Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Gerontol B Psychol Sci Soc Sci. Author manuscript; available in PMC 2010 December 8.
Published in final edited form as:
J Gerontol B Psychol Sci Soc Sci. 2003 July; 58(4): S225–S233.
PMCID: PMC2998990

Correlates of Emotional Support and Negative Interaction Among Older Black Americans



The study investigates the social and demographic, personality and social contact correlates of emotional support and negative interaction among older Black Americans.


The present analyses are based on the responses of 519 older African American respondents (55–96 years old) from the Americans’ Changing Lives Survey. Structural equation modeling (e.g., LISREL) was used to estimate the direct and indirect effects of sociodemographic variables (i.e., age, gender, education, income, marital status, and presence of children), social contact factors (i.e., frequency of face-to-face and telephone contact), and personality (i.e., extraversion and neuroticism) on emotional support and negative interaction with relatives and friends.


Gender, marital status, and extraversion were significantly associated with emotional support. Age, education, marital status, having a child, and neuroticism were identified as correlates of negative interaction.


The majority of older Black adults in this sample participated in emotionally supportive relations with relatives and friends, whereas negative social interactions were reported with less frequency. The distinctive group of predictors explaining emotional support and negative interaction suggests that they are different social occurrences. This analysis particularly underscores the role of personality factors in these processes.

The social networks and social relationships of older adults have raised fundamental questions in social gerontology since its inception as a field of study (e.g., disengagement and activity theories). Although much of the prior literature is predominated by a focus on emotional support among older adults (Ingersoll-Dayton, Morgan, & Antonucci, 1997), current theory and research incorporate notions of conflict and other negative occurrences (e.g., criticism or demands) as inevitable features of interpersonal relations with family members and others. Emotional support refers to exchanges between individuals that involve expressing care and affection, validating the person’s worth, and confiding (listening to problems). Negative interaction, in contrast, connotes social exchanges between individuals that are perceived by the recipient as nonsupportive, critical, harmful, or otherwise inconsequential to their needs (see Lincoln, 2000, for a review). Emotional support and negative social interactions are important antecedents to mental health and well-being outcomes (Finch, Okun, Barrera, Zautra, & Reich, 1989; Rook, 1984); emotional support has a beneficial impact on health and well-being, whereas negative interaction (e.g., being too demanding or critical) is detrimental to health. Although negative interaction occurs with less frequency than emotional support, it appears to be more important for predicting health and well-being (Finch et al., 1989; Ingersoll-Dayton et al., 1997; Rook, 1984).

African Americans and Social Interaction

The majority of African American elderly people (Barker, Morrow, & Mitteness, 1998; Johnson, 1999; Taylor & Chatters, 1986) are involved in strong and viable social support networks of immediate and extended family and nonkin associates (e.g., friends, fictive kin, and church members) and interpersonal relationships that involve exchanges of social support (e.g., emotional support and instrumental aid). However, African Americans are also likely to experience conflictual relations with the members of their support networks. Unfortunately, there is little research on the degree to which older Black adults, or minority elderly adults, more generally, are involved in unpleasant, conflictual, and other negative interactions with family and friends. The purpose of this article is to investigate the correlates of emotional support and negative interaction among older Black Americans. This analysis utilizes the older African American subsample of the Americans’ Changing Lives data set and examines demographic, social contact, and personality factors (extraversion, neuroticism) as correlates of emotional support and negative interaction.

Much of the literature on the informal support networks of African Americans focuses on general support exchanges, with little emphasis on negative social interaction. Further, the available research examines the consequences of emotional support and negative interaction with respect to health and well-being outcomes in defined groups. For example, Gant and Ostrow’s (1995) study of HIV-positive men found that negative interaction (e.g., feeling misunderstood) with social network members had a greater impact on measures of mental health than did social support and was highly correlated with tension or anxiety, depression, anger or hostility, and loneliness. Studies of pregnant and parenting adolescent mothers and their relationship with their parents (Davis & Rhodes, 1994; Davis, Rhodes, & Hamilton-Leaks, 1997) indicate that adolescents who receive more parental support (i.e., from their own mothers and fathers) have lower levels of depression (Davis et al., 1997). In contrast, relationship problems (i.e., negative interaction) such as disappointment, intrusiveness, and criticism are associated with psychological distress (Davis & Rhodes, 1994) and social adjustment problems (Davis & Rhodes, 1994). Further, low parental support and negative interaction with fathers is associated with higher levels of depression (Davis et al., 1997). Despite problems of small and specialized, non-probability samples, this research is a promising initial attempt at exploring the impact of negative social interaction on the supportive networks of Black Americans.

Theoretical Perspectives

Current approaches to examining social networks and interaction in later life (Ajrouch, Antonucci, & Janevic, 2001) are increasingly concerned with both emotional support and negative aspects of social interaction. Socioemotional selectivity theory (Carstensen, 1995; Lang, Saudinger, & Carstensen, 1998) specifically examines issues of social network size and the emotional quality of later life relationships. The theory posits that, because of awareness of one’s mortality, older persons are motivated to pursue social relationships that optimize short-term goals of emotional gratification, and hence they actively select emotionally “close” social partners in lieu of individuals who are less intimate. Although socioemotional selectivity results in declines in the overall size of the social network with advancing age, these networks are characterized as emotionally close with high levels of social integration. Comparable findings have been noted elsewhere in the literature among samples of the U.S. population (Antonucci & Akiyama, 1987), the German elderly population (Lang et al., 1998), and African American elderly population (Ajrouch, Antonucci, & Janevic, 2001), verifying that small, emotionally close networks are typical for the very old.

Demographic and Social Correlates of Emotional Support and Negative Interaction

Findings for age effects for various dimensions of social networks (i.e., size, composition, and contact) are generally consistent with socioemotional selectivity theory. Fung, Carstensen, and Lang (2001) found that among both Blacks and Whites, older and younger persons report comparable numbers of close social partners, but older persons have fewer peripheral social partners in their social networks. Furthermore, Lansford, Sherman, and Antonucci (1998) found that, across three age cohorts of nationally representative samples of Black and White adults, older persons tended to express greater satisfaction with the size of their informal networks, while at the same time they reported less frequent contact with networks overall. Compared with Caucasians, African American respondents had smaller networks with a higher proportion of family members and reported more contact with their networks. Taylor and Chatters (1986) found that more than a quarter of older Blacks indicated that emotional assistance (e.g., companionship, advice, and comfort) was the most important type of aid received from their support networks; best friends were important for providing companionship, whereas church members were important with respect to advice and encouragement. Finally, Johnson and Barer (1995) found that although Black and White respondents receive comparable levels of emotional support from children and grandchildren, Blacks are more likely to receive emotional support from extended kin.

In contrast, several studies involving older African Americans indicate either no age differences in level of contact and emotional closeness to extended kin networks (Taylor & Chatters, 1991) or declines in support from family and church members (Taylor, 1985; Taylor & Chatters, 1986) with advanced age. With regard to negative interaction, Rook’s (1984) study of a small sample of elderly widows failed to find any differences in negative interaction across a variety of indicators, including age. In contrast, other studies indicate that age is inversely associated with negative social interaction. Younger versus older (mean age of 68.9 vs. 73.6 years) stroke patients (Stephens, Kinney, Norris, & Ritchie, 1987), younger versus older (mean age of 62.6 vs. 68.8 years) caregivers (Pagel, Erdly, & Becker, 1987), and younger versus older (28–59 vs. 60–92 years) respondents (Okun & Keith, 1998) indicated more negative interactions with their family and social networks (i.e., spouse, children, other relatives, and friends).

Findings for other sociodemographic correlates are less consistent. With respect to gender, women’s greater overall investment in interpersonal and supportive relationships (so-called relationship work) suggests that they may receive higher levels of emotional support relative to men. In contrast, women’s greater social involvement could also mean that they are potentially more vulnerable to experiencing negative aspects of these relationships. The research evidence is equivocal and indicates that women are more likely than men to report negative interactions with their social networks (Pagel et al., 1987), Black women are more likely than Black men to receive support from adult children (Mutran, 1985), and older Black women and men are equally likely to receive expressive support from extended family members (Johnson, 1999). Education effects on social interaction indicate that persons with higher levels of education tend to receive greater amounts of assistance from family (Mutran, 1985), whereas other research indicates that education is unrelated to negative interaction (Pagel et al., 1987; Rook, 1984).

Marital and parental statuses represent primary and intimate personal relationships that facilitate feelings of social embeddedness and emotional closeness (Lang et al., 1998). Compositional differences in the social networks of elderly adults who are married and parents (i.e., as opposed to unmarried and childless) embody greater levels of socioemotional selectivity and maximize opportunities for emotionally satisfying relationships. Particular nonmarried statuses (i.e., divorced) may result in decreased levels of support from family (Taylor, 1986) and church networks (Taylor & Chatters, 1986). Further, being married and a parent provides opportunities for older individuals to be engaged in social interaction with others (in addition to spouses and children), is a link to other social resources such as extended family (Taylor, 1985, 1986; Taylor & Chatters, 1986) and church members (Taylor & Chatters, 1986), and is associated with larger helper networks (Chatters, Taylor, & Jackson, 1985, 1986) and more positive appraisals of affective dimensions of family life (Taylor & Chatters, 1991). Social contact (i.e., telephone and face to face) not only provides opportunities to engage in social interaction with family and friends but is associated with feelings of emotional closeness (Krause, Liang, & Keith, 1990; Taylor, 1985, 1986). Conversely, social contact is likely diminished for those personal relationships that are viewed as being detrimental.

Personality Correlates

Extraversion and neuroticism are two primary dimensions of personality that have received considerable attention in connection with social networks and relationships (e.g., Krause et al., 1990). Extraverts are characterized as individuals who are talkative, warm, and sociable, enjoy the company of others, prefer large groups, and actively seek out the opportunity to socialize. These attitudes and behaviors, in turn, create and maintain positive interaction and feedback from others. Extra-version is associated with larger networks, perceptions of support availability, and a greater propensity for social network formation and maintenance (Russell, Booth, Reed, & Laughlin, 1997). Among older adults, extraversion was associated with social contact (e.g., visits and contact by phone or letter) with family and friends (Krause et al., 1990) and larger social networks (Lang et al., 1998), but it was unrelated to perceptions of emotional closeness with the network (Lang et al., 1998). In contrast, persons with high levels of neuroticism are quiet and reserved, ineffective interpersonally, feel anxious in social situations, and withdraw or disengage from social interaction. Neuroticism is associated with conflictual relationships (Berry, Willingham, & Thayer, 2000), lower levels of support (Russell et al., 1997), and, among older adults (Lang et al., 1998), negative interaction (bivariate association).

The Proposed Model

Figure 1 illustrates the conceptual model of the proposed relationships between demographic and family status factors, personality, telephone and face-to-face contact, and emotional support and negative interaction. The notation used in the model is in accordance with that devised by Jöreskog and Sörbom (2001), in which the etas represent the latent or unobserved constructs, the xis represent the latent exogenous constructs, and the psis represent the correlation between the disturbance terms. The disturbance terms in the equation predicting extraversion and neuroticism (η1 and η2) are correlated. This specification suggests that extraversion and neuroticism are two dimensions of the same construct, personality, that may be related in some manner. Similarly, the disturbance terms predicting emotional support and negative interaction (η5 and η6,) are correlated. This specification suggests that the experience or perception of emotional support from network members is related, in some manner, to the negative interaction experienced or perceived. Correlating the disturbance terms in this manner makes it possible to consider the potential association between the factors without specifying a particular causal direction.

Figure 1
Proposed measurement model of sociodemographic factors, personality, social contact, emotional support, and negative interaction.

The proposed model is a reflection of theoretical suggestions from socioemotional selectivity theory, as well as empirical findings on the personality and sociodemographic correlates of social relationships. Focusing first on endogenous constructs, the model posits that personality is associated with frequency of social contact, emotional support, and negative interaction. Extraverted individuals are predicted to have more frequent telephone and face-to-face contact and to receive more emotional support from network members, whereas the reverse is anticipated for those high on neuroticism. Individuals who have more frequent social contact with members of their social network, both by means of the telephone and face to face, are expected to report more emotional support and have less negative interaction with network members, whereas the opposite is expected for persons with low levels of social contact.

Consistent with expectations from socioemotional selectivity theory, age status will exhibit a number of specific relationships with social contact and social relationship factors that reflect older persons’ tendency to be more selective in their social relations and to maximize these interactions. Accordingly, we anticipate that age will be positively associated with emotional support and inversely related to negative interaction with relatives and friends. However, as noted previously, despite generally high levels of satisfaction with social networks, older respondents tend to report less frequent contact (Lansford et al., 1998). Consequently, we expect that age will also be inversely associated with levels of telephone and face-to-face contact.

Finally, research in the area of personality suggests that differences in social networks and relationships may, in part, be explained by basic personality distinctions. In particular, prior research notes that older persons differ from younger individuals with respect to personality characteristics, such that older age is associated with higher levels of extraversion and lower levels of neuroticism. The model’s inclusion of age status and personality factors will provide an estimation of their independent effects on social contact and relationships and some assessment of the validity of the selectivity hypothesis within the context of individual factors that are known to be associated with social interaction. That is, we can determine, to some extent, whether observed differences in social contacts and relationships are attributable to socioemotional selectivity processes (which are associated with advanced age) or to differences in personality characteristics (i.e., extraversion or neuroticism) that are themselves known to be associated with social factors (e.g., network size, social contact, and conflict).

Previous research suggests that sociodemographic status factors are associated with personality characteristics and are important in terms of access to social resources and in patterning social relationships. With respect to gender, women are expected to report higher levels of emotional support and negative interaction, and social contact and neuroticism, but lower levels of extraversion, as compared with men. Education effects are less clear cut, but we anticipate that more education is associated with higher levels of emotional support and social contact but lower levels of neuroticism and negative interaction. Marital and parental statuses will be associated with model components in a similar manner; married persons and parents will demonstrate higher levels of emotional support, social contact, and extra-version and lower levels of negative interaction and neuroticism. Finally, because interactions with relatives and friends can reflect both beneficial and detrimental aspects, we anticipate that emotional support and negative interaction will have a weak, inverse relationship (Okun & Keith, 1998).

This investigation of the correlates of emotional support and negative interaction among older African Americans addresses several questions in the research literature. First, it examines demographic and social contact factors (e.g., face-to-face and telephone contact) as correlates of emotional support and negative interaction from relatives and friends. Second, it examines the influence of personality factors (neuroticism and extraversion) on emotional support and negative interaction and assesses their contribution to social interactions. Third, because the analysis is based on a large subsample of older African Americans from a major national sample, it provides greater sociodemographic variability and representativeness of findings. Finally, structural equation modeling techniques are used that provide the opportunity to specify a theoretically driven causal model of these relationships that assesses direct, indirect, and total effects of model components.



The Americans’ Changing Lives study is a multistage stratified probability panel survey of persons 25 years of age and older and who live in the coterminous United States. Face- to-face interviews were conducted in 1986, resulting in 3,617 completed interviews and a response rate of 67%. Blacks and persons aged 60 and older were oversampled at a rate twice that of Whites under the age of 60. This data set includes 1,174 African American respondents. The present analyses are based on the responses of 519 African American older respondents (55–96 years of age; M = 68.18 and SD = 8.14). Approximately 68% of the respondents are women, 38% are married, the average number of years of education is 8.73 (SD = 3.89), and 67% are parents.

Analysis Issues

Table 1 presents the items used to develop the model constructs depicted in Figure 1. All analyses were conducted by using covariance matrices as input and the maximum likelihood estimator in LISREL 8.51 (Jöreskog & Sörbom, 2001). The standardized factor loadings and measurement errors (indicated next to each item) provide information about the psycho-metric properties of the measures. Although there are no firmly established guidelines regarding cutoff points, researchers generally agree that items above .40 have acceptable psychometric properties. The data in Table 1 indicate that individual factor loadings were moderate to high in magnitude, ranging from .441 to 1.000.

Table 1
Factor Loadings for the Proposed Model of Correlates of Negative Interaction Among Older African Americans

There has been much discussion concerning the appropriate number of indicators for latent constructs. In general, the strategy for selecting items for latent variables involves selecting multiple observed variables that reliably and validly demonstrate good indicators of the construct in question (John & Benet-Martinez, 2000). Although there are no rules regarding the appropriate number of indicators per factor, recommendations range from 1 to 12 items (Marsh, Hau, Balla, & Grayson, 1998). Results from the current analyses indicate that the measured items have face validity and function within the model as would be theoretically expected (i.e., the hypothesized relations concerning them hold). Given that the items reflect the construct, they are related to other variables as expected under the theory, and the model fits well, we believe that we are tapping the constructs of interest.


Emotional support and negative interaction

Two indicators measuring emotional support from relatives and friends ask respondents the extent to which network members (a) make them feel loved and cared for and (b) are willing to listen to them discuss worries or problems. Two indicators of negative interaction with relatives and friends assess the extent to which they (a) make too many demands (e.g., interference or demands) and (b) are critical of the respondent and what they do (e.g., criticism or ridicule). Each item has five response categories, ranging from “not at all” to “a great deal,” with a high score corresponding to more emotional support or negative interaction.

Frequency of telephone and face-to-face contact

One indicator measuring frequency of telephone contact asked respondents to indicate how often they talk on the telephone with friends, neighbors, or relatives in a typical week. Response categories range from “never” to “more than once a day,” with higher scores representing more frequent telephone contact. The measure of frequency of face-to-face contact asked respondents to indicate how often they get together with friends, neighbors, or relatives to socialize. Response categories range from “never” to “more than once a week,” with higher scores reflecting more frequent contact.


Extraversion was measured with three indicators that reflect the degree to which respondents fell that they (a) take the initiative to make friends, (b) prefer to stay in the background on social occasions (reverse coded), and (c) are mostly quiet with others (reverse coded). Neuroticism was measured with three indicators that asked respondents to indicate the extent to which they (a) consider themselves moody, (b) felt fed up, or (c) were tense or high strung. Response categories for all items were “no,” “sometimes,” and “yes,” with a higher score on these indicators reflecting high levels of extraversion or neuroticism.

Sociodemographic controls

All analyses were conducted with the effects of age and education (coded continuously), gender (0 = male; 1 = female), marital status (0 = married; 1 = unmarried), and parental status (0 = no children; 1 = has children) taken into account.


The means, standard deviations, and range for the study variables are presented in Table 2. Consistent with previous research (e.g., Finch et al„ 1989; Rook, 1984), overall, positive interactions (e.g., emotional support) occur with greater frequency than negative interactions. Respondents reported, on average, that their relatives and friends made them feel loved and cared for (M = 4.18; SD = 0.97) and were willing to listen to their problems (M = 3.83; SD = 1.18) more often than their relatives and friends made too many demands (M = 1.46; SD = .85) and were critical of them (M = 1.60; SD = 0.97). Further, older African Americans reported a high frequency of telephone (M = 4.48; SD=1.58) and face-to-face contact (M = 3.95; SD= 1.67).

Table 2
Dependent and Independent Variables for Correlates of Negative Interaction: Ranges, Means, and Standard Deviations

Table 3 and Figure 2 present the completely standardized maximum likelihood parameter estimates for the proposed causal model. A significant but weak negative correlation between emotional support and negative interaction among relatives and friends (β =−.180; p < .05) was observed. An examination of the direct, indirect, and total effects provides a comprehensive picture of the process by which the model components are interrelated. Extraversion is positively associated with telephone (β= .205; p < .05) and face-to-face (β= .422; p < .01) contact and emotional support (β= .503; p < .01). Although extraversion leads to greater social contact and emotional support, there are no mediating relations among these constructs. The indirect effect, although statistically significant, is small and may be the result of the effects of other variables that were not included in the model. Neuroticism is related to negative interaction, such that high neuroticism scores are associated with negative interaction with relatives and friends (β = .421; p < .01). Table 4 presents the coefficients for the effects of exogenous variables on the latent constructs. With respect to age differences, persons of advanced age are more extraverted (β = .165; p < .05), less neurotic (β= −.1 17; p < .05), and report fewer negative interactions (β =−.279; p < .01) with relatives and friends, compared with their younger counterparts. Age influences face-to-face interaction both directly and indirectly; although face-to-face contact declines with age (β = 7minus;.1 14; p < .05; direct effect), this relationship is offset (i.e., mediated) by a positive indirect effect of age by means of extraversion (β = .059; p < .05). Compared with men, women report higher levels of neuroticism (β = .241; p < .01), more frequent telephone contact (β = .390; p < .001), and more emotional support (β =.195; p < .01) from relatives and friends. Although the data indicate a direct effect of gender on face-to-face contact (β = −.1 17; p; < .05), the total effect is reduced to insignificance (β= −.054;ns ) once the indirect effect by means of extraversion is taken into account. Further, the direct effect of gender on negative interaction (β= −.222;p < .05) is offset by its indirect effect by means of neuroticism (β= .125; p < .05). Although women report less face-to-face contact and fewer negative interactions with relatives and friends, this relationship is explained, in part, by the higher levels of neuroticism among women.

Figure 2
Estimated measurement model of personality, face-to-face and telephone contact, emotional support, and negative interaction.
Table 3
Completely Standardized Maximum Likelihood Parameter Estimates for the Correlates of Negative Interaction: Decomposition of Effectsa
Table 4
Completely Standardized Exogenous Effects of the Proposed Model

Persons with higher levels of education are more extraverted than their counterparts (β= .300;p < .001). Education has both direct and indirect effects on telephone contact and emotional support; higher levels of education are associated with more telephone contact (β = .221; p < .01), which is due, in part, to higher levels of extraversion (β = .058; p < .05; indirect effect) among older persons with more education. Although education has no direct effect on face-to-face contact, education operates through extraversion (β=.123; p < .05; indirect effect) to affect the frequency of face-to-face contact (β = .097; p; < .05; total effect). Older persons with more years of formal education also report lower levels of emotional support from network members (β = −.157; p < .05). However, this direct effect is mediated by the indirect effect by means of extraversion (β = .156; p < .01). Thus, the total effect of education on emotional support is not significant (β = −.001; ns). With respect to education and negative interaction, older African Americans with more education report more negative interaction with relatives and friends, compared with their less educated counterparts (β = . 1 17; p < .05; total effect).

With respect to marital status, older persons who are unmarried are less extraverted (β= −.144; p < .05) and report less frequent telephone contact with relatives and friends (β = −.174; p < .01; total effect) than their married counterparts. An indirect effect of marital status on emotional support by means of extraversion indicates that unmarried persons report less emotional support from network members, partially because of their lower levels of extraversion (β = − .153; p < .05; total effect). In addition, marital status is significantly related to negative interaction, such that unmarried persons report more conflict with relatives and friends (β = .121; p < .05; total effect). Finally, respondents who have children report lower levels of negative interaction with relatives and friends (β = −.184; p < .05; total effect), compared with those without children. The squared multiple correlations for the structural equations (the equivalent of R2) indicate that the set of factors explain 22.5% of the variance in emotional support and 30% of the variance in negative interaction.


This investigation found that older Black Americans were more likely to report emotionally supportive, as opposed to negative interactions with relatives and friends. This is consistent with the findings reported by Carstensen (1995) and others (Rook, 1984) on Whites. Further, the significant but weak (Okun & Keith, 1998) inverse relationship between emotional support and negative interaction indicates that these dimensions are not independent and that older African Americans can experience both supportive and conflictual relationships with relatives and friends simultaneously. The remainder of this section discusses the correlates of emotional support and then turns to a consideration of the factors associated with negative interaction. Because of space constraints, and because the primary focus here is on emotional support and negative interaction, findings related to the correlates of personality and social contact (i.e., telephone and face-to-face interaction) are not discussed.

Emotional Support

Emotional support from relatives and friends was associated with personality, demographic, family status, and social contact factors. Similar to research indicating that extraverts are more likely than nonextraverts to socialize with network members (Krause et al., 1990; Russell et al., 1997), extraverts in this sample reported that they were more likely to receive emotional support from relatives and friends. Neuroticism, in contrast, was unrelated to emotional support from relatives and friends. With regard to social and demographic status factors, age was unrelated to reports of emotional support from relatives and friends. Although predictions from socioemotional selectivity theory suggest that age is associated with enhanced support resources and positive appraisals of support relationships, it is important to note that research findings are equivocal as to the association between age and supportive relationships, noting both increases (Carstensen, 1995; Lang et al., 1998) and decreases in support with advanced age (Taylor, 1986). These discrepancies could be attributed to the different types of support measures that have been used in previous studies (e.g., social network size and relationship quality). Further, the null findings in the present analysis could have resulted from using a measure in which the focal support group combined relatives and friends. In other words, socioemotional selectivity processes may operate differentially in relation to specific groups (e.g., relatives vs. friends). Additional refinements in measurement may clarify both the relevant social support dimensions (i.e., network size and relationship quality) and the focal groups (i.e., kin vs. nonkin) involved in socioemotional selectivity processes. Although socioemotional selectivity theory does not focus on negative interactions, it is noteworthy that advanced age was associated with lower levels of reported problematic interactions with relatives and friends, suggesting that selectivity processes may involve divergent and independent dimensions of social relations.

Women in this study received more emotional support than their male counterparts, a finding that is consistent with research indicating that older Black women have more family and friends who provide emotional assistance (Barker et al., 1998), receive support from family members more frequently (Taylor, 1985), and have a larger pool of available helpers when confronted with a health problem (Chatters et al., 1985). Taken together, these findings suggest that older women are advantaged with respect to informal support networks (Antonucci, 1994). Persons who are not married were less likely to receive emotional support from relatives and friends than their married counterparts. The finding is consistent with that of Lang and colleagues (1998) and others (Krause et al., 1990) regarding the importance of social context (e.g., availability of nuclear family) for emotional closeness. However, previous work (Chatters et al., 1985, 1986) suggests that in order to compensate for the absence of a spouse, unmarried older Blacks were more likely than their counterparts to rely on broader social networks (i.e., siblings, friends, and neighbors) for assistance in the event of illness. Further, the lack of an effect of parenthood on emotional support is at odds with socioemotional selectivity theory and prior findings that adult children are important for securing support from family (Taylor, 1985, 1986; Taylor & Chatters, 1991). In this analysis, parents were no more likely than nonparents to receive emotional support from relatives and friends. As noted previously, discrepancies with prior research may reflect differences in the measures of support (e.g., potential helpers vs. enacted emotional support), as well as the focal support group (relatives and friends) used. Further, the present analysis specifically examined emotional support, whereas prior research (e.g., Taylor, 1985, 1986) assessed general support, which included instrumental assistance (e.g., help during illness and financial help) and emotional support.

Negative Interaction

Consistent with previous research, neuroticism was significantly related to negative interaction with relatives and friends (extraversion was unrelated to negative interaction). The relationship between neuroticism and negative interaction may reflect a complex interpersonal dynamic. Those with high levels of neuroticism may respond to ambiguous social situations (e.g., Larsen & Ketelaar, 1991; McCrac & Costa, 1991) in ways that generate negative social exchanges. For instance, because they are more likely to engage in self-recrimination and doubt and to be sensitive to ridicule, those with high levels of neuroticism are easily disturbed by awkward social situations and tend to perceive social interactions as negative. These reactions, in turn, may create situations that elicit negative interactions with others. Consistent with previous findings (e.g., Okun & Keith, 1998; Pagel et al., 1987; Rook, 1984; Stephens et al., 1987), older adults reported fewer negative interactions with relatives and friends, compared with their younger counterparts. Socioemotional selectivity theory would predict that because of increasing awareness of mortality, the very old actively choose social networks that emphasize intimate and positive emotional ties and reduce contact with peripheral relationships. The current investigation provided only limited evidence that gender is associated with negative interaction. Although the direct gender effect indicated that African American women experienced fewer negative interactions than men, this was reduced to insignificance once neuroticism and telephone and face-to-face contact were taken into account. This finding is supported by the work of others who report no gender differences in exposure to negative interaction (e.g., Pagel et al., 1987; Stephens et al., 1987).

Although previous studies report that education and negative interaction are unrelated (e.g., Pagel et al., 1987; Rook, 1984), the current findings indicated that older African Americans with more education were more likely to report negative interaction with relatives and friends. This finding is puzzling and suggests that there may be circumstances in which the dynamics of support exchanges and social interactions within social networks contribute to negative interactions. For example, persons who have more social and economic resources (e.g., higher educational level) and who are often expected to provide assistance to others may find that traditional norms governing social support to network members are burdensome. In these instances, support providers may become overtaxed if asked to repeatedly provide assistance. In addition, unbalanced support exchanges (e.g., lack of reciprocity) may generate negative interactions between network members (DiMatteo & Hays, 1981), including providers’ feelings of being unappreciated and taken advantage of and recipients’ feelings of indebtedness, embarrassment, and being a burden. Additional research into the dynamics of support exchanges may further clarify this relationship. Finally, with respect to marital and parental statuses, older African Americans who were married and had children experienced fewer negative interactions with members of their social network. This finding suggests that these social roles confer certain advantages on older adults with respect to supportive exchanges and interactions vis-à -vis their social networks (Krause et al., 1990). In particular, children are reliable sources of social support (Chatters et al., 1986), buffer the adverse effects of negative social interactions with relatives and friends (Okun & Keith, 1998), and broker supportive relationships and social interactions with others (Taylor, 1985, 1986). Further, Krause and colleagues (1990) noted that being married is an important correlate of the use of social support in later life.

Contrary to expectations (Krause et al., 1990; Taylor, 1985, 1986), the findings indicated that neither face-to-face nor telephone contact directly influenced emotional support or negative interaction among older Black adults. Previous multivariate analyses (Krause et al., 1990) examining personality and social contact factors (combining visits and contact by telephone and mail) in relation to emotional support from family or friends indicated that contact with family predicted received emotional support, whereas contact with friends and social extraversion were unrelated to support. Similarly, Lang and associates (1998) noted that extraversion is related to social network size but not the number of emotionally close relationships of older adults. In the present study, the significant effects for personality factors would seem to indicate that personality, rather than social contact, is of relatively greater importance for determining social interaction. Further investigation is needed to clarify the independent effects of personality and social contact factors on discrete measures of social networks and relationships. Finally, the pattern of findings in which extraversion predicted emotional support and neuroticism predicted negative interaction is worthy of note and suggests that dimensions of personality are domain specific with respect to their effects on social interactions (see Ingersoll-Dayton et al., 1997). That is to say, elements of personality that embody negative traits (i.e., neuroticism) predict negative social interactions, whereas positive aspects of personality (i.e., extraversion) are associated with positive social interactions.

Study Limitations

An interpretation of these findings should be considered within the context of the study’s strengths and limitations. First, there are acknowledged limitations associated with the measures of social interaction available in the Americans’ Changing Lives data set. As opposed to source-specific measures of negative interaction and emotional support from spouse, children, and parents, this analysis used global measures to assess relationships with nonimmediate family members and nonkin (i.e., relatives and friends combined). The use of the source-specific measures focusing on spouse, children, and parents, although providing more discrete information, would have limited sample selection criteria to respondents who are married, have adult children, or have a living parent. These stipulations would have severely restricted sample size and prohibited the use of more rigorous statistical techniques (structural equation modeling) in testing these relationships. The measures assessing interactions with relatives and friends were the most generous with respect to sample size considerations, and thus they were preferred for this initial investigation of these issues among older African Americans. Furthermore, given the prominence of extended family and nonkin relations for supportive exchanges among African Americans, it is important to examine these relationships involving relatives and friends, particularly among older adults.

This investigation only assessed two dimensions of personality. Given that the influence of personality on social interactions among older African Americans has received little attention in the literature, it is possible that other personality dimensions may be contributing to observed relationships. For example, there is evidence that individuals high in agree-ableness are more likely to interpret their interactions as less confrontational and perceive their networks as supportive (Graziano & Eisenberg, 1997) and that agreeableness is linked to positive and negative social interactions in ways similar to extraversion and neuroticism (Finch & Graziano, 2001). However, given the dearth of literature in this area that focuses on African Americans in general and older Blacks in particular, the current study provides initial empirical evidence that: at least two dimensions of personality are linked to social interaction in important ways.


The present analysis found that factors associated with emotional support and negative interaction were distinct. Future research should continue this line of investigation in distinguishing between the correlates of emotional support and negative interaction and, further, identify particular factors (e.g., intense support needs or poverty status) that potentially expose older adults to negative interactions with social networks. Despite the several innovations used in this study, the model tested is preliminary and in need of elaboration. Further investigations of the proposed relationships should consider other sources (e.g., spouse or children) and types of social support (e.g., instrumental assistance) and negative interaction (e.g., manipulation), as well as other important factors related to social interaction, such as stressful events.

Investigations of this sort will help us to more fully understand the processes that influence social support and negative interaction with the social networks of older African Americans.


This work was partially supported by a National Institute on Aging Minority Dissertation Research Grant in Aging (1R03-AG 18965) entitled “Social Interaction across the Life Course” awarded to Dr. Lincoln and by a National Institute on Aging Grant (R01 AG18782) entitled “Church-Based Assistance and Older Blacks.” The principal investigator was Linda Chatters; the coprincipal investigator was Robert .J. Taylor. Support was also provided to Robert Joseph Taylor as coinvestigator with the Michigan Center for Urban African American Aging Research (MCUAAAR; Grant P30 AG 15281); James S. Jackson was the principal investigator.

We thank Cameron McIntosh for his invaluable assistance with portions of this manuscript.


  • Ajrouch KJ, Antonucci TC, Janevic MR. Social networks among Blacks and Whites: The interaction between race and age. Journal of Gerontology: Social Sciences. 2001;56B:S112–S118. [PubMed]
  • Antonucci TC. A life-span view of women’s social relations. In: Turner BF, editor. Women growing older: Psychological perspectives. Thousand Oaks, CA: Sage; 1994. pp. 239–269.
  • Antonucci TC, Akiyama H. Social networks in adult life and a preliminary examination of the Convoy Model. Journal of Gerontology. 1987;42:519–527. [PubMed]
  • Barker JC, Morrow J, Mitteness LS. Gender, informal social support networks, and elderly urban African Americans. Journal of Aging Studies. 1998;12:199–222.
  • Berry DS, Willingham JK, Thayer CA. Affect and personality as predictors of conflict and closeness in young adults’ friendships. Journal of Research in Personality. 2000;34:84–107.
  • Carstensen LL. Evidence for a life-span theory of socioemotional selectivity. Current Directions in Psychological Science. 1995;4:151–156.
  • Chatters LM, Taylor RJ, Jackson JS. Size and composition of the informal helper networks of elderly Blacks. Journal of Gerontology. 1985;40:605–614. [PubMed]
  • Chatters LM, Taylor RJ, Jackson JS. Aged Blacks’ choice for an informal helper network. Journal of Gerontology. 1986;41:94–100. [PubMed]
  • Davis AA, Rhodes JE. African American teenage mothers and their mothers: An analysis of supportive and problematic interactions. Journal of Community Psychology. 1994;22:12–20.
  • Davis AA, Rhodes JE, Hamilton-Leaks J. When both parents may be a source of support and problems: An analysis of pregnant and parenting female African American adolescents’ relationships with their mothers and fathers. Journal of Research on Adolescence. 1997;7(3):331–348. [PubMed]
  • DiMatteo MR, Hays R. Social support and serious illness. In: Gottlieb BH, editor. Social networks and social support. London: Sage; 1981. pp. 117–149.
  • Finch JF, Graziano WG. Predicting depression from temperament, personality, and patterns of social relations. Journal of Personality. 2001;69(1):27–55. [PubMed]
  • Finch JF, Okun MA, Barrera M, Jr, Zautra AJ, Reich JW. Positive and negative social ties among older adults: Measurement models and the prediction of psychological distress and well-being. American Journal of Community Psychology. 1989;17:585–605. [PubMed]
  • Fung HH, Carstensen LL, Lang FR. Age-related patterns in social networks among European Americans and African Americans: Implications for socioemotional selectivity across the life span. International Journal of Aging & Human Development: Special Issue. 2001;52(3):185–206. [PubMed]
  • Gant LM, Ostrow DG. Perceptions of social support and psychological adaptation to sexually acquired HIV among White and African American men. Social Work. 1995;40(2):215–224. [PubMed]
  • Graziano WG, Eisenberg NH. Agreeableness: A dimension of personality. In: Hogan R, Johnson J, Briggs S, editors. Handbook of personality psychology. San Diego, CA: Academic Press; 1997.
  • Ingersoll-Dayton B, Morgan D, Antonucci TC. The effects of positive and negative social exchanges on aging adults. Journal of Gerontology: Social Sciences. 1997;52B:S190–S199. [PubMed]
  • John OP, Benet-Martinez V. Measurement: Reliability, construct validation, and scale construction. In: Reis HT, Judd CM, editors. Handbook of research methods in social and personality psychology. New York: Cambridge University Press; 2000. pp. 339–369.
  • Johnson CL. Family life of older black men. Journal of Aging Studies. 1999;13:145–160.
  • Johnson CL, Barer BM. Childlessness and kinship organization: Comparisons of very old Whites and Blacks. Journal of Cross-Cultural Gerontology. 1995;10:289–306. [PubMed]
  • Jöreskog K, Sörbom D. L1SREL. 8.51 and PRELIS 2.30 [Computer software] Chicago: Scientific Software International, Inc; 2001.
  • Krause N, Liang J, Keith V. Personality, social support, and psychological distress in later life. Psychology and Aging. 1990;5(3):315–326. [PubMed]
  • Lang FR, Staudinger UM, Carstensen LL. Perspectives on socioemotional selectivity in late life: How personality and social context do (and do not) make a difference. Journal of Gerontology: Psychological Sciences. 1998;53B:P21–P30. [PubMed]
  • Lansford JE, Sherman AM, Antonucci TC. Satisfaction with social networks: An examination of socioemotional selectivity theory across cohorts. Psychology & Aging. 1998;13:544–552. [PubMed]
  • Larsen RJ, Ketelaar T. Personality and susceptibility to positive and negative emotional states. Journal of Personality and Social Psychology. 1991;61:132–140. [PubMed]
  • Lincoln KD. Social support, negative social interactions, and psychological well-being. Social Service Review. 2000;74(2):231–252. [PMC free article] [PubMed]
  • Marsh HW, Hau KT, Balla JR, Grayson D. Is more ever too much: The number of indicators per factor in confirmatory factor analysis. Multivariate Behavioral Research. 1998;33:181–220.
  • McCrac RR, Costa PT., Jr Adding Liebe and Arbeit: The full five-factor model and well-being. Personality and Social Psychology Bulletin. 1991;17:227–232.
  • Mutran E. lntergenerational family support among Blacks and Whites: Response to culture or to socioeconomic differences. Journal of Gerontology. 1985;40:382–389. [PubMed]
  • Okun MA, Keith VM. Effects of positive and negative social exchanges with various sources on depressive symptoms in younger and older adults. Journal of Gerontology: Psychological Sciences. 1998;53B:P4–P20. [PubMed]
  • Pagel MD, Erdly WW, Becker J. Social networks: We get by with (and in spite of) a little help from our friends. Journal of Personality and Social Psychology. 1987;53(4):793–804. [PubMed]
  • Rook KS. The negative side of social interaction: Impact on psychological well-being. Journal of Personality and Social Psychology. 1984;46:1097–1108. [PubMed]
  • Russell DW, Booth B, Reed D, Laughlin PR. Personality, social networks, and perceived social support among alcoholics: A structural equation analysis. Journal of Personality. 1997;65(3):649–692. [PubMed]
  • Stephens MAP, Kinney JM, Norris VK, Ritchie SW. Social networks as assets and liabilities in recovery from stroke by geriatric patients. Psychology and Aging. 1987;2:125–129. [PubMed]
  • Taylor RJ. The extended family as a source of support to elderly Blacks. The Geronlologist. 1985;25:488–495. [PubMed]
  • Taylor RJ. Receipt of support from family among Black Americans: Demographic and familial differences. Journal of Marriage and the family. 1986;48:67–77.
  • Taylor RJ, Chatters LM. Extended family networks of older Black adults. Journal of Gerontology: Social Sciences. 1991;46:S210–S217. [PubMed]
  • Taylor RJ, Chatters LM. Church-based informal support among elderly Blacks. The Gerontologist. 1986;26:637–642. [PubMed]