Tensions in the Parent and Adult Child Relationship: Links to Solidarity and Ambivalence

Kira S. BirdittLaura M. Miller and Karen L. FingermanEva S. LefkowitzAuthor information Copyright and License information DisclaimerThe parent-child relationship is one of the most long-lasting and emotionally intense social ties. Although often positive and supportive, this tie also includes feelings of irritation, tension, and ambivalence (Luescher & Pillemer, 1998). Indeed, parents and their children report experiencing tensions long after children are grown (Clarke, Preston, Raksin, & Bengtson, 1999; Fingerman, 1996; Morgan, 1989; Shaw, Krause, Chatters, Connell, & Ingersoll-Dayton, 2004; Talbott, 1990). There is a lack of information, however, regarding the topics that generate more intense tensions for parents and their adult children, and whether mothers, fathers, and their sons and daughters report tensions of similar intensity. In addition, it is unclear whether tensions are associated with the overall quality of the relationship. Describing variations in perceptions of tensions and whether tensions are associated with relationship quality among parents and adult children is crucial due to implications this tie may hold for overall quality of life, depressive symptoms, and health (Fingerman, Pitzer, Lefkowitz, Birditt, & Mroczek, in press; Lowenstein, 2007; Silverstein & Bengtson, 1997).The present study examined the topics that generate tensions for parents and their adult children to achieve two aims: 1) examine whether the intensity of tension topics varied by generation, gender, and age of adult children, and 2) assess associations between tension intensity, solidarity, and ambivalence.Go to:

Topics of Tension in the Parent and Adult Child Relationship

Broadly defined, interpersonal tensions are irritations experienced in social ties. Tensions may therefore range from minor irritations to overt conflict. The developmental stake and developmental schism hypotheses provide a useful framework for understanding why tensions exist in the parent and adult child relationship across the lifespan. According to the developmental stake hypothesis, parents are more emotionally invested in the relationship than are adult children and this generational difference remains consistent across the lifespan (Bengtson & Kuypers, 1971; Rossi & Rossi, 1990; Shapiro, 2004). Fingerman (1996; 2001) expanded on the developmental stake hypothesis with the concept of the developmental schism in which she proposed that tensions occur in the parent-child relationship due to discrepancies in the developmental needs of parents and their children. Two of the schisms that characterize the parent and adult child tie include independence (also referred to as care of self) and the importance placed on the relationship (Fingerman, 1996). These schisms may lead to different topics of tension and variations in perceptions of tensions between family members.Qualitative studies have described tension topics in the parent and adult child tie, establishing that tensions are common and cover a wide range of issues (Clarke et al., 1999; Fingerman, 1996; Morgan, 1989; Shaw et al., 2004; Talbott, 1990). These studies mainly focused on describing tensions between adults and their parents, without providing theoretical explanations of why tensions occur or the intensity of these topics. Furthermore, little is known about how perceptions of tensions vary within or between families or the implications of these tensions for relationship quality.We considered two theoretical categories of tension topics that may explain differences in relationship qualities between adults and their parents. Tensions may reflect either the parameters of the relationship or the behaviors of one of the individuals in the relationship (Braiker & Kelley, 1979; Fingerman, 1996). We refer to these tensions as relationship and individual tensions. Relationship tensions refer to how the dyad interacts and encompass issues of emotional closeness and cohesion or lack thereof. Individual tensions pertain to the behaviors of one member of the dyad and often have to do with independence or self-care. We used these categories to group the tensions found in the literature (Clarke et al., 1999; Fingerman, 1996; Hagestad, 1987; Morgan, 1989, Talbott, 1990). Relationship tensions include unsolicited advice, contact frequency, personality differences, child rearing, and past relationship problems. Individual tensions include job/education, finances, housekeeping, lifestyle, and health. This study included a quantitative measure of these tensions allowing for a comparison of parents' and adult children's ratings of the intensity of relationship and individual tensions. We defined intensity as the degree to which the particular topic causes tension.Go to:

Perceptions of Tensions by Generation, Gender, and Age

Developmental schisms and the tensions that result from them may vary by structural and developmental contexts. We consider three factors that are particularly salient in the parent and adult child relationship: generation, gender, and age (Rossi & Rossi, 1990). First, due to variations in developmental needs and investment in the relationship, parents and adult children may have different perceptions of tension topics. Clarke and his colleagues (1999) found that adult children reported more tension regarding communication and interaction style (relationship tension) compared to other topics. Fingerman (1996) found that daughters reported more tensions regarding feeling intruded upon (relationship tension) than did their mothers. These variations may result from the schism between parents and children in their views about the importance of the relationship (Bengtson & Kuypers, 1971; Fingerman, 1996). Because parents feel more invested in the relationship than do their children, they may report fewer tensions regarding fundamental dyadic interaction problems whereas their adult children may report more relationship tensions due to parents' efforts to have a closer relationship. For example, parents may make more demands to have greater contact or provide more unsolicited advice (relationship tension) than do their children.Because of these developmental schisms, perceptions of individual tensions may also vary by generation. Clarke and his colleagues (1999) found that parents reported more tensions regarding adult children's habits and lifestyle (how they spend their time and money, health related concerns) than other tension topics. Indeed, parents often expect their adult children to start careers, gain financial independence, get married, and have children. Parental well-being often depends on their children's success in these roles (Ryff, Lee, Essex, & Schmutte, 1994). Because parents experience strong desires for their children to achieve adult status and independence (Fingerman & Pitzer, 2007), they may perceive more intense tension regarding their adult children's independence and ability to care for themselves and report more intense individual tensions than their adult children.Much of the research thus far on parent and adult child tensions has focused on the mother-daughter tie or, when including fathers and sons, has not examined the findings separately by gender. Perceptions of tensions, however, may vary by gender. Relationships with daughters tend to be more emotionally intense involving more closeness and conflict (Fingerman, 2001; Smetana, Daddis & Chuang, 2003). Mothers also tend to have more intimacy and conflict with their children than do fathers (Collins & Russell, 1991). Overall, tensions may be more intense with mothers or with daughters than with fathers or with sons.Because of developmental changes and age-related variations in developmental schisms, the tensions that parents and adult children report may also vary by adult children's age (Fingerman, 1996). For example, families with older adult children may experience less intense tensions due to increases in adult children's autonomy. As adult children gain employment and start new relationships, parents may feel less concerned about their adult children's lack of independence. Having less contact as children grow older may also lead to less intense tensions (Akiyama, Antonucci, Takahashi, & Langfahl, 2003).Go to:

Implications of Tensions for Affective Solidarity and Ambivalence

Tensions most likely have implications for relationship quality. Two dimensions of relationship quality considered in the present study include affective solidarity and ambivalence. Affective solidarity refers to positive sentiments between family members including affection, emotional closeness, trust, and respect (Bengtson & Roberts, 1991; Bengtson, Giarrusso, Mabry, & Silverstein, 2002). Because of the developmental stake, parents tend to report greater affective solidarity with their offspring than their offspring report with them (Shapiro, 2004).In contrast to solidarity, intergenerational ambivalence includes conflicting feelings or cognitions that arise when social structures do not include clear guidelines for interpersonal behaviors or relationships (Connidis & McMullin, 2002). This sociological or structural ambivalence occurs when roles include contradictory expectations for behavior. Gender and generation are important structural determinants of ambivalence in the parent and adult child relationship. This structural ambivalence leads to psychological ambivalence, which is defined as experiencing positive and negative sentiments about the same relationship (Luescher & Pillemer, 1998). For example, a daughter may experience simultaneous feelings of love and irritation regarding her mother. It is most likely that factors in addition to social roles predict more or less ambivalence. In particular, certain tensions may be associated with ambivalence.The implications of tensions for affective solidarity and ambivalence may vary depending on the tension topic. In fact, a small number of studies indicate that relationship tensions in the parent and adult child tie are associated with less regard for the relationship and ambivalence. Fingerman (1996) found that mothers who felt excluded by their daughters reported less regard for the relationship. Likewise, daughters who reported tensions about their mothers' provision of unsolicited advice conveyed less regard for the relationship. Parents reported greater ambivalence when their children were too busy to spend quality time with them (Peters, Hooker, & Zvonkovic, 2006). Adult children reported greater ambivalence with parents who were rejecting and hostile earlier in life (Willson, Shuey, & Elder, 2003).Some research suggests a possible link between individual tensions and relationship quality. Fingerman (1996) found that mothers and daughters who attributed tensions to annoying behaviors/habits reported greater regard for the relationship. Parents tend to report greater ambivalence when their children have not achieved adult statuses (marriage, children, and employment) or have financial difficulties (Fingerman, Chen, Hay, Cichy, & Lefkowitz, 2006; Pillemer & Suitor, 2002; Willson, Shuey, Elder, & Wickrama, 2006). Parents also report more ambivalence when adult children provide help and support for their health problems (Spitze & Gallant, 2004). Similarly, adult children tend to report ambivalence when they anticipate parent caregiving and health concerns (Willson et al., 2003; Wilson et al., 2006). Relationship tensions most likely have a greater impact on overall perceptions of the relationship than do individual tensions because they have to do with fundamental tensions in how the dyad interacts.Go to:

Present Study

The present study sought to contribute to the intergenerational relationship literature in several ways. Unlike previous research which described tensions and/or excluded fathers and sons, we included ratings of tensions from mothers, fathers, and their young-adult and middle-aged sons and daughters. In addition, we examined links between tensions, affective solidarity, and ambivalence. This study examined two questions:1) Do perceptions of relationship and individual tensions vary by generation, gender of parent, gender of adult child, and age of adult child?Based on the developmental stake and schism hypotheses and the previous qualitative research regarding tensions, we predicted that parents would report more intense individual tensions than their adult children (due to concerns for their children's independence) and adult children would report more intense relationship tensions than their parents (due to irritations regarding parents greater investment in the tie; Clark et al., 1999; Fingerman, 1996). In addition, based on the literature regarding gender and parent-child conflict, we predicted that dyads with mothers or with daughters would report more intense tensions than dyads with fathers or with sons. We also predicted that families with older adult children would report less intense tensions due to increasing autonomy of adult children and decreased contact.2) Do relationship and individual tensions relate to relationship quality (affective solidarity and ambivalence)?Because relationship tensions have to do with overall dyadic interaction problems rather than the behaviors of one dyad member, we predicted that family members reporting more intense relationship tensions would report lower solidarity and higher ambivalence. We hypothesized that the association between relationship tensions and relationship quality would be greater than the association between individual tensions and relationship quality.Go to:

Method

Participants

Participants were from the Adult Family Study (Fingerman, Lefkowitz, & Hay, 2004) which included 158 (N = 474) family triads (mother, father, adult child) living in the greater Philadelphia metropolitan area. Participants completed individual telephone and videotaped interviews as well as paper and pencil assessments of tensions and relationship quality. Table 1 includes a sample description. The selection of participants involved a stratified sampling method by age of the adult child (22 to 33, 34 to 49), gender, and ethnicity of adult children. Recruitment of the majority of the sample occurred from a list purchased from Genesys Corporation (85%) and the remaining (15%) from convenience sampling (e.g., snowball, advertisements, and church bulletins). The convenience sampling method helped to increase the sample of African Americans and to gain more variance in relationship quality (Karney, Davila, & Cohan, 1995). The sampling procedures occurred in equal distributions across the stratification groups of gender, age, and ethnicity.

Parents completed measures regarding tensions and the quality of their relationship with the target child and the target child reported on each of their parents. The triads consisted of adult children (ages 22 to 49; 48% men) and their mothers and fathers (ages 40 to 84) who lived within 50 miles of one another. The investigators did not include those who co-resided in the study. One-third of the sample consisted of African Americans and the remaining included European American participants. A total of 61% of adult children were married and 87% of the parents were married to each other. A total of 4 fathers did not complete the tensions measure.

Measures

Tensions

Participants completed a 16-item measure assessing the extent to which they experienced tensions with their adult child/mother/father in the past 12 months regarding specific tension topics. We derived the tension topics from previous research regarding parent and adult child tensions (Clarke et al., 1999; Fingerman, 1996, Hagestad, 1987; Morgan, 1989; Talbott, 1990).The directions for the tension measure were as follows: “Below are issues that can cause tension between parents and their grown son or daughter. Tension means at least one person is bothered, even if they don't talk about it. Please indicate the extent to which each of these issues causes tension for you or your parent/child on a scale from: 1 (not at all), 2 (a little), 3 (some), 4 (quite a bit), to 5 (a great deal).” Thus, each tension topic received a rating of tension intensity.When tensions pertained to either the parent or the adult child (i.e., housekeeping, job, finances, health, and child rearing), we asked about tensions for parent and adult child separately. When tensions referred to dyadic interactions (e.g., personality differences) we asked about tensions only once.We examined the tensions as two scales: relationship and individual tensions. We created the scales by computing the mean of the items. To examine the validity of the two theoretical categorizations, eight social science researchers grouped the topics into the two categories. If at least six raters (.75) agreed on a topic, we considered it a valid example of the category. Relationship tensions included: contact frequency, personality differences, unsolicited advice, past relationship problems, and child rearing. The individual tensions included: finances, housekeeping, lifestyle, job/education, and health. All but two tensions, politics/religion and treatment of other, fit into the two categories. Cronbach's alpha demonstrated the statistical reliability of the two category theory; .75 for the relationship tensions and .86 for the individual tensions.

Affective solidarity

We assessed positive feelings about the relationship with the Bengtson Affective Solidarity index (Bengtson & Schrader, 1982) in which participants indicated how much they trust, understand, respect, feel affection toward, and feel their adult child/mother/father is fair from 1 (not well) to 5 (extremely well). The total score contains the sum of the five items (α = .85).

Ambivalence

As is common in the emerging literature pertaining to ambivalence (Fingerman et al., 2006; Willson et al., 2003), we created a measure of ambivalence with ratings of positive and negative aspects of the relationship originally used in the American Changing Lives study (Umberson, 1992). Although indirect assessments may reflect the perception of attitudinal differences between social partners (Priester & Petty, 2001), this approach is associated with other measures of ambivalence (Willson et al., 2003) and may be more effective than directly asking participants about their mixed feelings (Pillemer & Suitor, 2002). People may have difficulty estimating their mixed feelings but have less trouble estimating how positive and negative they feel (Luescher & Pillemer, 1998). The positive measure included two items (how much does he/she make you feel loved and cared for, how much does he/she understand you). The negative measure included two items (how much does he/she criticize you, how much does he/she make demands on you) rated from 1 (not at all) to 5 (a great deal). We used Griffin's similarity and intensity of components formula to calculate ambivalence [(positive + negative)/2-|positive − negative|] + 1.5 (Thompson, Zanna, & Griffin, 1995). Higher scores reflect greater ambivalence.

Generation, gender, and age

Generation and gender of parent included four categories: 1 (adult child reporting on father), 2 (adult child reporting on mother), 3 (mother reporting on adult child), and 4 (father reporting on adult child). We categorized families by age and gender of adult child as: 0 (ages 22 to 33), 1 (ages 34 to 49) and 0 (daughter), 1 (son).

Covariates

We coded ethnicity as 0 (European American), 1 (African American). Education consisted of the number of completed years of schooling (range = 7 to 18). Participants rated their health from 1 (poor) to 5 (excellent). Participants also completed the 10-item impression management scale of the Balanced Inventory of Socially Desirable Responding (Paulhus, 1984; 1991) to examine the extent to which participants presented a positive image rather than responding truthfully. Participants indicated agreement from 1 (not true) to 7 (very true) with items such as: “I sometimes tell lies if I have to” and “I have never taken things that don't belong to me” (α = .66).

Analysis Strategy

Because the data included multiple respondents within the same family (i.e., adult child, mother, and father) and children reported on both parents, we used multilevel modeling to account for the nested data (SAS PROC MIXED; Littell, Milliken, Stroup, & Wolfinger, 1996; Singer, 1998). PROC MIXED is ideally suited for this type of data in which there are multiple dependencies (e.g., between the family members and between responses of the same individual). In particular, we estimated models that included a random family effect (implying a correlation of observations within the same family) and a random parent/child within family effect that allowed mother and father reports on the same child and the child's reports on both parents to be correlated. The models included two levels in which upper level variables included characteristics of the family triad (e.g., age of adult child, gender of adult child, ethnicity) and the lower level variables included characteristics of the parent or adult child (e.g., parent's gender, ratings of tensions or relationship quality, socially desirable responding, health). Upper level characteristics varied between families whereas lower level variables varied within family. We considered ethnicity, education, and health as covariates because they are often associated with the quality of relationships (Birditt & Antonucci, 2007; Hill & Sprauge, 1999) and socially desirable responding because people may report less intense tensions to appear socially desirable (Birditt & Fingerman, 2003). For parsimony in presentation, covariates do not appear in the tables.The analyses included three steps. First, we conducted descriptive analyses. Next, we estimated models to assess whether the intensity of tension topics varied by generation, gender of parent, gender of adult child, and age of adult child. Thus, the scales for relationship topics and individual topics were the outcome variables. Finally, we estimated models to examine associations between tension topic intensity, affective solidarity, and ambivalence. Individual and relationship tensions were the predictors and the relationship quality scales were the outcomes.Go to:

Results

Results are presented in three sections. In the first section, we summarize the descriptive statistics. In the second and third sections we describe results of two sets of multilevel models examining whether tensions varied by characteristics of families, parents and their adult children, and whether tensions predicted variations in relationship quality.

Descriptives

We calculated descriptive statistics to provide an overall picture of the intensity of tensions aggregated across parents and adult children. Relationship tensions were slightly more intense (M = 1.89, SD = .76) than individual tensions according to a paired sample t-test (M = 1.83, SD = .75; t = 2.46, p <. 05). Although the mean intensity ratings were rather low, 94% of the participants reported at least a little tension regarding at least one of the relationship or individual tension topics. Correlations revealed a relatively high association between relationship and individual tensions (r = .71, p < .01). Although research suggests that only correlations above .80 are problematic (Licht, 1995), we considered whether there were problems due to multi-collinearity in later analyses. Table 2 includes all other correlations.

Tension Topics by Generation, Gender and Age of Adult Child

We used two-level multilevel models to examine whether the intensity of relationship and individual tensions varied by generation, gender, and age of adult child with two models: one for relationship tensions and one for individual tensions (Table 3). Because none of the interactions between age, gender and generation were significant, we removed them from the models. To examine variations by gender of parent and generation, we included four planned contrasts (mother vs. child, father vs. child, mother vs. father, child perception of mother vs. father).As expected, relationship tensions varied by gender (of parent and adult child) and by age of the adult child, but the age effect was opposite of the expected direction. Estimated means are provided in the text to help interpret the planned contrasts presented in Table 3. Offspring reported more intense relationship tensions with their mothers (M = 1.92, SE = .06) than with their fathers (M = 1.76, SE = .06). Families with daughters (M = 1.96, SE = .05) reported more intense relationship tensions than did families with sons (M= 1.79, SE = .06). Families with older adult children (M = 1.96, SE = .06) reported more intense relationship tensions than families with younger adult children (M = 1.79, SE= .06). Relationship tensions did not vary between mothers and fathers or by generation.Individual tensions varied by generation and gender of child. As hypothesized, mothers (M = 1.94, SE = .06) and fathers (M = 1.90, SE = .06) reported more intense individual tensions than did adult children regarding their mothers (M = 1.76, SE =.06) or fathers (M = 1.67, SE = .06). Families with daughters reported more intense individual tensions (M = 1.90, SE = .06) than did families with sons (M = 1.74, SE = .06). Individual tensions did not vary by gender of parent or age of adult child.

Tensions, Affective Solidarity, and Ambivalence

In the next analyses, we used two-level multilevel models to examine associations between relationship and individual tensions, affective solidarity, and ambivalence (Table 4). Because of the high correlation between individual and relationship tensions, we estimated models with the two scales together as predictors in the same models as well as the two scales as predictors in separate models to examine whether there were problems in the parameter estimation due to multi-collinearity. When the predictors were entered together, we found that individual tensions predicted greater affective solidarity but showed no association with ambivalence whereas relationship tensions predicted lower affective solidarity and greater ambivalence. We found different results when the predictors were entered separately, which indicates possible estimation problems because of multi-collinearity. Thus, we present the models with the scales entered separately. We estimated four models with affective solidarity and ambivalence as outcomes and individual tensions (Models 1 and 3) and relationship tensions (Models 2 and 4) as separate predictors. We included generation/gender of parent, gender of adult child, age of adult child, self-rated health, social desirability, education, and ethnicity as covariates.The models revealed significant associations between tensions, affective solidarity, and ambivalence. As hypothesized, more intense relationship tensions predicted less affective solidarity. More intense individual tensions also predicted lower solidarity. The models predicting ambivalence revealed that more intense relationship and individual tensions predicted greater ambivalence. Because previous literature indicated that relationship tensions predict lower relationship quality than do individual tensions, we compared the strength of the associations between relationship tensions and relationship quality with the strength of the association between individual tensions and relationship quality. In particular, we statistically compared the goodness of fit of the models with the −2 log likelihood estimations (Singer & Willett, 2003). The goodness of fit comparison involves subtracting the −2 log likelihood estimations of two models and examining the difference on a chi-square distribution with a degree of freedom of 1. The comparison of the goodness of fit indicators revealed that the models with relationship tensions as the predictor of affective solidarity and ambivalence possessed a significantly better fit than the models with individual tensions as the predictor of affective solidarity and ambivalence (p < .01).Go to:

Discussion

Using a unique sample of European and African American parent and adult child triads, this study supports the developmental schism hypothesis and contributes to the literature in several ways. The present study revealed that the majority of parents and adult children experienced at least a little tension with one another. In addition, the implications of those tensions for relationship quality varied by tension topics with some tension topics more highly associated with overall relationship quality than others. This study also indicates that structural or developmental differences in tension intensity are not always consistent across topics of tension. For example, although offspring reported more intense relationship tensions with mothers than with fathers, there were no differences between mothers and fathers in their perceptions of either relationship or individual tensions. There were also age differences in reports of relationship tensions but no such differences in reports of individual tensions.

Tensions by Generation, Gender and Age

Parents and adult children in the same families had different perceptions of tension intensity. Interestingly, in spite of the high correlation between relationship and individual tensions, the predictors of those tensions varied. Consistent with our hypothesis, mothers and fathers reported more intense individual tensions (e.g., finances, education, and health) than did their adult children. This finding is similar to the adolescent literature which suggests that parents are more upset by conflicts with adolescent children and that they tend to ruminate over these interactions more than their children do (Larson & Richards, 1994; Steinberg, 2001). This finding is particularly interesting because it indicates that parents are still more upset when children are older but that the generational difference is specific to individual tensions and not relationship tensions. This finding may reflect the developmental stake hypothesis or the concept of the developmental schisms in which parents' greater investment in the tie also may lead to parents' greater tension (Fingerman, 1996). Generational differences in perceptions of individual tensions may be a sign of parents' desires for their children to reach independent status. Parents are often concerned about their children's independence and continue to make efforts to socialize their children into adulthood (Fingerman & Pitzer, 2007). Parents may choose to attribute tensions to individual tensions rather than relationship tensions as a means of maintaining close relationships with their children.Although we predicted that adult children would report more intense relationship tension topics than their parents, we found no generational differences in reports of these tension topics. Tensions regarding the parameters of the relationship may result from ongoing interpersonal dynamics rather than structural variables. For example, these tensions may include problems established early in the relationship and continue across the lifespan. In addition, because these tensions have to do with perceptions of dyadic interactions it is possible that both individuals are more likely to perceive that there is a problem as compared to individual tensions that have to do with one of the individuals and not the dyad. For instance, parents may never communicate their irritations regarding their adult child's financial situation (individual tension) whereas a tension regarding personality differences may be more obvious when interacting with one another.This study also demonstrated that gender of child differences in the intensity of conflict typically found in studies of parents and adolescents appear to continue into adulthood (Smetana et al., 2003). Families with daughters reported that relationship and individual tensions were more intense than families with sons. This finding is consistent with research indicating that relationships with daughters are more emotionally intense than relationships with sons (Rossi & Rossi, 1990). Families may have more intense tensions with daughters because parents have more contact with daughters than with sons.We had predicted that mothers would report more intense tensions than fathers. However, there were no differences between mothers and fathers in their reports of tension intensity. The lack of variation between mothers and fathers in their perceptions is somewhat unexpected given that mothers often report greater investment and emotional intensity regarding their children than do fathers (Collins & Russell, 1991; Rossi & Rossi, 1990). This finding is also surprising given the adolescence literature indicating that parent's gender is often a more significant predictor of interaction patterns than child's gender (Hauser et al., 1987; McHale, Crouter, & Whiteman, 2003). Differences between mothers and fathers may dissipate across adulthood as parents grow older and the tension topics become less gender specific. For example, gender intensification theory suggests children experience an intensification of gender roles during adolescence, which coincides with greater sex typed parental socialization (Hill & Lynch, 1983). The empirical literature regarding this theory has revealed, however, that this intensification highly depends on the context (Alfieri, Ruble & Higgins, 1996; Crouter, Manke, & McHale, 1995). The issues in adulthood may therefore be more gender neutral (as compared to adolescence) and may elicit fewer gender differences in how parents view the relationship. Parents may consequently experience decreases in their gender specific parenting roles leading to more idiosyncratic relationships.Interestingly, although mothers and fathers had similar perceptions of tensions, offspring reported more intense relationship tensions with their mothers than with their fathers. This finding may be due to mothers' greater feelings of closeness with their offspring than fathers (Rossi & Rossi, 1990). Mothers may make more demands for closeness and may generally be more intrusive than fathers (Fingerman, 1996).We had predicted that families with older children would report less intense tensions overall due to age related increases in children's autonomy and decreasing contact frequency, but instead found that families with older adult children reported more intense relationship tensions. Consistent with the developmental schism hypothesis, parents and adult children may experience increasingly discrepant perceptions regarding the importance of their relationship with one another. Middle-aged children may be less invested in the parent-child tie than young-adult children because they are more likely to have formed their own families and experience multiple role demands. Thus, at the same time that parents become more invested in their relationship with their adult children, adult children may become increasingly less invested as they grow older creating even more intense relationship tensions.

Tensions, Affective Solidarity and Ambivalence

As hypothesized, relationship tensions were more highly associated with relationship quality than were individual tensions. Both relationship and individual tensions predicted greater ambivalence and less affective solidarity, but relationship tensions were more highly associated with relationship quality than individual topics of tension. These findings are important because they indicate that although the majority of parents and adult children experience at least a little tension, some tension topics may be more harmful to relationships than others. It is important for parents and their children to maintain good relationships across the lifespan for a number of reasons. For example, the quality of the relationship is associated with well-being and health (Fingerman et al., in press; Lowenstein, 2007; Silverstein & Bengtson, 1991) and the parent-child relationship is an important source of support and assistance for both parents and children (Silverstein & Bengtson, 1997; Shaw et al., 2004).In addition, it is interesting that tensions regarding particular topics may be detrimental to how parents and children view one another in general. Relationship tensions have to do with fundamental dyadic interaction problems. Thus, it makes intuitive sense that relationship tensions would have greater implications for overall negative opinions about the relationship. It is possible that these tension topics are detrimental because they represent longstanding tensions that are difficult to change. Indeed, researchers have found that negative childhood experiences are associated with ambivalent feelings in adulthood (Willson et al., 2003). Researchers have also found that unsolicited advice is associated with less regard for one another in the mother-daughter relationship (Fingerman, 1996). These more global relationship tensions may have broad influences on how parents and children view one another in general which may eventually have implications for support exchange, health, and well-being.The finding in the present study that individual tensions predicted lower relationship quality is consistent with research findings regarding ambivalence in the parent-child relationship. These studies examined links between structural variables (e.g., caregiving, marriage, career, finances) and ambivalence (Pillemer & Suitor, 2002; Willson et al., 2006) and found that parents and adult children report greater ambivalence when children have not reached adult statuses and independence. The individual tensions in this study may reflect parents' worries and irritations regarding their children's progress as adults. This study takes these findings a step further and indicates that parents and adult children who report these tensions also report more ambivalence and less affective solidarity. It is interesting that individual tensions appear to be less detrimental for relationship quality than relationship tensions. It may be that parents and children are less likely to communicate their irritations regarding individual tensions. For example, parents may experience irritations regarding their children's finances or education that they never communicate and thus these problems are less detrimental to the relationship overall. It is also possible that these tensions are less detrimental because they reflect worries or concerns for one another rather than fundamental relationship problems.

Limitations and Directions for Future Research

There are several limitations that should be addressed in future studies. This sample is somewhat unusual and may be highly functional because the majority of parents were still married to one another and willing to participate in an extensive survey. Thus, although we sought to develop a more comprehensive assessment of tensions, we may have underrepresented families that are less functional and that may experience more severe tensions such as neglect, abuse, chemical dependency, and psychological disorders. It is also unclear from the cross-sectional design whether relationship quality (ambivalence, affective solidarity) predicts changes in tension intensity or the reverse and future studies should examine these associations over time. In addition, scholars have criticized the indirect measurement of ambivalence because high scores may reflect a differentiated view on the topic or the perception of attitudinal disagreement between social partners rather than simultaneous positive and negative feelings (Priester & Petty, 2001). Future work should consider the implications of tensions for both indirect and direct assessments of ambivalence. Finally, further research should assess the types of coping strategies used in response to tensions. For example, some parents and adult children may avoid discussing a particular tension whereas others may argue. These differing styles may have unique implications for relationship quality (Birditt & Fingerman, 2005; Fingerman, 1998).This study advances the field by examining perceptions of tension topics among mothers, fathers, and adult children and the implications of those tensions for affective solidarity and ambivalence. This study is also highly unusual due to the large number of African American families included. The majority of studies in the family literature have only included European Americans. Thus, our findings are more generalizable to a diverse population. This study demonstrates the importance of considering multiple perspectives of relationships. Parents and adult children who are in the same relationship have different perceptions of the causes of tensions and those perceptions may have differential implications for relationship quality. Tensions are associated with greater ambivalence and lower affective solidarity. It is important for researchers and practitioners to be aware that the perceptions of tensions vary between families, within families, and within person in regards to different relationships. This study also indicates that structural and developmental variations in tensions depend widely on the topic of tension and that certain topics of tension may be more harmful to the relationship than others. These findings have important implications due to the long-lasting and far-reaching effects of the parent-child relationship on well-being, health, and support. Next steps include examining how parents and adult children cope with tensions and the implications of those tensions for relationship quality over time.Go to:

Acknowledgments

This study was supported by grant R01AG17916 “Problems between Parents and Offspring in Adulthood” and R01 AG027769, “The Psychology of Intergenerational Transfers” from the National Institute of Aging. We would like to thank Kristine Ajrouch and the Life Course Development Program for their helpful comments. We would also like to thank Kristina Hartman and Nicole Frizzell for their assistance with manuscript preparation and Brady West for his assistance with the statistical models.