Transmission of Religiosity From Parent to Child: Moderation by Perceived Parental Depression and Anxiety

Abstract

Although parents have a strong influence on their children’s religiosity, little research has been conducted on how different parental anxiety and depressive problems affect the transmission of religiosity. The current study asked emerging adults to report on the religiosity and internalizing behaviors of their mothers and fathers as well as their own religiosity. Structural equation modeling was used to measure whether perceived parental internalizing problems, parent gender, and participant gender moderated the relationship between parental and emerging adult religiosity. Results indicated that the interaction terms between perceived maternal religiosity and both depressive and anxiety problems were significant only for females. Participant gender did not suggest a three-way interaction among variables, though parent gender did as the interaction term between maternal religiosity and both anxiety and depressive problems were significantly stronger than paternal ones.

Keywords

emerging adulthood religiosity internalizing problems gender differences parenting

Although many factors influence thoughts and behaviors, religiosity is one of the few that permeates nearly all aspects of individuals’ lives. Indeed, nationwide surveys in the United States have shown that 89% of people believe in god and 13% believe in some type of higher power (Pew Research Center, 2014). Thus, religion is part of most people’s lives, and parental factors like parenting practices, parental religiosity, and parental internalizing problems all influence the religiosity of their children (Jacobs, Miller, Wickramaratne, Gameroff, & Weissman, 2012). Indeed, often parents attempt to instill their values in their children by modeling religious behavior (Flor & Knapp, 2001). Many things can hinder a child from acquiring the religiosity of his or her parents, such as parental incarceration, parenting behaviors, and internalizing problems (Jacobs et al., 2012; Stearns & McKinney, 2017a).

In fact, in a study of young mothers and their young children, maternal depression inhibited the importance that children placed on religion (Jacobs et al., 2012). Specifically, Jacobs et al. (2012) expanded on an earlier study (Gur, Miller, Warner, Wickramaratne, & Weissman, 2005) that found that the transmission of religiosity (as measured by importance of religion and religious attendance) from mother to child was reduced when mothers displayed depression. Jacobs and colleagues (2012) examined a younger set of mothers who were the original offspring of the initial study done in 2005. Through semistructured interviews, Jacobs et al. assessed mothers and their children for depression and religiosity through reported religious attendance and personal spirituality twice over a 10-year span. They found that religious importance but not attendance hindered the transmission of religiosity from mother to child.

However, Jacobs et al. (2012) did not examine the influence of parental anxiety, the influence of parental internalizing problems with regard to child gender, or examine maternal and paternal variables. For the purpose of this study, internalizing problems are conceptualized as negative behaviors that are focused inward, such as fearfulness, sadness, and withdrawal. Thus, the current study advances prior research by replicating and expanding on the study by Jacobs et al. and examining how perceived maternal and paternal anxiety and depressive problems moderate the transmission of religiosity in the parent–emerging adult–child relationship. The current study also aims to provide information about how internalizing problems inhibit the transmission of values from parent to child. Thus, researchers may gain a better understanding of parent–child interaction processes, which can in turn allow for more effective interventions (e.g., aiding in fostering better parent–child relationships and assisting a family whose parents are depressed).

Parental and Child Religiosity

Although childhood, adolescence, and adulthood have been well studied, emerging adulthood is a relatively new area that needs to be explored further. Emerging adulthood is the phase after adolescence but before adulthood, generally encompassing the ages between 18 and 25 years (Arnett, 2000). Specifically, Arnett (2000) suggests that emerging adults are characterized by a delay in three areas: extended period of education, delay in leaving home, and choosing to wait to get married. College students generally meet two of the three criteria (higher education and delay of marriage), sometimes remain living with their parents or close enough to visit often, and still rely on their parents (McKinney & Power, 2012). Moreover, this phase is a period during which individuals attempt to establish their own identities, specifically in the areas of work, romance, and their worldviews. Since religiosity is often associated with individuals’ worldviews and identities, it makes sense to study religiosity during emerging adulthood. In particular when examining parental influence on child religiosity, it is useful to study the emerging adult when children are first beginning to separate from their parents and have the freedom to choose their religion. Moreover, emerging adulthood is a time when children begin to reexamine parent–child relationships and religious issues (Barry, Nelson, Davarya, & Urry, 2010).

Many influences affect the religiosity of emerging adults, specifically the religiosity of their parents. Spilman, Neppl, Donnelan, Schofiend, and Conger (2012) found that parental religiosity during their children’s adolescence was positively correlated to the adolescents’ religiosity, which likewise predicted religiosity in emerging adulthood. Specifically, mothers’ religious affiliation, attendance, and ideology during their children’s childhood served as predictors for religious ideology during emerging adulthood (Pearce & Thornton, 2007). Additionally, perceived similarity to parents’ religious beliefs, faith support, and attachment to fathers predicted emerging adult religiosity (Leonard, Cook, Boyatzis, Kimball, & Flanagan, 2012). The current study seeks to examine other variables that may affect the transmission of religiosity from parent to child.

Parental Internalizing Problems, Parenting, and Religiosity

Studies have investigated the role of maternal psychopathology on parenting practices. Indeed, children are likely to observe their parents behaving in manners associated with depression and anxiety, such as anhedonia, disengagement, as well as controlling and nervous behaviors. Moreover, maternal depression and anxiety (internalizing problems) have been associated with negative parenting behaviors such as aggression toward the child, overprotectiveness, and coercive control (Neppl, Conger, Scaramella, & Ontai, 2009). When mothers display internalizing symptoms or diagnoses, they also tend to be more detached in their parenting, such as being less affectionate, less structured, and spending less time with their children (Bailey, Hill, Oesterle, & Hawkins, 2009). Lovejoy, Graczyk, O’Hare, and Neuman (2000) determined from their meta-analysis that maternal depression was more strongly associated with negative/hostile parenting behaviors than disengaged or positive parenting.

Given that parental internalizing problems are likely to be noticed by children and have a large influence on parenting practices, it follows that it also has a hand in how children adhere to the religiosity of their parents. Jacobs et al. (2012) found that in a sample of young mothers and children, maternal depression moderated the importance that children ascribed to religiosity, but not church attendance or denomination, such that maternal depression decreased the transmission of religious importance. Similarly, studies have shown that in a population of adults maternal depression can moderate the likelihood that adult offspring will acquire the religion of their parents in the domains of religious importance and church attendance (Gur et al., 2005). Thus, previous literature has established that maternal depression seems to hinder the transmission of religiosity; however, these studies did not examine effects between males and females, and paternal effects; they only focused on depression and used either children or adults. Thus, a need exists to fill the gap by examining emerging adults to better understand why a sample of children did not find that maternal depression hindered church attendance but a sample of adults did, as well as to examine paternal variables, child gender, and other psychopathologies (i.e., anxiety; Gur et al., 2005; Jacobs et al., 2012). Moreover, this attenuation of transmission of religiosity may be because depressed parents tend to be less involved with their children than nondepressed parents; indeed, when parents are depressed, they are more likely to use harsher discipline practices more frequently and have less positive interactions with their children (Lovejoy et al., 2000).

According to social learning theory (Bandura, 2001), parents with internalizing problems may not be ideal models and thus are less likely to transmit religiosity to their children. Indeed, parents with internalizing disorders are less religious and therefore have less religiosity to model to their children (Gur et al., 2005). Similarly, studies have found that the strongest predictor of adolescent religious views was parent modeling of religious behavior, showing that families indeed have a strong effect on religious development (Flor & Knapp, 2001). Many studies have attempted to understand the transmission of religiosity from parent to child, and some have suggested that perceived parent–child relationship quality may play a role in whether religious practices instilled in emerging adults during their childhood remain later in life (Barry, Padilla-Walker, & Nelson, 2012). Goeke-Morey, Papp, and Cummings (2013) also found that as maternal religiosity increased, the family functioned better and parent–child attachment improved; the association between parent–child attachment security and family stressors also was moderated by maternal religiosity such that increased religiosity was associated with better attachment despite family stressors like family conflict, child adjustment problems, and maternal distress.

Thus, children are likely to notice when their parents are acting in depressed (i.e., disengaged, anhedonic, and withdrawn) or anxious (over controlling or nervous) manners, which will negatively affect the likelihood that the children accept the parents’ message of religious activities and importance (if the parents are religious and want to pass a religious message to their children). Admittedly, while previous literature suggests that maternal depression hindered religious transmission from parent to child (Jacobs et al., 2012), parental anxiety may function differently. For example, it is possible that anxious parents may attempt to deal with their anxiety by appealing to a higher power through increased religiosity in a way similar to individuals with Obsessive Compulsive Disorder who deal with their obsessive and compulsive thoughts through repetitive prayer (Al-Solaim & Loewenthal, 2011); in turn, they may wish to pass their religious values on to their children with greater zeal than nonanxious parents. However, the majority of evidence suggests that parental psychological disorders such as depression and likely anxiety lessen the transmission of religiosity from parent to child.

Therefore, the current study suggests that transmission depends on other variables within the parent–child relationship, such as parental internalizing problems, and seeks to further examine in what contexts children acquire the religiosity of their parents. Similarly, based on Bandura’s (2001) theory of social modeling, it is likely that same-sex modeling (i.e., mother–daughter and father–son) will facilitate the transmission of religiosity. However, as stated below in the gender section, evidence exists for opposite-sex dyads being particularly powerful with regard to the transmission of parent–child values.

Gender

Another variable that determines if religiosity will be passed from parents to children is gender, both that of the parent and of the child. The argument has been made that mothers are the primary influence on their children; however, this may not be the case, especially if the mother has internalizing behaviors (Gere et al., 2013). Admittedly, many studies suggest that mothers are the primary conductor of religiosity (Boyatzis, 2006). In part, this relationship may be due to women having higher levels of spirituality and religiosity than men (Spilka, Hood, Hunsberger, & Gorsuch, 2003). Fathers also are important, however, as mothers are more likely to engage in conversations about religion when fathers are present (Boyatzis, 2006).

The gender of children also plays a role in their religiosity. Adolescent boys, for example, seem to be more heavily influenced by the religiosity of their parents (Flor & Knapp, 2001). Adolescent girls, on the other hand, have reported more intrinsic religiosity than boys (Henry, Plunkett, Robinson, Huey, & McMichael, 2009). Based on previous studies, opposite-gender dyads are particularly helpful for the facilitation of transmission from parent to child. For example, research has found increased transmission among father–daughter pairs (Halgunseth, Jensen, Sakuma, & McHale, 2016; Stearns & McKinney, 2017a, 2017b). Admittedly, as stated above, females are more religious in general, but fathers and daughters also may experience a better relationship during adolescence (Nielsen, 1996). Similarly, evidence exists that sons are more likely to acquire the religiosity of their mothers given that Dickie, Ajega, Kobylak, and Nixon (2006) found that sons who reported an increased closeness to their mothers similarly reported that they felt an increased closeness to god and greater religiosity; daughters, however, who were close to their mothers were only likely to feel an increased closeness to god but did not report more religiosity. These studies collectively emphasize the importance of examining both parent and child gender in the context of parent–child dyads. Few studies have examined gender differences (both maternal and paternal as well as that of the child), and the current study examines how gender dyads may play a role in the transmission of religiosity, particularly in the presence of parental internalizing problems.

Current Study

The current study advances previous literature by examining how the religiosity of emerging adults is influenced by the perceived internalizing problems (i.e., anxiety and depression) and religiosity of both their mothers and their fathers. Specifically, this study replicates and expands on Jacobs et al. (2012) by not just using the same analyses (moderation) but also by adding paternal variables and parental anxiety, using emerging adults, and further moderating the relationship by participant gender (i.e., female and male emerging adults). Moreover, whereas Jacobs et al. had examined only depressed mothers, it was suspected that parental anxiety would lead to the same outcomes as depression had in their sample, though perhaps through different parenting mechanisms (e.g., children of depressed parents experiencing rejection whereas children of anxious parents dealing with controlling and anxious parenting behaviors). Additionally, previous research has suggested that mother–son and father–daughter dyads result in particularly powerful religiosity transmission (Dickie et al., 2006; Stearns & McKinney, 2017a). The current study seeks to determine whether the facilitating power of these identified gender dyads still remain in the presence of parental internalizing problems.

Similar to many other studies, the current study used perceptions of parental religiosity and internalizing problems based on several studies that have indicated that children’s perceptions of their parents may be just as important as reality (Finley, Mira, & Schwartz, 2008; Yahav, 2006). Thus, this method may provide important and unique information about the process through which religiosity is transmitted from parent to child based on how the participants perceive their parents’ religiosity and internalizing problems. Specifically, asking emerging adults to report on their perceptions of their parents allows for a better understanding of how their parents have presented themselves to their children and how they have been perceived by their children. Naturally, this method has the weaknesses of shared-method bias as well as the difficulty inherent in asking emerging adults who are not spending all their time with their parents to report on parental internalizing problems that are by nature internal. Admittedly, parental internalizing problems may affect children even if they do not perceive it on a conscious enough level to report. However, previous studies have validated certain measures, such as the Adult Behavior Checklist (ABCL), for cross-informant ratings of psychopathology, suggesting that children may be able to report on their parents’ psychopathology (Achenbach & Rescorla, 2003).

The following hypotheses were made:

Hypothesis 1: Perceived parental religiosity would be positively associated with emerging adult religiosity.

Hypothesis 1a: Perceived maternal religiosity would be a stronger predictor of emerging adult religiosity than perceived paternal religiosity.

Hypothesis 1b: Perceived maternal and paternal religiosity would be a stronger predictor of female emerging adult religiosity than male emerging adult religiosity.

Hypothesis 2: Perceived parental internalizing problems (i.e., depressive and anxiety problems) would moderate the relationship between parental and emerging adult religiosity, in that increased parental internalizing problems would weaken the relationship between parent and child religiosity.

Hypothesis 2a: The interaction terms between maternal religiosity and both anxiety and depressive problems would be stronger than the paternal interaction terms.

Hypothesis 2b: the interaction terms would be stronger for emerging adult females than males.

That is, three-way interactions were hypothesized where the interaction term would be further moderated by parent and child gender.

Method

Participants

Participants were recruited from a general Psychology Research Pool (PRP) made up of Introduction to Psychology students who are given class credit in exchange for their participation in various studies via a brief description of the project stating that participants would fill out questionnaires about religiosity and their family. The demographics of the sample are representative of the Introduction to Psychology subject pool at large and corresponds to other PRP study demographics (i.e., majority female, between the ages of 18 and 25 years, and made up of undergraduates). Given that the title of the study was “Religiosity and Family Study,” some participants may have self-selected to take part in the study if they felt that religiosity was important to them. The PRP pool in that semester had approximately 965 students who could have participated in the study.

The sample (N = 435; 295 female, 140 male) consisted of emerging adults between the ages of 18 and 25 years (M = 20.62, SD = 1.76) who were attending a large Southern U.S. university. After removing 22 participants who were missing data from an entire measure for either maternal or paternal data, 413 remained (281 female, 132 male). Participants identified their race as Caucasian (65.5%), African American (25.9%), Latino (2.1%), Asian (4.2%), or Other (2.6%). A high percentage of participants reported being Christian-other (47.5%), whereas others were Baptist (16.3%), Catholic (8.9%), Protestant (3.1%), Atheist (4.0%), Other (3.3%), Methodist (3.3%), Neo-pagan (2.8%), and Spiritual (2.5%). Most participants reported that their parents had a bachelor’s degree (mother 32.1%, father 27.5%) or high school diploma (mother 25.2%, father 33.2%); other responses for maternal and paternal education included 18.2% and 16.8% who had a master’s degree, 16.0% and 13.1% who had an associate’s degree, 3.5% and 6.1% who had a doctorate, and 5.0% and 4.9% other, respectively. Moreover, 69.5% of participants reported that they grew up with both their biological mother and father, 19.5% with only their biological mother, 3.4% with only their biological father, and 7.7% reported some other living arrangement. Participants also reported that 50% spend at least 1 hour a day with their fathers, and 70% replied that they spend at least 1 hour with their mothers. Similarly, half of the participants reported that they currently live with their fathers and 75% that they currently live with their mothers.

Measures

Stearns–McKinney Assessment of Religious Traits

The Stearns–McKinney Assessment of Religious Traits (SMART) was developed as a new scale designed to measure various dimensions of religiosity (Stearns & McKinney, 2018). The overall scale includes 53 statements describing religious activities, feelings, and beliefs and is scored on a Likert-type scale from 0 = not true to 7 = very true. Factor analysis indicated a higher order Religiosity factor, which consists of 5 lower order factors: Private Religiosity (e.g., I try to live my life according to my religious beliefs), Social Support (e.g., I consider myself active in my faith or church), Coping (e.g., I find comfort in my religion or spirituality), Conviction (e.g., I will always believe in a divine being/God), and Conservative Religiosity (e.g., I strictly follow my religious beliefs with regard to my appearance). Items range from 9 to 13 for each of the four factors. Factor loadings of the five factors onto the overall Religiosity factor ranged from .75 to .99, and item loadings onto each of the five factors ranged from .65 to .84. Internal consistency for the five factors ranged from .88 to .95. Due to low factor loadings, the conservatism factor was dropped for this study. Validity has been demonstrated by comparing the SMART with several established scales including the Religious Well-Being Scale, the Santa Clara Strength of Religious Faith Scale, and the intrinsic subscale of the Religious Orientation Scale–Revised (Gorsuch & McPherson, 1989; Paloutzian & Ellison, 1982; Plante & Boccaccini, 1997). Strong correlations among the overall religiosity scale of the SMART and the other scales ranged from .70 to .77 and from .50 to .76 with a mean of .67 for the five factors, indicating good convergent validity (Stearns & McKinney, 2018).

Adult Behavior Checklist

The ABCL(Achenbach & Rescorla, 2003) consists of 123 statements used to assess the internalizing and externalizing behaviors of others over the past 6 months. The ABCL problem behaviors are scored with 0 = not true, 1 = somewhat or sometimes true, and 2 = very true or often true. Factors (or subscales) include depressive (e.g., Feels worthless or inferior) and anxiety problems (e.g., Nervous, highstrung, or tense). The current study used the 14-item Diagnostic and Statistical Manual of Mental Disorders (DSM) scale for Depressive problems and 7-item DSM scale for Anxiety problems (Achenbach & Rescorla, 2003). Internal consistency alphas have ranged from .87 to .93 in past studies (Achenbach & Rescorla, 2003). Cross-informant correlates ranged from .30 to .79, with a median of .42, indicating good psychometrics with regard to measuring the behaviors of others (Achenbach & Rescorla, 2003).

Control Variables

Based on previous literature, several control variables were added to the study: race, home status, and parental education to determine their effects on emerging adult religiosity. For example, research has indicated that African American adolescents were more likely to describe themselves as religious and attend church meetings when compared with Caucasian adolescents (Molock & Barksdale, 2013). Moreover, increased parental education levels were associated with decreased religious transmission from parent to child (Leon & van Liew, 2017). Finally, it is likely that if emerging adults lived with both their mothers and their fathers (instead of with only one parent) they were more likely to acquire the religiosity of their parents. For example, Boyatzis, Dollahite, & Marks (2006) found that mothers were more likely to discuss religious topics with children when fathers were present. Thus, these variables were placed into the model to rule out their effects on moderation of parental internalizing problems regarding the transmission of religiosity from parent to child.

Procedure

On institutional review board approval, the questionnaires were posted to an online survey system. Potential participants could read about the study and provided informed consent prior to their participation. Participants completed maternal and paternal versions of the ABCL as well as personal, maternal, and paternal versions of the SMART separately; completed these measures in random order; and were instructed to base their ratings on current perceptions. Thus, all the measures were completed from the participants’ current perspective. On voluntary withdrawal or completion of the study, participants were given a debriefing form and received credit for their participation.

Planned Analyses

Structural equation modeling was conducted using AMOS 24.0. Latent variables included perceived maternal religiosity, perceived paternal religiosity, and emerging adult religiosity. First a measurement model was constructed to ensure that the indicator variables (private religiosity, coping, social religiosity, conviction, and conservatism) were loaded onto the latent religiosity variables (mother, paternal, and emerging adult). Once the measurement model with latent variables was determined, a structural with latent and observed variables was analyzed. Observed variables included perceived maternal and paternal anxiety and depressive problems and were examined using two different models, one for maternal and paternal anxiety problems and one for maternal and paternal depressive problems. The maximum likelihood method of covariance structure analysis was used. Model fit was examined with the comparative fit index (CFI), the Tucker–Lewis index (TLI), the standardized root mean square residual (SRMR), and the root mean square error of approximation (RMSEA). According to Hu and Bentler (1999), CFI and TLI values >.90, SRMR values <.10, and RMSEA values <.08 indicate acceptable fit; similarly, CFI and TLI values >.95, SRMR values <.08, and RMSEA values <.06 indicate good model fit.

Hypothesis 1 stated that perceived parental religiosity would be positively associated with emerging adult religiosity and was tested by examining the path coefficients for the observed variables shown in Figures 2 and 3. Hypothesis 2 stated that perceived parental internalizing problems (i.e., depressive and anxiety problems) would moderate the relationship between parental and emerging adult religiosity and was tested by examining the path coefficients for the interaction effects shown in Figures 2 and 3 and plotted in Figures 4 to 7. Based on the latent religiosity variables (maternal, paternal, and emerging adult) created by AMOS, a composite “observed” variable was imputed into SPSS. The religiosity, anxiety, and depressive problems variables were centered and turned into interaction terms within SPSS. Interaction terms included perceived maternal religiosity × maternal anxiety and depressive problems for a total of two maternal interactions, as well as perceived paternal religiosity × paternal anxiety and depressive problems for a total of two paternal interactions, totaling four interaction terms altogether. These interaction terms then were examined in AMOS with path analysis. Significant interaction terms were interpreted by plotting them using simple slope analyses at ±1 SD.

Pairwise parameter comparisons, a statistical test comparing the difference between path coefficients or multiple group analysis, were used to test Hypotheses 1a, 1b, 2a, and 2b (see Figures 2 and 3 for conceptual models and Figures 4-7 for interactions). This comparison produces a z score indicating the statistical difference between two-path coefficients (i.e., between the male and the female paths; Byrne, 2013). Specifically, male and female as well as maternal and paternal path coefficients were compared to determine relationships moderated by gender. For example, this comparison allows for the path between maternal religiosity and emerging adult religiosity to be compared between males and females to indicate if one is sufficiently stronger than the other (i.e., Hypothesis 1b). Analyses were first conducted with the overall sample and then separately for males and females to determine any gender differences.

Results

Table 1 shows correlations, alphas, and descriptive statistics among latent variables based on gender. The original measurement model with latent perceived religiosity variables as described above and shown in Figure 1 provided good model fit (SRMR = .09, CFI = .98, TLI = .97, RMSEA = .05). All factor loadings except paternal conservatism (.32) exceeded .79 (all ps < .001), indicating convergent validity. Results were controlled for race, home status (who lived in the house with the child growing up, such as mother, father, or both), and parental education using dummy coding and by loading the control variables onto the dependent variable of emerging adult religiosity (see Table 2 and Figures 2 and 3). Results remained the same when controlling for the above variables.

Table 1.

Correlations, Mean, Standard Deviation, and Alphas Among Variables in Measurement Model Based on Gender.

MGA based on gender	1	2	3	4	5	6	7	α	M	SD
1. Maternal Religiosity	1	.50	.42	−.11	ns	−.30	ns	.99	72.14	19.50
2. Emerging Adult Religiosity	.65	1	.42	−.15	ns	−.25	ns	.98	70.00	21.68
3. Paternal Religiosity	.47	.42	1	ns	.27	−.14	.26	.89	38.11	6.37
4. Maternal Anxiety	−.16	−.26	−.22	1	.20	.71	.24	.89	10.86	3.34
5. Paternal Anxiety	ns	ns	ns	ns	1	ns	.73	.85	17.42	3.60
6. Maternal Depressive	−.38	−.30	−.32	.65	ns	1	.12	.95	18.27	5.31
7. Paternal Depressive	ns	ns	ns	ns	.78	ns	1	.78	35.62	6.03
α	.98	.97	.90	.87	.84	.90	.69	—	—	—
M	70.13	63.23	36.52	10.65	16.57	18.35	34.12	—	—	—
SD	17.94	22.11	7.25	3.13	3.75	5.52	6.45	—	—	—

Note. MGA = multiple group analysis. All ps < .05 unless noted as “ns.” Correlations appear below the diagonal and are shaded for males and above the diagonal for females.

Figure 1.

Measurement model factor loadings (all ps < .001).

Table 2.

Control Variable Betas.

Control variables	Anxiety model		Depression model
Control variables	Males	Females	Males	Females
1. Maternal Education	−.11	−.03	−.07	.08
2. Paternal Education	−.08	.09	−.08	−.02
3. Lived with Mom	.04	.22*	.01	.20*
4. Lived with Dad	.05	.04	.04	.04
5. Lived with Both	−.03	.20*	−.08	.21*
6. Caucasian	−.33*	−.08	−.36*	−.05
7. African American	−.29*	.03	−.33*	.06
8. Asian	−.18*	−.14*	−.18*	−.13*
9. Hispanic	.07	.02	.07	.01
10. Other	−.18*	−.09	−.20*	−.08

p < .05.

Figure 2.

Structural model indicating moderation of religiosity transmission by parental anxiety problems. See Table 3 for path coefficients.

Figure 3.

Structural model indicating moderation of religiosity transmission by parental depressive problems. See Table 3 for path coefficients.

The structural model for anxiety problems shown in Figure 2 provided acceptable to good model fit (SRMR = .09, CFI = .97, TLI = .96, RMSEA = .05), as did the structural model for depressive problems shown in Figure 3 (SRMR = .09, CFI = .96, TLI = .95, RMSEA = .07). See Table 3 for path analysis results. Confirming Hypothesis 1, results indicated that both perceived maternal and paternal religiosity shared a positive association with emerging adult religiosity in both the anxiety and the depressive problems models. Additionally, analyses confirmed that Hypothesis 1a in both models as perceived maternal religiosity were a significantly stronger predictor of emerging adult religiosity than perceived paternal religiosity (anxiety model: z = 1.84, one-tailed p = .03; depressive model: z = 1.82, one-tailed p = .03).

Table 3.

Results for Path Analyses Predicting Emerging Adult Religiosity.

Predictors	Anxiety model			Depression model
Predictors	Overall	Males	Females	Overall	Males	Females
1. Maternal Religiosity	.41	.58	.37	.35	.58	.35
2. Paternal Religiosity	.23	ns	.23	.23	ns	.23
3. Maternal Anxiety	−.13	−.15	−.11	—	—	—
4. Paternal Anxiety	Ns	ns	ns	—	—	—
5. Maternal Anxiety × Religiosity	Ns	ns	−.08	—	—	—
6. Paternal Anxiety × Religiosity	Ns	ns	ns	—	—	—
7. Maternal Depression	—	—	—	−.17	ns	−.17
8. Paternal Depression	—	—	—	ns	ns	ns
9. Maternal Depression × Religiosity	—	—	—	−.16	ns	−.16
10. Paternal Depression × Religiosity	—	—	—	ns	ns	ns

Note. All paths <.05 unless stated “ns.”

Hypothesis 1b was not confirmed in the maternal anxiety model as results were in the opposite direction; specifically, perceived maternal religiosity shared a positive association with emerging adult religiosity in both males and females, with the male path being stronger than the female path (z = 2.60, p < .001). Perceived paternal religiosity was associated with emerging adult religiosity in females only, supporting Hypothesis 1b, although the difference between males and females was not significant. Hypothesis 1b also was not confirmed in the depressive model as the positive association between perceived maternal religiosity and emerging adult religiosity was stronger in males than in females (z = 2.67, p < .001). Similar to the paternal anxiety model, perceived paternal religiosity was associated with emerging adult religiosity in females only, although the difference between males and females was not significant.

Hypothesis 2 was not confirmed in the anxiety model as neither the interaction term between perceived maternal religiosity and maternal anxiety problems nor the interaction term between perceived paternal religiosity and paternal anxiety problems were significant (Figures 4 and 6). Hypothesis 2 was partially confirmed in the depressive model as the interaction term between perceived maternal religiosity and maternal depressive problems was significant (Figures 5 and 7), such that increased maternal depression decreased the transmission of religiosity from mother to child. Results showed no moderation for the paternal variables. Hypothesis 2a was not confirmed in the anxiety model but was confirmed in the depressive model such that the interaction term between maternal religiosity and depressive problems was significantly stronger than the interaction term between paternal religiosity and depressive problems (z = 2.70, p < .001). These results suggest a three-way interaction such that the interaction terms between perceived parental religiosity and depressive problems were further moderated by parental gender with maternal depressive problems hindering the transmission of religiosity from mother to emerging adult child more than paternal depressive problems.

Figure 4.

Maternal Religiosity × Anxiety problems interaction with overall sample. Interaction was not significant.

Figure 5.

Maternal Religiosity × Depressive problems interaction with overall sample. Interaction was significant.

Figure 6.

Maternal Religiosity × Anxiety problems interaction. Interaction was significant for females but not for males.

Figure 7.

Maternal Religiosity × Depressive problems interaction. Interaction was significant for females but not for males.

Based on these results, when maternal internalizing problems were low, no gender differences occurred: Low maternal religiosity was related to lower male and female religiosity, and higher maternal religiosity was related to higher male and female religiosity. That is, maternal religiosity predicted male and female religiosity equally across gender when maternal internalizing problems were low. When maternal internalizing problems were higher, however, gender differences did occur (i.e., the interaction term between maternal religiosity and both anxiety and depressive problems was significant for females but not for males), which differed across the anxiety and depressive problems models. In the maternal anxiety model, males had especially low religiosity when their higher anxiety mothers also were lower in religiosity; when their mothers were higher in religiosity, males did not differ from those whose mothers were lower in anxiety (i.e., maternal anxiety had no effect on males when maternal religiosity was higher). For females, however, high maternal anxiety reduced their religiosity when maternal religiosity was higher (i.e., female religiosity was lower in the higher religiosity/higher anxiety group than the higher religiosity/lower anxiety group). Thus, maternal anxiety appeared to hinder transmission for females. The same pattern occurred in the depressive model, except that males in the lower maternal religiosity group were not different based on higher or lower maternal depression. Both male and female transmission was hindered when maternal depression was higher, and this effect was especially pronounced for females.

For Hypothesis 2b, pair wise parameter comparisons showed no statistical difference between males and females regarding the interaction term between perceived maternal religiosity and maternal anxiety problems; however, the interaction term between maternal religiosity and anxiety problems was significant for females when using a one-tailed test but was not significant for males (Figure 6). Results still showed no moderation for the paternal variables when female and male participants were examined separately. Similarly, the interaction between perceived maternal religiosity and maternal depressive problems was significant for females but not for males (Figure 7), although this difference was not statistically significant. Results still showed no moderation for the paternal variables when the sample was separated into males and females. Pair wise parameter comparisons did not indicate a three-way interaction for emerging adult gender in either the anxiety or the depressive problems models (i.e., no statistical differences were found between male and female interactions); however, several interactions were significant for females but not for males. See below for a discussion of these interactions.

Discussion

The current study examined the role of perceived parental internalizing problems in the transmission of religiosity from parents to children in emerging adult males and females. Specifically, the current study contributes to previous literature by examining both maternal and paternal variables, parental anxiety, and used participant gender as a moderating variable. Both perceived maternal and paternal religiosity predicted emerging adult religiosity likely due to the strong influence of parents on children, as found in other research (e.g., Pearce & Thornton, 2007; Stearns & McKinney, 2017a). Indeed, parents socialize the religiosity of their children by exposing them to religious beliefs and activities as well as by encouraging them to engage in religious communities (Boyatzis, 2006). However, mothers tended to be the primary influence on religiosity, likely due to being viewed as the primary caregiver and spending more time with children (Boyatzis, 2006).

The current study found that the father–daughter bond may facilitate the transmission of religiosity (Halgunseth et al., 2016; Stearns & McKinney, 2017a, 2017b). This association may be related to females being more religious in general, thus boosting the transmission of religiosity within the father–daughter dyad (Boyatzis, 2006). Moreover, fathers and daughters may have a special bond due to decreased conflict during the teenage years, in comparison to mothers and daughters, and a less competitive relationship than fathers and sons experience (Nielsen, 1996). Some researchers even have suggested that fathers have a strong influence on daughters’ ability to trust, which is likely to have an influence on their belief in a paternalistic deity (Flouri, 2005).

Additionally, the mother–son relationship influences the transmission of religiosity as prior research has suggested that sons, but not daughters, who reported an increased closeness to their mothers similarly reported that they felt an increased closeness to god and greater religiosity (Dickie et al., 2006). Some research also has shown that sons, specifically adolescent boys, are more likely to be influenced by the religiosity of their parents (Flor & Knapp, 2001). This effect is perhaps due to boys being more receptive to the teachings of their mothers and that mothers are more likely than fathers to communicate the importance of religiosity (Boyatzis, 2006).

Aligning with previous studies, the current sample showed that maternal (but not paternal) depression lessens the importance children place on religiosity and hinders the transmission of religiosity from parent to child. The current study also indicated that maternal anxiety, a previously unstudied component, hindered the transmission in the same way as depression, showing that maternal internalizing problems can be detrimental to the child’s acquisition of religiosity. These results are supported by previous research finding that maternal depression hindered the transmission of religiosity from mother to child (Gur et al., 2005; Jacobs et al., 2012). Indeed, when parents, particularly mothers, have internalizing symptoms, they are more likely to use negative parenting behaviors and spend less time with their children (Bailey et al., 2009; Neppl et al., 2009). Moreover, this effect is likely a result of detached parenting and decreased parent–child communication (Bailey et al., 2009; Boyatzis, 2006; Gur et al., 2005; Jacobs et al., 2012). Similarly, studies have shown that modeling is an important component to the transmission of religiosity, and parents with internalizing problems are less likely to model religiosity in a manner that effectively encourages their children to be religious (Flor & Knapp, 2001).

The current study contributes to the understanding of child religiosity by helping to explain gender differences among emerging adult children. Specifically, maternal internalizing problems served as a hindering factor for female children. Female children may be more sensitive to parental psychopathology than male children given their own increased likelihood of developing internalizing disorders (Muller et al., 2016). Otherwise, female children may be more influenced by the religiosity of their parents, and when the parent–child relationship is negative, they are less likely to be influenced by their parents than male children (Stearns & McKinney, 2017a). However, few studies have examined how gender dyads effect the transmission of religiosity, and while the current study provides new information to conceptualizing parent–child transmission, more work is left to be done to clarify the part that gender plays.

Implications for Research and Practice

Given the results of the current study, gender, perceived parental religiosity, and perceived parental internalizing problems clearly have a strong influence on the religiosity of emerging adults. Researchers still have much ground to explore in an effort to specify what mechanisms are causing differential results regarding gender. Specifically, it is important to examine why perceived parental internalizing problems hindered the transmission of religiosity for some dyads but not for others. Additionally, researchers should examine why the father–daughter and mother–son dyads are so strong regarding transmission of religiosity.

The implications stemming from this and previous studies involve the ramifications of perceived parental religiosity and parental internalizing problems on personal religiosity. As researchers continue to explore how parental internalizing problems affect their children, it is important to see how these effects extend into emerging adulthood. Increased religiosity has been associated positively with many mental health outcomes, particularly if the parents themselves have poor mental health (Brewer-Smyth & Koenig, 2014). Importantly, the current study found that parents with poor mental health might have trouble fostering religiosity in their children, possibly suggesting the need for intervention.

Limitations

This study must be viewed in the context of its limitations, such as using a college sample. While the current study aims to be generalizable to emerging adults as a whole, the sample was conducted on college students, and they do not encompass all emerging adults. However, as argued by Arnett (2000), college is typically representative of emerging adulthood. Moreover, the current sample was conducted in the Southern United States and some variables, such as religiosity, may differ across regions. Additionally, the sample was conducted on Introduction to Psychology students and the results may not be generalizable to Southern college students at large. Thus, the results of the current study must be generalized to college students at a Southern university with caution.

Additionally, the study asked participants to indicate the religiosity and psychological problems of their parents through their perspective, and a shared-method bias may exist due to relying on a single informant. Admittedly, it may be difficult for children who live separately from their parents to report on current parental anxiety and depressive problems; however, in the current sample, 50% of the participants responded that they spend at least 1 hour a day with their fathers, and 70% replied that they spend at least 1 hour with their mothers. Similarly, half of the participants reported that they currently live with their fathers and 75% that they currently live with their mothers. Of course, it is impossible to know whether or not participants who did not live with their mother or father still reported on their parent and thus was unable to provide accurate information. However, controlling for home status (whether the participants lived with their mother, father, or both) helps address this limitation. Furthermore, participants, for example, with mental health problems of their own may view their parents as having more problems merely as a result of their own mental health problems. Studies have shown, on the other hand, that children’s perceptions of their parents may be just as important as reality (Finley et al., 2008; Yahav, 2006). Thus, how the participants perceive their parents’ religiosity and psychopathology may provide unique information relative to parents’ reports. Finally, due to the cross-sectional nature of the study, causality and direction of effects cannot be determined; theory and prior research, however, support the directions examined here (e.g., Okagaki & Bevis, 1999).

Summary

The current study demonstrated the influence of perceived parental religiosity and parental internalizing psychopathology on emerging adults’ religiosity. Given that many previous studies have indicated that religiosity can serve as a protective element and personal religiosity is highly correlated with parental religiosity, it is important to identify factors that either facilitate or hinder the transmission of religiosity from parent to child. Although beyond the scope of this article, more research needs to be done regarding parental psychopathology and its effects on parent–child interactions and the parent–child relationship. Moreover, the current study highlighted that gender differences play an important role in whether or not perceived parental religiosity and parental psychopathology will have an effect on the religiosity of children. A meta-analysis examining gender findings involving religiosity and the transmission of religiosity from parent to child would be helpful in determining differential gender results. Similarly, researchers should examine the uniqueness of specific dyads (e.g., father–daughter, mother–son) and what makes them important regarding transmission of religiosity.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Melanie Stearns

References

Achenbach

Rescorla

(2003). Manual for the ASEBA adult forms and profiles. Burlington: University of Vermont, Research Center for Children, Youth, and Families.

Al-Solaim

Loewenthal

K. M.

(2011). Religion and obsessive-compulsive disorder (OCD) among young Muslim women in Saudi Arabia. Mental Health, Religion & Culture, 14, 169-182.

Arnett

(2000). Emerging adulthood: A theory of development from the late teens through the twenties. American Psychologist, 55, 469-480.

Bailey

Hill

Oesterle

Hawkins

(2009). Parenting practices and problem behavior across three generations: Monitoring, harsh discipline, and drug use in the intergenerational transmission of externalizing behavior. Developmental Psychology, 45, 1241-11226.

Bandura

(2001). Social cognitive theory: An agentic perspective. Annual Review of Psychology, 52, 1-26.

Barry

Nelson

Davarya

Urry

(2010). Religiosity and spirituality during the transition to adulthood. International Journal of Behavioral Development, 34, 311-324.

Barry

Padilla-Walker

Nelson

(2012). The role of mothers and media on emerging adults’ religious faith and practices by way of internalization of prosocial values. Journal of Adult Development, 19, 66-78.

Boyatzis

(2006). Advancing our understanding of religious dynamics in the family and parent-child relationship. International Journal for the Psychology of Religion, 16, 245-251.

Boyatzis

Dollahite

Marks

(2006). The family as a context for religious and spiritual development in children and youth. In Roehlkepartain

E. C.

Benson

P. L.

Wagener

(Eds.), The handbook of spiritual development in childhood and adolescence (pp. 297-309). Thousand Oaks, CA: Sage.

10.

Brewer-Smyth

Koenig

(2014). Could spirituality and religion promote stress resilience in survivors of childhood trauma? Issues in Mental Health Nursing, 35, 251-256.

11.

Byrne

(2013). Structural equation modeling with AMOS: Concepts, applications, and programming (2nd ed.). New York, NY: Routledge.

12.

Dickie

Ajega

Kobylak

Nixon

(2006). Mother, father, and self: Sources of young adults’ God concepts. Journal for the Scientific Study of Religion, 45, 57-71.

13.

Finley

Mira

Schwartz

(2008). Perceived paternal and maternal involvement: Factor structure, mean differences, and parental roles. Fathering, 6, 62-82.

14.

Flor

Knapp

(2001). Transmission and transaction: Predicting adolescents’ internalization of parental religious values. Journal of Family Psychology, 15, 627-645.

15.

Flouri

(2005). Fathering and child outcomes. West Sussex, England: Wiley.

16.

Gere

Hagen

Villabo

Arnberg

Neumer

Torgersen

(2013). Fathers’ mental health as a protective factor in the relationship between maternal and child depressive symptoms. Depression and Anxiety, 20, 31-38.

17.

Goeke-Morey

Papp

Cummings

(2013). Changes in marital conflict and youths’ responses across childhood and adolescence: A test of sensitization. Development and Psychopathology, 25, 241-251.

18.

Gorsuch

R. L.

McPherson

S. E.

(1989). Intrinsic/extrinsic measurement: I/E revised and single-items scales. Journal for the Scientific Study of Religion, 28, 348-354.

19.

Gur

Miller

Warner

Wickramaratne

Weissman

(2005). Maternal depression and the intergenerational transmission of religion. Journal of Nervous and Mental Disease, 193, 338-345.

20.

Halgunseth

Jensen

Sakuma

McHale

(2016). The role of mothers’ and fathers’ religiosity in African American adolescents’ religious beliefs and practices. Cultural Diversity and Ethnic Minority Psychology, 22, 386-394.

21.

Henry

Plunkett

Robinson

Huey

McMichael

(2009). Adolescent religious motivation, family functioning, and empathic concern. Journal of Religion and Spirituality in Social Work: Social Thought, 28, 343-361.

22.

Bentler

P. M.

(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1-55.

23.

Jacobs

Miller

Wickramaratne

Gameroff

Weissman

(2012). Family religion and psychopathology in children of depressed mothers: Ten-year follow up. Journal of Affective Disorders, 136, 320-327.

24.

Leon

van Liew

(2017). Intergenerational religious transmission and conversion in the United States: Are the effects of parental religion moderated by parental religion concordance, education, geography, and generation? Research in the Social Scientific Study of Religion, 28, 1-21.

25.

Leonard

Cook

Boyatzis

Kimball

Flanagan

(2012). Parent-child dynamics and emerging adult religiosity: Attachment, parental beliefs, and faith support. Psychology of Religion and Spirituality, 5, 5-14.

26.

Lovejoy

Graczyk

O’Hare

Neuman

(2000). Maternal depression and parenting behavior: A meta-analytic review. Clinical Psychological Review, 20, 561-592.

27.

McKinney

Power

(2012). Childhood playtime, parenting, and psychopathology in emerging adults: Implications for research and play therapists. International Journal of Play Therapy, 21, 215-231.

28.

Molock

S. D.

Barksdale

C. L.

(2013). Relationship between religiosity and conduct problems among African American and Caucasian adolescents. Journal of Child and Family Studies, 22, 4-14.

29.

Muller

Tronick

Zietlow

Nonnenmacher

Verschoor

Trauble

(2016). Effects of maternal anxiety disorders on infant self-comforting behaviors: The role of maternal bonding, infant gender and age. Psychopathology, 49, 295-304.

30.

Neppl

Conger

Scaramella

Ontai

(2009). Intergenerational continuity in parenting behavior: Mediating pathways and child effects. Developmental Psychology, 45, 1241-1256.

31.

Nielsen

(1996). Adolescence: A contemporary view. Ft. Worth, TX: Harcourt Brace.

32.

Okagaki

Bevis

(1999). Transmission of religious values: Relations between parents, and daughters’ beliefs. Journal of Genetic Psychology, 160, 303-318.

33.

Paloutzian

R. F.

Ellison

C. W.

(1982). Loneliness, spiritual well-being and the quality of life. In Peplau

L. A.

Perlman

(Eds.), Loneliness: A sourcebook of current theory, research, and therapy (pp. 224-237). New York, NY: Wiley Interscience.

34.

Pearce

Thornton

(2007). Religious identity and family ideologies in the transition to adulthood. Journal of Marriage and Family, 69, 1227-1243.

35.

Pew Research Center. (2014). Pew forum on religion and public life: Religious landscape study. Washington, DC: Author.

36.

Plante

Boccaccini

(1997). The Santa Clara strength of religious faith questionnaire. Pastoral Psychology, 45, 375-387.

37.

Spilka

Hood

Hunsberger

Gorsuch

(2003). The psychology of religion: An empirical approach (3rd ed.). New York, NY: Guildford Press.

38.

Spilman

Neppl

Donnelan

Schofiend

Conger

(2012). Incorporating religiosity into a developmental model of positive family functioning across generations. Developmental Psychology, 4, 762-774.

39.

Stearns

McKinney

(2017a). Parental religiosity and emerging adult psychological adjustment: Moderated mediation by gender and personal religiosity. Psychology of Religion and Spirituality, 9(Suppl. 1), S60-S69.

40.

Stearns

McKinney

(2017b). Perceived parent-child religiosity: Moderation by perceived maternal and paternal warmth and autonomy granting and emerging adult gender. Psychology of Religion and Spirituality. Advance online publication. doi:10.1037/rel0000142

41.

Stearns

McKinney

(2018). Stearns-McKinney Assessment of Religious Traits (SMART): Development of an improved measure of religiosity. Manuscript submitted for publication.

42.

Yahav

(2006). The relationship between children’s and adolescents’ perceptions of parenting style and internal and external symptoms. Child: Care, Health, and Development, 33, 460-471.