THE ROLE OF PERCEIVED RELIGIOUS SIMILARITY IN THE QUALITY OF MOTHER-CHILD RELATIONS IN LATER-LIFE: DIFFERENCES WITHIN FAMILIES AND BETWEEN RACES

Abstract

Despite evidence of the importance of value similarity in predicting parent-adult child relations, little attention has been given to the unique role of religious similarity. Using 1,407 dyads nested within 390 families, we examine whether religious similarity predicts the quality of mother-child relations in later life, and whether the strength of this association differs by race. Consistent with our hypotheses, religious similarity was found to be an important factor in predicting both closeness and conflict, particularly in Black families. These findings suggest that it may be important to give greater attention to religion when studying patterns of interaction and support in the later years, especially among Black families.

The role of similarity in has long been emphasized in scholarship on interpersonal relations. Specifically, it has been argued that structural and value similarities are important in maintaining strong relationships both within and outside of the family (Feld 1982; Heider 1959; Homans 1950; Lazarsfeld and Merton 1954; Suitor and Keeton 1997; Suitor et al. 1995b; Wellman and Wortly 1990). Religious similarity, which intersects structure and values, has been shown to be an especially salient characteristic in explaining relationship quality (Lichter and Carmalt 2009; Louch 2000; Vaaler, Ellison, and Powers 2009), including between parents and offspring in childhood and young adulthood (Pearce and Axinn 1998; Stokes and Regnerus 2009). However, little attention has been directed toward examining the role of religious similarity in relations between middle-aged adult children and older parents. In the present paper we address this issue, using data on 1407 mother-child dyads nested within 390 families from the Within-Family Differences Study (Suitor and Pillemer, 2001). We focus on two specific research questions: 1) Does similarity of religious values predict the quality of mother-child relationships in later life; and 2) do race differences in the salience of religiosity moderate the relationship between religious value similarity and the quality of mother-child relations?

Similarity and Parent-Child Relations

The importance of similarity in interpersonal relations has been highlighted in both classic and contemporary scholarship. Heider (1959), Homans (1950), and Merton (1968) argued that similarity played a central role in establishing and maintaining cohesion in both dyads and larger groups. For more than five decades empirical research has demonstrated that individuals are more likely to develop and maintain strong relationships with others who are similar to them on important social dimensions, such as, marital status, parental status, educational attainment, age, and gender (Feld 1982; Lazarsfeld and Merton 1954; Suitor and Keeton 1997; Suitor, Pillemer, and Keeton 1995a; Teo et al. 2003; Wellman and Wortley 1990).

Sharing social statuses appears to enhance relationship quality primarily because such similarity increases the likelihood that associates will have parallel experiences, leading to shared values and perspectives (Suitor et al., 1995a; Thoits, 1992). These processes have been found to be important for relations with kin and nonkin (Rossi and Rossi 1990; Suitor and Pillemer 1992; Suitor, et al. 1995a; Umberson 1995). Studies focused on parents and adult children have revealed that perceptions of similarity are among the strongest predictors of several dimensions of relationship quality, including closeness, confiding, preferences for care, conflict, and ambivalence (Pillemer et al. 2007; Rossi and Rossi 1990; Suitor and Pillemer 2006).

In fact, research that has taken both social structural and perceived value similarity into consideration has found that parents’ perceptions of value similarity with their adult children have substantially greater influence in shaping the quality of parent-adult child relations than do structural similarities (Pillemer et al., 2007; Suitor and Pillemer, 2006, 2007). This is particularly surprising and noteworthy given that the failure of children to achieve normative adult statuses such as employment, marriage and parenthood often creates stress in their parents’ lives which can translate into more tense intergenerational relations (Greenfield and Marks 2006; Lachs and Pillemer 2004; Pillemer, et al. 2007; Sechrist et al. in press). Further, it appears that perceived similarity of values, not similarity as measured by the congruence between parents’ and children’s separate reports of their values, is the dimension of similarity that affects relationships (Pillemer, et al. 2007; Roberts and Bengtson 1990; Suitor and Pillemer 2006).

In sum, there is ample theoretical and empirical scholarship to argue that similarity, particularly value similarity, is salient in shaping the quality of parent-adult child relations. However, despite these strong findings of the effects of value similarity more generally, little is known regarding the particular influence of similarity of religious beliefs on parent-child relations, especially in later-life.

Religious Similarity and Parent-Adult Child Relations

Religion has been found to play an important function in family processes through reinforcing norms regarding family roles, while providing the social support and negative sanctions that facilitate these goals (Bartkowski and Xu 2000; D’Antonio, Newman, and Wright 1982; Marks and Dollahite 2001). Further, traditional religious values and religious similarity have been shown to enhance marital quality and stability (Curtis and Ellison 2002; Ellison, Bartkowski, and Anderson 1999; Lichter and Carmalt 2009; Mahoney et al. 2003; Vaaler et al. 2009; Williams and Lawler 2003).

Religious values and practices have also been shown to affect parent-child relationships. Specifically, high levels of religious service attendance and religious salience have been found to enhance the quality of parent-adult child relationships (Aquilino 1999; Brody et al. 1994; King 2003; Pearce and Axinn 1998; Sechrist, et al. 2007). However, few studies have examined the influence of religious similarity on parent-child relations; and those that have focused on these relationships earlier in the life course. These studies found that similarity in level of participation and religious salience were strongly related to relationship quality between parents and children in adolescence and early adulthood (Pearce and Axinn 1998; Stokes and Regnerus 2009). Consistent with other research comparing the effects of similarity of values with similarity of structural positions (Pillemer et al., 2007; Suitor and Pillemer, 2006, 2007), these studies found similarity of religious denomination had little or no effect on relationship quality, relative to similarity of practices and values.

Despite the evidence that religious values and perceptions of value similarity shape the quality of other family relations, and the high salience of religiosity among older Americans (Idler et al. 2003), there has been little attention to the role of religious value similarity in later-life family relations. In fact, only two studies have broached this subject. Myers (2004) provided evidence that congruence of both religious participation and salience were important for support exchanges between parents and adult children. However, Rossi and Rossi (1990) found no consistent relationship between similarity of religious denomination and the quality of parent-adult child relations. This pair of findings is consistent with examinations of relationship quality for young adult children and their parents in that similarity of religious participation and values were important, whereas denomination was not. We believe that this is most likely because similarity of religious denomination may have effects on day-to-day activities and observances that are less influential for parents and adult children, whose day-to-day lives are typically less interdependent than at earlier stages of the life course. Further, in relationships in which there is not regular daily or weekly contact, value similarity plays a much greater role than does structural similarity (Suitor et al. 1995b). Thus, we believe that it is very likely that mothers will report greater closeness and less conflict with children whom they perceive hold religious beliefs more similar to their own.

Race, religion, and parent-adult child relations

Up to this point we have discussed the role of similarity in interpersonal relations without taking context into consideration. However, the importance of particular dimensions of value similarity may be context specific. Most relevant to the question at hand, we suggest that the effects of perceived similarity of religious values on parent-adult child relations may vary by race. Specifically, because religion plays a greater role in the lives of Blacks than Whites both at a cultural and individual level, we believe that perceptions of religious similarity may play a more important role in parent-adult child relations in Black than White families.

There are many indications of the larger influence of religion in the lives of Black than White Americans. First, the church is one of the most central institutions among Blacks in America, playing an important role in both the heritage and the everyday lives of Black individuals (Taylor, Thronton and Chatters 1987). Historically, Blacks have been excluded from many secular outlets for social interaction, leadership, and achievement due to slavery and segregation. Therefore, scholars contend that the church became the central institution for social activity and support within the Black community primarily because it was the only institution whose leadership was Black (Billingsley 1999; DuBois 2000; Lincoln and Mamiya 1990).

Further, the Black church has emphasized particular religious values linking these to the history of the Black community, creating a unique belief system (Cone 1975; DuBois 2000; Paris 1995). For instance, the Black church highlights messages of justice and equality in the sight of God; which have helped Black individuals to cope with racial adversity and maintain an important collective pride and identity despite the historical marginalization of their community (Cone 1975; Lincoln and Mamiya 1990; Maynard-Reid 2000; Paris 1995; Shenk, Zablotsky and Croom 1998). This link between religion and identity creates a bond between the Black individual and religion that is unique to Black culture (Billingsley 1999). Empirical evidence has shown that, in fact, Black individuals report greater religiosity on a variety of dimensions including attendance, private practices and salience (Levin, Taylor and Chatters 1994).

Religion has also been shown to have a greater influence on other aspects of Black individuals’ lives than for White individuals. For instance, religious participation is more strongly related to physical and psychological well-being for Blacks than Whites (Ferraro and Koch 1994; Krause 2003; 2004).

Finally, and of particular relevance to this study, there is evidence that religion plays an especially strong role in intergenerational relations in Black families. For example, Black parents are more likely than their White counterparts to have been influenced by their own parents about religion and religious values (Krause and Ellison 2007). Further, older Black family members place a high value on the transmission of religious values (Ellison 1997; Weddle-West 2000). Given that Black women report especially high levels of religious salience and service attendance (Chatters, Taylor, and Lincoln 1999; Taylor 1988), and have been shown to report higher levels of religiosity across behavioral and spiritual dimensions compared to White women, (Levin, Taylor, and Chatters 1994), religious similarity may be of particular importance to Black mothers in comparison to White.

Taken together, these findings suggest that religious similarity between mothers and their adult children may be more important in Black than White families. Thus, we anticipate that race will moderate the relationship between religious value similarity and the quality of mother-adult child relations, with stronger findings in Black than White families.

Summary and Hypotheses

In sum, given the importance of religiosity in the lives of older individuals, and the role of value similarity in later-life parent-child relations, we hypothesize that that similarity of religious values will be associated with greater closeness and lower conflict between mothers and adult children. Further, based on the greater importance of religion in Black culture, we hypothesize that religious value similarity will be a stronger predictor of relationship quality in Black families than White families.

METHODS

Study Design

Design Goals

The data for this study were collected as part of the Within-Family Differences Study (WFDS 1 and 2) conducted by Suitor and Pillemer between 2001 and 2009. The project was designed to provide data on within-family differences in parent-adult child relations in later life. Suitor and Pillemer’s approach was similar to those that have been used by developmental psychologists such as McHale and colleagues (McHale et al. 1995) in studying within-family differences in earlier stages of the life course. The original design involved selecting a sample of mothers 65–75 years of age with at least two living adult children and collecting data from mothers regarding each of their children. A further decision was made to include only community-dwelling mothers in the sample to reduce the likelihood that the women would be in need of extensive caregiving, thus, allowing them to study relationships outside of the context of caregiving. The second wave of this study was designed to provide data on changes in parent-child relationships, focusing particularly on changes in within-family differentiation and patterns of support.

Sampling

Beginning in the summer of 2001 Suitor and Pillemer used Massachusetts city and town lists as the source of the original WFDS sample. With the assistance of the Center for Survey Research (CRS) at the University of Massachusetts Boston, Suitor and Pillemer drew a systematic sample of women ages 65–75 with two or more children from the greater Boston area (see Suitor and Pillemer 2006 for a more detailed description of the sampling procedures for wave 1). The wave 1 sample consisted of 566 mothers, which represented 61% of those who were eligible for participation.

For the follow-up study, Suitor and Pillemer, with the assistance of CSR staff, attempted to re-contact each mother who participated in the original study. Data collection occurred between May of 2008 and January of 2010. In the second wave of the study, 411 mothers were interviewed, resulting in a response rate of approximately 80% taking into consideration both valid responses and deaths among mothers. This paper uses data on the 1,407 mother-child dyads nested within 390 families from the second wave of the WFDS for which there were complete data on all study variables.

Sample Characteristics

Mothers’ Characteristics

Table 1 presents the mothers’ demographic characteristics for the sample of mothers who participated at T1 and T2. At T2, the mothers were between 72 and 82 years of age (mean = 76.8; SD = 3.1). Forty-one percent of the mothers were currently married, 45% were widowed, and 14% were divorced or separated. Nineteen percent of the mothers had completed less than high school, 44% had completed high school, 13% had completed some college and 24% had a college degree. Forty-five percent of the women were Catholic, 39% were Protestant, and 16% were another religion or said that they had no religious denomination. Seventy-six percent of the mothers were White; twenty-four percent were Black. The number of living children of women in the sample ranged from two to thirteen (mean= 3.7; SD = 1.7).

Table 1.

Sample Characteristics

Mothers’ Characteristics (N=390)
Age in years (s.d)	76.8 (3.1)
Race (in %)
Black	24
White	76
Marital Status (in %)
Married	41
Separated/Divorced	14
Widowed	45
Education (in %)
Less Than High School	19
High School Graduate	44
At Least Some College	13
College Graduate	24
Number of Children (s.d)	3.7 (1.7)
Religion (in %)
Catholic	45
Protestant	39
Other	16
Children’s Characteristics (N=1407)
Age in years (s.d)	49.4 (5.8)
Daughters (in %)	52
Married (in %)	68
Parents (in %)	74
Education (in %)
Less Than High School	6
High School Graduate	31
Some College	12
College Graduate	50

Adult Children’s Characteristics

Table 1 presents the adult child characteristics for the sample. The adult children in the sample were between the ages of 28 and 68 (mean = 49.4, SD = 5.8). Fifty-two percent of the adult children were daughters. Sixty-eight percent were married. Six percent of the adult children had completed less than high school, 31% were high school graduates, 12% had attended some college, and 50% were college graduates. Seventy-four percent of the adult children had children of their own.

Measures

The item regarding religious similarity is only available in the T2 data; therefore, measurement of the primary dependent and independent variables are from the second wave of the study. The only variables used from T1 are demographic characteristics that would be highly unlikely to change from T1 to T2, such as race, gender, and educational attainment. Wave 2 data were used for characteristics that might be subject to change between waves, such as marital status.

Dependent Variables

Following the lead of several prominent studies of intergenerational relations (Aquilino 1999, Giarrusso et al. 2005; Rossi and Rossi 1990; Umberson 1989), we have included both positive and negative dimensions of relationship quality in the analysis.

Affection

The measure of affection is a scale constructed from the following two questions: 1) “Using any number from 1 to 7, where 1 is very distant and 7 is very close. What number would you use to describe the relationship between you and [child’s name] nowadays?”; and 2) How often, very often, fairly often, sometimes, rarely or never does [child’s name] make you feel loved and cared for?” Consistent with other studies using these measures (Lawton, Silverstein, and Bengtson 1994; Ward, 2008; Ward, Spitze, and Deane 2009), the distributions were highly skewed; for this reason, we collapsed the lowest categories of each item, so that the scores of both range from 1–4. Then we combined the two items to create a measure of affection which ranged from 2–8 (mean=6.76, s.d.=1.60, α=.74)

Conflict

The conflict measure is also a scale created using the following two items: 1) “Sometimes no matter how close we may be to someone, the relationship can also at times be tense and strained. Using any number from 1 to 7, where 1 is not at all tense and strained and 7 is very tense and strained, what number would you use to describe how tense and strained the relationship between you and your mother is nowadays?”; and 2) “How often, very often, fairly often, sometimes, rarely or never, would you say the two of you typically have disagreements or conflicts?” Although these items were less skewed than the closeness items, we felt it important to be consistent across measures. Therefore, the response categories were also transformed into a 1–4 range for both variables, after which they were combined to create the conflict measure which ranged from 2 to 8 (mean=4.14, s.d.=1.77; α=.64).

The correlation between the affection and conflict scales is −.53.

Independent Variables

The independent variable of interest is similarity of religious values. Mothers were asked, “Parents and children are sometimes similar to each other in their views and opinions and sometimes different from each other. Would you say that you and [child’s name] share very similar views, similar views, different views, or very different views on religion?” This measure was coded 4=very similar views to 1=very different views. Table 2 presents the distribution of our main independent variable, religious value similarity, for the full sample and for the White and Black sub-samples. In all samples, most mother-child dyads have similar or very similar views on religion. Slightly more Black mother-child dyads have similar views compared to the White dyads, but this difference was not statistically significant.

Table 2.

Distribution of Perceptions of Religious Similarity in the Full Sample and by Race

Religious Similarity	Full-Sample (N=1407 in 390 families)	White (N=1033 in 296 families)	Black (N=374 in 94 families
Very different-1	7.5	7.2	8.3
Different-2	21.6	22.6	19.0
Similar-3	40.2	38.6	44.7
Very Similar-4	30.7	31.7	28.1
Mean (s.d.)	2.94 (.90)	2.95 (.91)	2.93 (.89)

Race

Race was measured at T1. Respondents were asked two questions concerning their race and ethnic background: “Are you of Hispanic or Latino origin?” and “(In addition to being Hispanic) What is your race? Select one or more of the following - White, Black or African American, Asian, American Indian or Alaska Native, Native Hawaiian or other Pacific Islander or Other.” For this analysis we used only those respondents who reported White or Black/African American. If Black and any other race/ethnicity was reported, respondents were coded as Black. The present analyses were restricted to Black and White mothers because the number of Asian and Latina mothers in the study was too small to allow for meaningful comparisons with Black and White mothers (n=3 and n=7, respectively).

Control Variables

Several factors that have been found to predict parent-adult child relationship quality were included as controls. These factors were both mother-level and child-level items. These included family size, mother’s age, educational attainment and religious denomination; and child’s age, gender, marital status, and educational attainment. With the exception of gender, educational attainment and age, data on all sociodemographic characteristics were measured at T2.

Family size was measured by the number of living adult children at wave 2. Mothers’ age was calculated by adding seven years to the age of the mother given at wave 1, the age range for the mothers is 72–82 (mean=76.8. s.d. 3.1). To measure mothers’ educational attainment, mothers were asked, “What was the highest grade in school that you completed?” Mothers’ education was coded 1=less than high school, 2=high school graduate, 3=some college and 4=college graduate. At wave 1 mothers were asked, “What is your religious preference -Catholic, Protestant, Jewish, or something else?” For this analysis, we created a set of dummy variables identifying mothers as Catholic, Protestant, or Other; in all analyses Other was the referent category.

Child’s age was calculated by adding seven years to the age of each child provided by their mother at T1. The age range for children was 28–68 (mean=49.4, s.d. 5.8). To ascertain the children’s marital status, the mothers were asked, “Is [child’s name] married or living with a partner as married?”; this is coded as 1=married and 0=unmarried. To measure child’s educational attainment, mothers were asked, “What was the highest grade in school that [child’s name] completed?” Child’s education was coded 1=less than high school, 2=high school graduate, 3=some college and 4=college graduate. Gender of child was coded 1=female, 0=male.

Plan of Analysis

Because there are multiple mother-child dyads nested within families, we used multilevel modeling to analyze these data. This procedure allows for non-independence among the cases and adjusts for the correlated error structures. Because our families only range from 2–13 adult children, the groups are not large enough to obtain reliable estimates when using interaction terms across levels. In circumstances where there are small numbers of cases in each group, it is recommended that random intercept models are used (Raudenbush and Bryk 2002). Therefore, to examine race differences in the effects of religious similarity on relationship quality, we ran separate analyses by race for each of the dependent variables and compared the similarity coefficients across models (Paternoster, Brame, Mazerolle, and Piquero, 1998):

$$\text{t}=\frac{{\text{b}}_1-{\text{b}}_2}{\surd ({{\text{SEb}}_1}^2+{{\text{SEb}}_2}^2)}$$

All of the analyses were conducted using SPSS 17.

RESULTS

Positive Relationship Quality

Table 3 presents the findings for mothers’ affection with their adult children. The first column shows the findings using the full sample. Consistent with our hypothesis, this model shows that perceptions of religious similarity are positively related to affection. The middle and last columns in Table 3 present the models for the White and Black subsamples, respectively. These models show that religious similarity predicts affection for both Black and White mothers. Further, as hypothesized, tests of the difference between the coefficients by race indicated that the religious similarity coefficient was significantly larger for the Black than the White subsample (p<.05, one-tailed).

Table 3.

Mixed Models Predicting Positive Relationship Quality in Mother-Adult Child Relations

	Full Sample (N=1407 in 390 families)			White Sub-Sample (N=1033 in 296 families)			Black Sub-Sample (N=374 in 94 families)
Predictors	Estimate	SE	df	Estimate	SE	df	Estimate	SE	df
Intercept	3.314**	1.45	369.03	2.639**	1.65	278.47	3.670**	3.12	90.02
Child/dyad Characteristics
Religious similarity	.410**	.04	1377.78	.361**	.05	1020.56	.524**a	.09	354.81
Age	−.017*	.01	1318.60	−.017	.01	977.58	−.017	.01	338.37
Female	.295**	.07	1253.85	.405**	.09	926.24	.002	.15	320.38
Education	.079	.05	1386.80	.091	.06	1027.86	.070	.09	357.90
Marital Status	.226**	.08	1279.30	.247**	.10	949.45	.165	.15	331.89
Mother Characteristics
Age	.033	.02	410.85	.043	.02	318.953	.018	.04	94.87
Catholic	..555**	.18	416.24	.514**	.19	312.70	.576	.68	89.54
Protestant	.425*	.19	409.80	.254	.20	307.30	1.133	.57	89.35
Family size	−.026	.03	300.64	−.016	.04	220.59	−.045	.06	78.86
Mother’s education	−.089*	.04	422.80	−.098*	.04	319.92	−.036	.08	100.819
Black	.078	.17	396.55	--	--	--	--	--	--
Model Statistics
BIC	5014.313			3702.536			1361.653
AIC	4941.025			3638.376			1311.170

^*p<.05.

^**p<.01.

^aSignificantly different from White coefficient at the .05 level (one-tailed).

Negative Relationship Quality

Table 4 presents the findings for negative relationship quality. As shown in the first column, religious similarity predicts lower levels of conflict. The final two columns present the results for the White and Black subsamples. These models indicate that religious similarity is associated with lower levels of conflict for both Black and White mothers. However, as in the case of positive relationship quality, tests of the difference between the similarity coefficients revealed that religious similarity is significantly more strongly related to negative relationship quality for Black than White mothers (p<.05, one-tailed).

Table 4.

Mixed Models Predicting Negative Relationship Quality in Mother-Adult Child Relations

	Full Sample (N=1407 in 390 families)			White Sub-Sample (N=1033 in 296 families)			Black Sub-Sample (N=374 in 94 families)
Predictors	Estimate	SE	df	Estimate	SE	df	Estimate	SE	df
Intercept	7.348**	1.69	378.67	7.923**	1.89	288.37	5.683**	3.89	90.24
Child/dyad Characteristics
Religious similarity	−.306**	.05	1356.36	−.248**	.05	1000.10	−.477**a	.10	349.53
Age	..005	.01	1286.84	.002	.01	948.98	.010	.02	331.04
Female	.118	.08	1225.02	.116	.09	900.48	.079	.17	314.23
Education	−.193**	.05	1380.25	−.160**	.06	1016.61	−.24**	.10	358.89
Marital Status	−.604**	.09	1248.58	−.800**	.10	922.49	−.198	.17	324.43
Mother Characteristics
Age	−.015	.02	417.10	−.022	.02	324.91	.010	.05	94.71
Catholic	−.092*	.17	379.05	−.386	.21	315.41	−.712	.85	89.55
Protestant	−.556*	.22	411.36	−.453	.23	310.55	−.949	.71	89.27
Family size	−.043	.04	320.20	−.077	.05	240.43	.031	.07	80.66
Mother’s education	.050	.04	429.53	.055	.05	326.77	.011	.10	100.82
Black	−.025	.20	394.23	--	--	--	--	--	--
Model Statistics
BIC	5258.336			3821.365			1473.350
AIC	5185.048			3757.205			1422.867

^*p<.05.

^**p<.01.

^aSignificantly different from White coefficient at the .05 level (one-tailed)

Religious denomination also appears to influence relationship quality. In the full sample, Protestant and Catholic mothers appear to have more positive and less negative relations with their adult children; however, these differences were not observed in the sub-sample analyses, likely due to the smaller sample sizes at the family level.

In sum, these findings provide consistent support for both of our hypotheses. Specifically, mothers reported greater relationship quality with children who they perceived as having more similar religious values; further, the effects of religious similarity on relationship quality were greater in Black than White families.

Discussion

Our goals in this paper were to examine the role of religious value similarity in parent-adult children relations, and to explore whether these patterns differed by race. The findings that we have presented demonstrated that mothers’ perceptions of religious value similarity to their adult children were important predictors of both positive and negative aspects of relationship quality. This pattern is similar to that found in studies of parents and offspring in childhood and adolescence (Pearce and Axinn 1998; Stokes and Regnerus 2008), demonstrating the important role played by perceptions of religious similarity across the life course. Although our findings revealed that religious similarity shapes parent-adult child relations for both Black and White mothers, it is noteworthy that this dimension of similarity is more important for Black than White mothers, as predicted. We believe that this finding reflects the greater role religion plays in Black families and the Black community more broadly (Barnes 2005; Patillo-McCoy 1998).

The present study extends our understanding of the role of religion in adult family relationships by focusing on the perceptions of similarity of religious values, rather than on similarities of denominational membership, religious behaviors or salience. Previous studies of religious similarity in parent-child relationships have measured reports of similarity of denomination, attendance behaviors and/or salience (Myers 2004; Pearce and Axinn 1998; Rossi and Rossi 1990; Stokes and Regnerus 2009). However, the broader literature on interpersonal relations points to the even greater importance of perceptions of similarity than similarity in role partners’ reports of their own values or behaviors. For example, empirical analyses of both marital relationships and intergenerational relationships highlight the importance of perceptions compared to actual accounts (Acock and Bengtson 1980; Frisco and Williams 2003; Lavee and Katz 2002; Suitor 1991; Wilcox and Nock 2006). These findings are consistent with Bengtson and Kuypers (1971) classic article concerning the different stake older parents and their adult children hold in a relationship, in which they proposed that parents’ and children’s perceptions of one another might be as important, if not more important, in predicting relationship quality than actual differences in attitudes and behaviors. We hope that future research will make direct comparisons regarding the differential effects of perceived religious value similarities and objective measures of religious value similarity on parent-adult child relationship quality.

These findings also raise several other important questions that we hope will be pursued in future research. One of these questions focuses on the mechanisms through which similarity of religion enhances relationship quality. We believe that this process takes several avenues. First, mothers and adult children who have similar beliefs may share experiences such as attending church together and have fewer conflicts over religion directly. Second, religious beliefs often guide attitudes and behaviors beyond those directly related to religion (Leege and Kellstedt 1993; Mencken, Bader, and Embry 2009; Sherkat and Ellison 1999), therefore shared religious beliefs would serve to enhance relationship quality by minimizing violations of both family and societal norms and creating fewer disruptions within mother-child relationships. Although this is likely the case in all families, the elevated importance of religion in Black communities (Barnes 2005; Patillo-McCoy 1998) may make this particularly important for Black mothers and their adult children. For instance, in Anderson’s (1999) Code of the Street, there are links to the importance of religion in avoiding the problems of the street; additionally, there are references to the importance of religious similarity in maintaining “decent family” relationships (pp. 60). Although, our study offers evidence that religious similarity has a strong relationship to relationship quality within mother-adult child relations, it is unable to identify the specific causal patterns and mechanisms of the relationship. Therefore, an important next step for future studies would be to examine the pathways through which religious similarity influences relationship quality.

We believe that the findings we have presented set the stage for addressing other questions regarding religious similarity and parent-adult child relations. For example, mothers and fathers differ in terms of both the level of closeness and conflict and the predictors of relationship quality with their adult children (Suitor et al. in press). Most relevant to the present issue, similarity appears to be a more important factor in explaining relationship quality for mothers than fathers (Suitor and Pillemer, in press); a finding that is consistent with other research on gender differences in the effects of homophily (Suitor and Pillemer, 2000, 2002). Further, studies have shown that religion is more salient in the lives of women than men (Argyle and Beit-Hellahmi 1975; Miller and Hoffman 1995; Collett and Lizardo 2009), particularly for Black women (Levin, Taylor, and Chatters 1994; Taylor 1988), suggesting that the effects of religious similarity on fathers’ relationships with their adult children may not mirror those found for mothers. However, some studies have found that fathers play a key role in the religious socialization of their children (Baker-Sperry 2001; Zhai et al. 2007), which may mean religious similarity is just as important for father-adult child relations. Therefore, further research is needed to address the role of religious similarity in fathers’ relationships with their children.

Another important question for future research is the examination of children’s perceptions of the role of religious similarity in their relationships with their parents. Studies have shown that parent’s and adult children’s perceptions of their relationships often differ (Giarrusso et al., 1995; Shapiro, 2004; Suitor, Sechrist, Steinhour, Pillemer 2006). Further, although older parents place great importance on the transmission of their values, adult children may not place as high of importance on this continuity as they go on to live their independent lives (Allport 1961; Bengtson and Kuypers 1971). Therefore, the relationship between religious similarity and relationship quality may not be as strong when examining it from the adult children’s perspective.

Additionally, research has indicated that there are considerable variations of values and beliefs both across and even within specific religious denominations (Warner 1988, 1989; Wuthnow 1988); therefore, identifying similarity of beliefs rather than simply denomination is important across all faith groups. However, it may be that affiliation with specific groups may make similarity of beliefs even more important. For instance, within Protestant denominations, strict churches have been shown to require higher commitment levels (Olson and Perl 2001) while also producing greater retention of members (Sherkat and Wilson 1995; Smith and Sikkink 2003). Thus, parents affiliated with strict denominations may place greater importance on religious value transmission and religious similarity with their adult children. Further, variations in strictness and fundamentalism in non-Judeo-Christian religions may also produce different effects of religious similarity on parent-child relations. The issue of religious denomination is an important factor to pursue in future research regarding the effects of religious similarity on interpersonal relations.

The race differences found here also suggest that it may be fruitful to explore the way in which religious similarity shapes family relations in other ethnic groups. For instance, similar to Black Americans, religion in Mexican families has traditionally been very important, yet with immigration to the U.S., acculturation has encouraged a pulling away from traditional religious values, particularly among younger Mexican Americans (Williams 1990; Sanchez 1995). It is unclear how religious similarity might influence relationship quality in Mexican American families. Future research should examine how the role of religious similarity may differ for Mexican American families as well as other immigrant groups in which religion is central to their traditional cultural values, but younger generations are breaking from these traditions (Alba, Raboteau, and DeWind 2009).

In sum, the analyses presented in this study lend further evidence to the strong influence of value similarity in intergenerational relations. We believe that these findings shed new light on the role of religious similarity in later-life families, which until now has received little attention. Further, these analyses highlight an issue not adequately addressed by research at any point of the life course—race differences in the effect of similarity. We hope that our findings will spur researchers to take religious similarity into consideration when studying family relations, and to expand the study of this issue to other ethnicities, religions and cultural groups.