This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation

Negative Interactions in Close Relationships Across the Life Span

¹ Institute for Social Research, University of Michigan, Ann Arbor.
² Department of Social Psychology, University of Tokyo, Japan.
³ Department of Psychology, University of Michigan, Ann Arbor.
⁴ Department of Psychology, University of the Sacred Heart, Tokyo, Japan.

Address correspondence to Hiroko Akiyama, 5080 Institute for Social Research, University of Michigan, Ann Arbor, Michigan 48106. E-mail: akiyama{at}isr.umich.edu

	Abstract

TOP Abstract Methods Results Discussion References

 
This article examines age differences in positive, but especially negative, interactions in close relationships. Two community-based representative samples of people from 13 to 96 years of age from the United States and Japan were used to explore these relationships. Analyses indicate stability across age in positive interactions but a general decline in negative interactions in both countries. Three possible explanations for this age difference in negative interactions were examined: social maturity, familiarity, and contact frequency. Results provide most support for the contact frequency explanation. And finally, intercorrelations across relationships, that is, spouse, mother, father, child, and same-gender best friend, indicate moderate to high intercorrelations across all ages with a slight increase among older Japanese age groups. These results can be summarized as lending support to a generalized reduction in negative interactions with age but exceptions are noted in specific relationships and cultural traditions.

Close relationships can be positive and negative. They provide people with joy, comfort, and support. At the same time, they can be a source of conflict, frustration, and disappointment (Antonucci, Akiyama, & Lansford, 1998; Ingersoll-Dayton, Morgan, & Antonucci, 1997; Levitt, Silver, & Franco, 1996; Okun & Keith, 1998; Rook, 1997; Walen & Lachman, 2000). Findings from prior research suggest a general tendency of increasing positivity and decreasing negativity in affect and social relations as people age (Andrews & Withey, 1976; Campbell, Converse, & Rodgers, 1976; Hansson, Jones, & Fletcher, 1990; Smith & Goodnow, 1999; Veroff, Douvan, & Kulka, 1981; Veroff, Kulka, & Douvan, 1981). This empirical finding has been replicated, predominantly, in the United States and other Western industrialized countries (e.g. Baltes & Mayer, 1999; Carstensen, Isaacowitz, & Charles, 1999; Lansford, Sherman, & Antonucci, 1998; Okun & Keith, 1998). It still remains to be empirically investigated whether the age-related patterns and their explanations replicate in non-Western industrialized countries. The present investigation briefly considers positive relations but focuses on age differences in negative interactions in both Western and non-Western industrialized countries.

Various possible explanations for age differences are considered. The literature suggests three plausible explanations for decreasing negativity in close relationships: maturity, familiarity, and contact frequency. The first explanation suggests that with increasing age people mature. They acquire social skills from experience, learn to control their emotions (Carstensen, Gottman, & Levenson, 1995; Charles, Reynolds, & Gatz, 2001; Diehl, Coyle, & Labouvie-Vief, 1996; Lawton, Kleban, Dean, Rajagopal, & Parmelee, 1992), and improve their emotional understanding (Labouvie-Vief & DeVoe, 1991). For example, Diehl and colleagues (1996) examined coping and defense strategies among people from 10 to 70 years of age and reported what might be interpreted as a demonstration of this maturity, albeit with cross-sectional data. They found that older people used more impulse control and had a tendency to positively appraise conflict situations whereas younger people were more likely to be outwardly aggressive and less likely to use impulse control.

The second explanation suggests that because older people are more likely to have known these close relations (e.g., spouse) longer than younger people, they are more familiar with them, know them better, and have learned to anticipate and successfully cope with problems that have repeatedly arisen in their relationship over the years. Drawing from the social cognitive perspective, it may be that these individuals hold relational schemas that allow them to anticipate and manage conflict in their close, long-term relationships (Baldwin, 1992). Alternatively, they may become accustomed to such negative aspects of relations and no longer be upset by them or perceive them as problematic. Reis and Downey (1999) have recently noted that social cognition is an important aspect of relationships that has previously been ignored. People have attitudes, ideas, beliefs, and opinions about those people with whom they have close social relationships. The relational schemas Baldwin (1992) refers to are a good example of social cognition interacting with, and fundamentally influencing, how individuals interpret their social relationships. Familiarity, in social cognition terms, means a well-developed cognitive schema for those people with whom they have close relationships. These involve distinct schemas, which identify each person including their most common positive and negative attributes, behaviors, and/or relational styles.

The third possible explanation is that negative interactions are a function of the frequency of contacts. In general, evidence indicates that frequency of contact with social network members declines with age (Carstensen, 1992; Carstensen et al., 1995; Fredrickson & Carstensen, 1990). This argument suggests that the more contacts or interactions people have, the more conflicts and frustrations they tend to experience. If Carstensen (1992) is correct that older people tend to have fewer interactions, this reduction in contact frequency may explain the decreased negative interactions sometimes noted among the elderly. Traditional social exchange theories (Roloff, 1981) and interdependence theories (Kelley, 1979), however, suggest that increased contact is likely to lead to increased conflict and negative interaction but only under certain conditions, such as inequity in the relationship or a lack of shared rewards. On the basis of a cognitive interpretation of relationships, these perspectives argue that more frequent contact requires a joint view of life goals and experiences. Thus, agreement about or successful achievement of targeted outcomes serves to maximize positive aspects of relationships, whereas failure to achieve or discrepancy concerning goals and outcomes lead to an increase in negative aspects of relationships. Similarly, Baumeister, Wotman, and Stillwell (1993) suggested that close relationships characterized by mutuality tend to be positive but can be negative in the absence of mutual commitment. Because people change their life goals as they age, they may reduce their frequency of contact with people who do not share their goals or with whom a lack of mutuality becomes clear, thus minimizing the number of negative social interactions and relationships to which they are exposed.

Another interesting issue to consider is the degree to which a target person's interactions are generalized across relationships, that is, the degree to which all relationships might be positive or negative. Empirical evidence varies, with some research suggesting that negative interactions tend to be specific to only one relationship (Baumeister et al., 1993; Hansson et al., 1990; Levitt et al., 1996) whereas other research suggests that negativity generalizes to several or most relationships of some people (see Okun & Keith, 1998). Thus, in the former case, the negative interaction is hypothesized to be only with one specific person such as a spouse, child, or friend, whereas in the latter case all relationships (i.e., with spouse, child, and friend) are characterized by negative interactions. Addressing this question of specific versus generalized negative interactions empirically may provide confirming evidence that will advance our understanding of the causes of negative interactions. Negative interactions with just one person would suggest that characteristics of that one person explain the negative relationship, whereas generalized negative interactions across several people would suggest that characteristics of the focal person or of the situation are more appropriate explanations for the negative interactions. We are fortunate in that the data available in the present study provide specific information, both positive and negative, about each of the respondents' close personal relationships.

The objectives of this study are threefold: (a) to identify age patterns of positive and negative interactions in close relationships, (b) to empirically explore the three alternative explanations for the observed age patterns of negative interactions outlined above, and (c) to examine the consistency of negative interactions across relationships over the life course. The analysis focuses on five dyadic relationships: mother, father, spouse, child, and same-gender best friend.

An additional distinctive feature of this study is a comparison of negative interactions in close relationships in two vastly different cultures. We assume that age affects all individuals across cultures. However, as individuals age in sociocultural contexts, aging may not affect individuals' social relations uniformly across cultures. We believe that the examination of negative interactions in two similarly developed but culturally diverse populations, that is, Japan and the United States, will enhance our understanding of how age affects these specific characteristics of social relationships. Compared with the United States, Japan is characterized by a larger proportion of three-generation households, the tradition of elderly being cared for by family members, the obligatory nature of marital relationships, and a significantly lower divorce rate.

On the basis of the aforementioned theoretical and empirical grounds, we propose the following hypotheses.

Hypothesis 1-1: Negative interactions in close relationships will decrease with age in both countries.

Hypothesis 1-2: Positive interactions in close relationships will increase with age in both countries.

Although we expect that all three explanations, to various degrees, account for the decline of negative interactions as people age across the two nations, more specifically we hypothesize the following.

Hypothesis 2-1: Social maturity will affect the age-related decline of negative interactions similarly for all close relationships in the two countries.

Hypothesis 2-2: Familiarity is more likely to affect those relationships from which people have virtually no escape regardless of relationship quality. More specifically, negative interactions with child where the relationship is both permanent and expectancy based, will be more affected by familiarity than those with best friends in both countries. Furthermore, familiarity will be more strongly associated with negative interactions with spouse in Japan, where divorce is rarely an option, than with spouse in the United States.

Hypothesis 2-3: Due to a higher prevalence of intergenerational coresidence in Japan, the mediating effect of contact frequency will be more marked in parent–child relations (i.e., relationships with mother, father, and child) in Japan than in the United States.

Hypothesis 3: As we expect that negative interactions decrease with increasing age across all relationships in the aggregate, we hypothesize that negative interactions in different relationships covary at an individual level as well. In other words, there will be a consistency across relationships in negative interactions within an individual in both nations.

These hypotheses will be examined using cross-national survey data that were collected from probability samples aged 13–93 in the two countries.

	Methods

TOP Abstract Methods Results Discussion References

 
Sample
This study uses data from the cross-national survey "Social Relations and Mental Health Over the Life Course." The regional probability samples at comparable sites in the United States and Japan (the greater Detroit and Yokohama metropolitan areas) consist of 1,702 Americans ranging in age from 8 to 93 and 1,842 Japanese ranging in age from 8 to 96. Older people (age 60+) are oversampled by doubling the sampling probability. Professional interviewers administered in-home, structured interviews approximately 1 hr in length. Because a slightly different version of the questionnaire was used for the children (ages 8–12), they were excluded from this study. Findings reported in this article are based on the data from 1,498 Americans and 1,641 Japanese aged 13 to 92. Table 1 provides the demographic characteristics of the two samples. The response rates were 72% and 67% for the U.S. and Japanese surveys, respectively.

Because the primary purpose of the original study was to assess the linkage between several domains of positive interactions and mental health, only two questions about negative interactions were included in the original questionnaire. The negative interactions score is the mean of the responses to the following two items: (a) "My (e.g., mother) makes too many demands on me"; and (b) "My mother gets on my nerves." The Cronbach's alphas of the negative interaction scale also range from adequately to high:.58 (mother),.59 (father),.61 (spouse),.68 (child), and.53 (best friend) for the U.S. sample and.79 (mother),.82 (father),.82 (spouse),.79 (child), and.81 (best friend) for the Japanese sample. Thus, both positive interaction and negative interaction scores range from 1 to 5, with higher scores indicating higher levels of positive and negative interactions, respectively.

Explanatory factors
Three explanatory factors were explored: social maturity, familiarity, and contact frequency. Because assessment of social maturity was not a planned part of the original study, a social maturity scale was constructed utilizing items from the Bem's Sex Role Inventory. On the basis of factor analysis of 40 items of Bem's Sex Role Inventory, we selected four items to form a social maturity scale. Thus, social maturity was measured by a 4-item scale comprised of the following characteristics: sensitive to the needs of others, understanding, compassionate, and eager to soothe hurt feelings. Respondents were asked to rate how well each characteristic describes themselves on a 7-point scale (1 = almost never true, 7 = almost always true). Social maturity was assessed as the mean of the responses to these four items. The Cronbach's alphas of this scale are.77 for the U.S. sample and.85 for the Japanese sample. Familiarity was simply and directly measured by the number of years a respondent had known a partner (e.g., spouse) prior to the interview. And finally, frequency of contact was assessed from a single self-report item asking on a 5-point scale how often the respondent had contact with his or her mother, father, spouse, child, or same-gender best friend. Contact frequency was coded as follows: 5 = live together or daily contact, 4 = once a week or more, 3 = once a month or more, 2 = once a year or more, 1 = none.

	Results

TOP Abstract Methods Results Discussion References

 
This section is organized as follows. A description of positive and negative interactions with mother, father, spouse, child, and same-gender best friend in both the United States and Japan is provided first. Next, each of the three alternative explanations for differences in negative interactions across age is examined empirically and the success of each explanation is considered. And finally, the degree to which negative interactions are person specific or generalized across all relationships is explored. A copy of the intercorrelations of all variables is available from the first author upon request.

Age Patterns in Positive and Negative Interactions
Figure 1 shows the amount of negative interactions with mother, father, spouse, child, and best friend reported by Americans in six age groups: 13–19, 20–34, 35–49, 50–64, 65–74, and 75+. It should be noted that some of the older categories are eliminated either because no one or too few people were present in the category. Figure 2 presents comparable graphs for the Japanese sample. An inspection of these 10 graphs reveals certain consistent age-related patterns. With increasing age, in general, people are less likely to have negative interactions with mother, father, child, and best friend, but not with spouse. On the other hand, their positive interactions (not shown in Figures 1 and 2) tend to remain at the same level across all age groups. We found these age patterns both in the U.S. and Japanese samples. Analyses of variance (ANOVA) confirmed significant age differences in negative interactions in all relationships, except spouse in Japan. Among Americans there is significant age effect in the relationship with spouse due to the decline in negative interactions in middle age. There were no significant age differences in positive interactions in any of the five relationships across the two nations.

View larger version (27K): [in this window] [in a new window]   Figure 1. Age differences in negative interactions across five relationships in the United States

View larger version (24K): [in this window] [in a new window]   Figure 2. Age differences in negative interactions across five relationships in Japan

Exploration of Alternative Explanations of Age Patterns in Negative Interactions
Table 2 provides the results of the regression analyses designed to identify the nature of age-related patterns of negative interactions. First, the linear relationship between age and negative interactions was tested. Then, in order to explore the possibility that there is a curvilinear association of age and negative interactions, age² was added in the regression model.

Thus, these findings largely support Hypothesis 1-1. With increasing age, negative interactions decline for all close relationships except spouse. However, negative interactions with parents slightly increase in old age. Given the finding of no age differences in positive relationships in ANOVA, Hypothesis 1-2 is not supported.

In order to test the Hypotheses 2-1, 2-2, and 2-3, we used path analyses and examined the mediating effect of the three potential explanatory variables (social maturity, familiarity, and contact frequency) for the age effect on negative interactions. For the relations with fathers and mothers in both countries where the age² variable was significant in the regression analyses, we tested the pathways from both age and age² to negative interactions. The path models are shown in Figure 3. Only one explanatory variable (i.e., social maturity, years known, or contact frequency) was entered into each path model as a mediator. Following the strategy suggested by Baron and Kenny (1986), we estimated the indirect effects of age (and age²) on negative interactions through the mediator using the Sobel Test (Sobel, 1982). It should be noted that familiarity was not tested for the relations with mother and father, because the number of years a parent is known is almost always equal to the age of the respondent. Also, because more than 97% of the married couples in the sample lived in the same household, contact frequency was not tested in the relation with spouse.

View larger version (15K): [in this window] [in a new window]   Figure 3. Path models of mediators between age, age2, and negative interactions

We also explored the effect of coresidence. We repeated the path analyses replacing contact frequency with coresidence. The indirect effect of age on negative interactions through coresidence was significant in exactly the same relations where the path model with contact frequency was significant in Japan. In the United States, the indirect effect through coresidence was significant not only in the relation with child where the model with contact frequency was significant but also in the indirect effects of both age and age² in the relation with mother. These results confirm that living arrangements do affect the decline of negative interactions in intergenerational relations as people age.

In sum, the results of path analyses indicate the following: (a) social maturity mediates the link between age and negative interactions with best friend in Japan; (b) familiarity does not explain the age-related decline of negative interactions in any relationships; and (c) contact frequency accounts for some portion of the age differences in negative interactions in the relationships with one's child in the United States and in the relationships with mother, father, and child in Japan. In those relationships, in general, social maturity increases and contact frequency decreases with age. More maturity and less contact are associated with fewer negative interactions. At the same time, the quadratic patterns in negative interactions (i.e., a slight increase in old age after a steady decline from young age through middle age) in the relationships with mother and father in Japan are also explained by increased contact with mother and father in old age.

Thus, Hypothesis 2-1 is only supported in the relationship with best friend in Japan. Hypothesis 2-2 is not supported at all. Our findings mostly support Hypothesis 2-3. We found a significant mediating effect of contact frequency in the intergenerational relationships (with mother, father, and child) but not in the relationship with friend in Japan. Rather unexpectedly, contact frequency is found to explain the age-related decline of negative interactions with one's child in the United States as well.

Consistency in Negative Interactions Across Relationships
Table 4 presents the correlation coefficients among negative interactions in each of the five close relationships. The numbers above each diagonal are for Japan, and the numbers below each diagonal are for the United States. The correlation coefficients indicate the degree to which the respondent reported a similar level of negative interactions across specific relationships. This addresses the question of whether an individual who has one relationship, such as with mother, which is characterized by negative interactions, is likely to have similar negative interactions with most or all of their other close relationships, such as with father, spouse, child, and best friend as well. Examination of these correlation coefficients across the six age groups in the two countries highlights three interesting patterns. First, both in the United States and Japan, negative interactions in all five close relations are moderately correlated in the first three age groups (up to the 35–49 age group). Second, distinctive correlation patterns are observed in the two countries for the older age groups. In the United States, negative interactions with parents do not appear to permeate other relationships, that is, spouse, child, and best friend, whereas negative interactions in all five relations continue to be correlated at a significant level in Japan. Third, these data indicate that the negative interactions with all five relations are intercorrelated particularly among the older age groups in Japan, and in some cases these correlations are quite high (e.g., r =.77) between negative interaction with mother and negative interaction with father in the 50–64 age group. Thus, Hypothesis 3 is largely supported.

	Discussion

TOP Abstract Methods Results Discussion References

These data, from two random probability samples in the United States and Japan, are generally consistent with previous findings (e.g. Levitt et al., 1996; Okun & Keith, 1998; Rook, 1997) indicating a gradual decrease of negativity in close relationships with increasing age. We benefit from our separate analyses by relationship, however, because spousal relations appear to be an exception in both countries. The data reported in this study suggest that negative interactions with spouse neither increase nor decrease linearly with age. In the United States, a curvilinear pattern is evident with the lowest levels of negative spouse interactions apparent among the middle rather than the older age groups. Among the Japanese, negative interactions with spouse appear to be relatively stable across different age groups, although the analysis indicates a slight decline with age. We interpret this finding as suggesting that decreasing negativity with age is not generalized across all close relationships, but rather is relation specific. Levitt and colleagues (1996) found that most people reported that they coped with troublesome relationships by terminating or withdrawing from them. Short of divorce, this is especially difficult to accomplish with a spouse. It may be that middle-aged people, because of their multiple role commitments and intergenerational obligations to parents and children, have less time to interact with their spouse. Because our data obviously include spouse information among people who have not divorced, the inability to withdraw from a spouse, especially among older people who have fewer outside commitments, may explain the lack of a decrease in negative interactions with spouse in old age in the United States. An unexpected finding is that negative interactions with mother and father increase in old age. This age-related pattern was observed consistently in the two countries. Why do negative interactions increase only in these specific relationships? The most plausible explanation is that the relationship with parents in old age is more likely to involve care of dependent parents, which often brings about feelings of frustration, irritation, and guilt toward parents. This may be an increasingly visible cohort effect. Because of the prolonged life expectancy, a considerable portion of older people now have living parents. Those parents are likely to be advanced in age and require various levels of care. The respondents themselves or their spouse may also have health problems. It is not hard to imagine that such situations can be detrimental to one's relationship with parents. Adding these new strains to previously existing ones may lead to the observed increase in negative interactions.

In keeping with our three hypothesized explanations for decreasing negativity with age, we examined three attributes: social maturity, familiarity, and contact frequency. First, social maturity was found to affect negativity only in the relationship with best friend in Japan. A similar trend was observed in the U.S. friend relationship as well, but failed to reach a statistically significant level. We hypothesized that people socially mature with age and that social maturity would reduce negative interactions across all relationships. Obviously this is not the case. It appears that increasing maturity with age affects negative interactions with best friend but not parent–child relationships. This finding suggests that the nature of negative interactions with friend and parent is different. Friendship is a voluntary relationship and, therefore, may benefit from social maturity (e.g., social skills), whereas the parent–child relationship, being both obligatory and stable, may suffer from less changeable, less positive interactive patterns.

Second, familiarity, which was measured by the number of years a partner was known, is not linked to the age-related decline in negative interactions in close relationships. Close relationships are mostly long-term and intimate in nature. In such relationships, the number of years a partner is known appears not to make a significant difference in familiarity after the first few years.

Third, our analyses suggest that contact frequency accounts for a part of the observed decreasing negativity in intergenerational relationships with age. We found this indirect effect of contact frequency in the relations with child in the United States and with mother, father, and child in Japan. People tend to have fewer contacts with mother, father, and child as they age, which reduce the number of negative interactions in those relationships. Furthermore, we hypothesized that because of a higher prevalence of intergenerational coresidence in Japan, the mediating effect of contact frequency would be more marked in Japan than in the United States. When two persons live in the same household, share resources, and are intimately involved in each other's life, they have more opportunities to engage in negative interactions, as well as positive ones, compared with those who do not live together. Persons who live together also have less control over the frequency of their interactions. Consequently, they are less likely to be able to avoid negative interactions. Thus, decreasing negativity in close relationships with age may, in part, be due to decreasing coresidence among parents and children. Coresidence may also explain the lack of an age difference in negative interactions with spouse that was observed both in the United States and Japan. Additional analysis of coresidence supports this interpretation of our contact frequency data.

The correlations among negative interactions in the five close relationships provide a deeper insight into the nature of close relationships. We found moderate to high correlations among negative interactions with all relationships included in this study for younger American adults and for Japanese adults of all ages. This finding leads us to conclude that the cause of negative interactions may be in the focal person rather than in the partner. However, the correlations of negative interactions among the older age groups in the American sample suggest that negative interactions are correlated within a specific cluster of social relationships. It appears that the close relationships of American adults are segmented into multiple clusters, for example, one segment for parents and another segment for spouse, child, and best friend. These two segments appear to be relatively independent, each with distinctive dynamics. The latter observation indicates that we cannot entirely attribute the cause of negative interactions to personal characteristics of a focal person. It is plausible that certain group dynamics within a specific segment of social relations is also likely to contribute to the explanation of negative interactions.

We found higher correlations in the older age groups compared with the younger groups in Japan. This might suggest an interesting aspect of age-related changes in close relationships. According to socioemotional selectivity theory, older people actively select relationships from which they derive the most emotional satisfaction (Carstensen et al., 1995). However, when the density or interaction among members is high, and not voluntary, as is the case in a smaller selected social network who coreside, both positive and negative interactions appear to permeate throughout the network.

Developmental explanations based on cross-sectional comparisons about age changes must be considered highly speculative. The cross-sectional comparisons also cannot rule out the possibility of cohort effects. For example, the declining negative interactions with age might be due to different expectations for relationships held by differing cohorts. These findings, and certainly the proposed explanations, need to be confirmed by longitudinal data. In interpreting our findings, we also need to consider the possibility of response bias among different age groups. In particular, our findings of reduced negative interaction with age might possibly be accounted for by a larger social desirability bias among older respondents. Rodgers, Andrews, and Herzog (1992) investigated measurement errors including response biases in data from a large survey and found no age differences in measurement errors. However, because we cannot assess social desirability bias empirically in this data set, we cannot completely rule out this possible explanation of our findings.

In sum, the results of the present study, though limited by the cross-sectional nature, contribute to our knowledge of age differences in social interactions in several ways. Our findings suggest that positive interactions are maintained at about equal levels over the life course and that negative interactions decline in all except the spousal relationship. An additional significant finding is that negative interactions with parents slightly increase in old age after the steady decrease from young adulthood to middle age. Among the three explanations that were explored in the present study, that is, social maturity, familiarity, and frequency of interaction, the latter explanation for age differences was best supported. More specifically, decline of coresidence accounts for age-related decrease (and increase in old age in the relations with mother and father) in negative interactions. And finally, significant intercorrelations in negative interactions among different relationships suggests that multiple explanations for negative interactions are likely, that is, characteristics of the focal person, of their close relationships, and the culture within which they live, all contribute to the nature and experience of negative interactions.

Although interest in the social relations of older people has increased significantly in recent years, much of the work has focused on positive interactions and their advantages. The current study supports previous findings but also notes the unique and perhaps critical contribution of negative interactions with those to whom we feel close, that is, mother, father, spouse, child, and friend. These findings also highlight the importance of taking a developmental perspective both at the individual and the family level. Our relationships clearly vary and are affected by age, development, and family life cycle. Future research on social relations needs to take these multiple perspectives into account.

	Acknowledgments

 
This research was funded in part by research grants to Toni C. Antonucci and Hiroko Akiyama (MH46549 and AG13490) and a grant to Hiroshi Azuma from the Ministry of Education in Japan. Elizabeth Langfahl was supported by a Survey Research Center Founder's Graduate Research Assistant Fellowship, the Elizabeth Douvan Award. We are grateful to Lisa Byrd for assistance with the final preparation of the article and to our colleagues in the Life Course Development Program at the University of Michigan for their feedback on earlier versions of this article. An earlier version of this article was presented at the 50th Annual Meeting of The Gerontological Society of America, November 15–19, 1995, Los Angeles, CA.

Received for publication March 20, 2001. Accepted for publication May 20, 2002.

	References

TOP Abstract Methods Results Discussion References

 

• Andrews, F. M., Withey, S. B. (1976). Social indicators of well-being: Americans' perceptions of life quality. New York: Plenum.
• Antonucci, T. C., Akiyama, H., Lansford, J. (1998). Negative effects of close social relations. Family Relations, 47,379-384.
• Baldwin, M. W. (1992). Relational schemas and the processing of social information. Psychological Bulletin, 112,461-484.
• Baltes, P., Mayer, K., (Eds.). (1999). The Berlin Aging Study: Aging from 70 to 100. New York: Cambridge University Press.
• Baron, R. M., Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51,1173-1182.[Medline]
• Baumeister, R. F., Wotman, S. R., Stillwell, A. M. (1993). Unrequited love: On heartbreak, anger, guilt, scriptlessness, and humiliation. Journal of Personality and Social Psychology, 64,377-394.
• Campbell, A., Converse, P. E., Rodgers, W. L. (1976). The quality of American life. New York: Russell Sage Foundation.
• Carstensen, L. L. (1992). Social and emotional patterns in adulthood: Support for socioemotional selectivity theory. Psychology and Aging, 7,331-338.[Medline]
• Carstensen, L. L., Gottman, J. M., Levenson, R. W. (1995). Emotional behavior in long-term marriage. Psychology and Aging, 10,140-149.[Medline]
• Carstensen, L. L., Isaacowitz, D. M., Charles, S. T. (1999). Taking time seriously: A theory of socioemotional selectivity. American Psychologist, 54,165-181.[Medline]
• Charles, S. T., Reynolds, C. A., Gatz, M. (2001). Age-related differences and change in positive and negative affect over 23 years. Journal of Personality and Social Psychology, 80,136-151.[Medline]
• Diehl, M., Coyle, N., Labouvie-Vief, G. (1996). Age and sex differences in strategies of coping and defense across the life span. Psychology and Aging, 11,127-139.[Medline]
• Fredrickson, B. L., Carstensen, L. L. (1990). Choosing social partners: How old age and anticipated endings make people more selective. Psychology and Aging, 5,335-347.[Medline]
• Hansson, R. O., Jones, W. H., Fletcher, W. L. (1990). Troubled relationships in later life: Implications for support. Journal of Social and Personal Relationships, 7,451-463.[Abstract]
• Ingersoll-Dayton, B., Morgan, D., Antonucci, T. C. (1997). The effects of positive and negative social exchanges on aging adults. Journal of Gerontology: Social Sciences, 52B,S190-S199.[Abstract]
• Kelley, H. H. (1979). Personal relationships: Their structure and processes. Hillsdale, NJ: Erlbaum.
• Labouvie-Vief, G., DeVoe, M. R. (1991). Emotional regulation in adulthood and later life: A developmental view. Annual Review of Gerontology and Geriatrics, 11,172-194.
• Lansford, J. E., Sherman, A. M., Antonucci, T. C. (1998). Satisfaction with social networks: An examination of socioemotional selectivity theory across cohorts. Psychology and Aging, 13,544-552.[Medline]
• Lawton, M. P., Kleban, M. H., Dean, J., Rajagopal, D., Parmelee, P. (1992). The factorial generality of brief positive and negative affect measures. Journal of Gerontology: Psychological Sciences, 47,P228-P237.
• Levitt, M. J., Silver, M. E., Franco, N. (1996). Troublesome relationships: A part of human experience. Journal of Social and Personal Relationships, 13,523-536.[Abstract]
• Okun, M. A., Keith, V. M. (1998). Effects of positive and negative social exchanges with various sources on depressive symptoms in younger and older adults. Journal of Gerontology: Psychological Sciences, 53B,P4-P20.[Abstract]
• Reis, H. T., Downey, G. (1999). Social cognition in relationships: Building essential bridges between two literatures. Social Cognition: Special Issue: Social cognition and relationships, 17,97-117.
• Rodgers, W. L., Andrews, F. M., Herzog, A. R. (1992). Quality of survey measures: A structural modeling approach. Journal of Official Statistics, 8,251-275.
• Roloff, M. E. (1981). Interpersonal communications: The social exchange approach. Beverly Hills, CA: Sage.
• Rook, K. S. (1997). Positive and negative social exchanges: Weighing their effects in later life. Journal of Gerontology: Social Sciences, 52B,S1-S3.[Medline]
• Smith, J., Goodnow, J. J. (1999). Unasked-for support and unsolicited advice: Age and the quality of social experience. Psychology and Aging, 14,108-121.[Medline]
• Sobel, M. E. (1982). Asymptotic intervals for indirect effects in structural equations models. In S. Leinhart (Ed.), Sociological methodology 1982 (pp. 290–312). San Francisco: Jossey-Bass.
• Veroff, J., Douvan, E., Kulka, R. (1981). The inner American: A self-portrait from 1957 to 1976. New York: Basic Books.
• Veroff, J., Kulka, R., Douvan, E. (1981). Mental health in America, 1957–1976. New York: Basic Books.
• Walen, H. R., Lachman, M. E. (2000). Social support and strain from partner, family, and friends: Costs and benefits for men and women in adulthood. Journal of Social and Personal Relationships, 17,5-30.[Abstract/Free Full Text]

This article has been cited by other articles:

				L. M. Miller, S. T. Charles, and K. L. Fingerman Perceptions of Social Transgressions in Adulthood J Gerontol B Psychol Sci Soc Sci, September 1, 2009; 64B(5): 551 - 559. [Abstract] [Full Text] [PDF]

				J.-H. Ha and B. Ingersoll-Dayton The Effect of Widowhood on Intergenerational Ambivalence J. Gerontol. B. Psychol. Sci. Soc. Sci., January 1, 2008; 63(1): S49 - S58. [Abstract] [Full Text] [PDF]

				K. L. Fingerman, P.-C. Chen, E. Hay, K. E. Cichy, and E. S. Lefkowitz Ambivalent reactions in the parent and offspring relationship. J. Gerontol. B. Psychol. Sci. Soc. Sci., May 1, 2006; 61(3): P152 - P160. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation