of companionship and instrumental supports, in addition to emotional support. That is, adults pursue common activities (i.e., companionship) that can be emotionally satisfying. Adults also provide instrumental support in terms of money, tools, and favors (e.g., repairing broken objects for others), which may generate emotional well-being.
As the baby boom generation ages, understanding the sources of emotional well-being has taken on greater importance because of recent findings that it protects against disease, disability, and mortality. The absence of emotional well-being may be implicated in the onset of cardiovascular disease (Musselman, Evans, & Nemeroff, 1998) and appears to elevate risk for physical disability and premature mortality in older adults (Ostir, Markides, Black, & Goodwin, 2000; see also Danner, Snowdon, & Friesen, 2001). Finally, there is mounting evidence that the risk of suicide, which increases dramatically with age, increases as emotional well-being decreases (Koivumaa-Honkanen et al., 2001; Weerasinghe & Tepperman, 1994).
Notes
1. The terminology emotional well-being and positive and negative affect are used interchangeably throughout this article (see, e.g., Bradburn, 1969).
2. All measures in this study were taken from the self-administered questionnaire.
3. The reciprocity correlations by type of relationship revealed that reciprocity was high and stable by age cohort for relationships with spouse, other family members, and anyone else (e.g., neighbor). The reciprocity correlation with parents and in-laws decreased with age, whereas the reciprocity correlation with children increased with age (results available on request).
4. The discrepancies between the hours of support given to and received from spouses, parents, in-laws, children, and other family members decreased with age. The only exception was the exchange of emotional support with parents. Perhaps indicative of the role of caregiver, the discrepancy in emotional support exchanged with parents increased between the ages of 25 and 64 (but not between 65 and 74).
|
Acknowledgments |
|---|
Address correspondence to Corey L. M. Keyes, Department of Sociology, Room 225, Tarbutton Hall, 1555 Pierce Drive, Atlanta, GA 30322. E-mail: ckeyes{at}emory.edu
Received for publication February 16, 2000. Accepted for publication March 4, 2002.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  C. Magai, N. S. Consedine, A. R. King, and M. Gillespie Physical Hardiness and Styles of Socioemotional Functioning in Later Life J. Gerontol. B. Psychol. Sci. Soc. Sci., September 1, 2003; 58(5): P269 - 279. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
