Who's your daddy? Family structure differences in attachment to God.

Murunga, Maurice S. ; Limke-McLean, Alicia ; Wright, Ronald W. 等

Recent research has demonstrated that individuals' relationships with God are attachment-based. However, research has not yet investigated differences in attachment to God by parents' marital status. Thus, the goal of the present study was to examine these links. To do so, 288 undergraduate students completed measures assessing family structure, attachment to fathers, attachment to mothers, and attachment to God. Results suggest support for the correspondence theory of attachment to God (i.e., individuals project their attachment to parents onto their attachments to God) for participants with married parents. In contrast, the compensation hypothesis (i.e., individuals seek relationships with God to fulfill unreliable relationships with parents) was supported for participants with divorced parents.

**********

Previous research on attachment has spanned the human developmental lifespan. Research on early childhood has focused on the genesis of attachment and attachment patterns (e.g., Ainsworth, Blehar, Waters, & Wall, 1978; Bowlby, 1969, 1973, 1980). Studies on young adulthood have focused on the manifestations of attachment styles in young adults (e.g., Bartholomew & Horowitz, 1991). Among adults, research has mainly focused on the concept of romantic relationships as an attachment process (e.g., Collins & Read, 1990; Freeney & Noller, 1990; Hazan & Shaver, 1987; van Ijzendoorn, 1995). More recently, research has also demonstrated individuals' relationship with God as attachment-based, conceptualizing God as a secure base (Beck, 2006c; Beck & McDonald, 2004; Cassiba, Granqvist, Costantini, & Gatto, 2008). Similarly, matters of faith--such as how individuals' faith development occurred relative to the types of attachment bonds they manifested with others (Hart, Limke, & Budd, 2010) and the link between psychological health and bonding with God (Miner, 2009)--have been investigated.

However, the focus on bonding with God has not addressed whether parental marital status predicts variability in attachment to God. That is, although attachment to God seems to correspond to attachment to fathers (but not mothers) in Christian samples (cf. Limke, Amin, Poojar, & Kamble, 2015; Limke & Mayfield, 2011), it is possible that access to father figures (through divorce or other circumstances) may moderate this relationship. In the current study, we expected that individuals from divorced homes (in which they lived primarily with their mothers) would report higher attachment anxiety towards God than those with present fathers.

Background

Attachment

The seminal work in attachment theory originated from observations of orphans during the Second World War (Bowlby, 1973, 1988). Bowlby observed that infants and children who had been deprived of their mothers' attention and care for a prolonged time eventually developed "detachment." Demonstrating what he termed the protest/despair/detachment sequence, Bowlby (1973) suggested that following such a separation, the infants manifested protestations (e.g., cried inconsolably, stomped their feet, and longed for a reunion with their mothers); then, they gave in to despondency (e.g., stopped crying and expecting the return of their mothers); and finally, they embraced detachment (e.g., exhibited disorganized behaviors due to failure to find suitable substitute mothers). Bowlby observed that when the separation from mothers had been prolonged, whether reunion between the mothers and infants occurred or not, infants' detachment from their mothers persisted (Bowlby, 1980). Based on these observations, Bowlby proposed that mother-infant attachment was preprogrammed and innate, ethological/evolutionary in nature, and served an adaptive function; that is, attachment was important for the protection and survival of the human infant and, by extension, was central to the survival and psychological wellbeing of the human species.

Extending Bowlby's work, Ainsworth et al. (1978) delineated a three-pattern classification of infant-to-mother attachment styles: (a) secure infants (became agitated at their mothers' departure from the Strange-Situation room, were ecstatic on their mothers' return, and then settled down to explore the room); (b) insecure-anxious resistant/ambivalent infants (became distressed at their mothers' departure, remained agitated upon their mothers' return, and clung onto their mothers rather than venture into exploration of their environments); and (c) insecure-avoidant (were apathetic to both mother's departure and return). Main and Solomon (1990) demonstrated yet another dimension to the mother-infant interaction by adding a fourth category: disorganized/disoriented infants (were inconsistent in dealing with separations and reunions with mothers). When they re-analyzed Ainsworth et al.'s (1978) Strange-Situation results, Main and Solomon (1990) observed an additional class of infants who consistently manifested an undifferentiated pattern to their milieu; these disoriented infants showed inconsistency both in their proximity seeking and avoidance towards mothers.

This previous research on attachment has highlighted the role of the mother figure in parental attachment with children but has not sufficiently addressed father-child attachment (Ainsworth et al., 1978; Bowlby, 1988). More recently, research has found that fathers who perceived their parental responsibility as an important function tended to be more likely to raise securely-attached children than those who did not (Wong, Mangelsdorf, Brown, Neff, & Schoppe-Sullivan, 2009). Similarly, involved fathers buffered children from the effects of bullying. Furthermore, the fathers' involvement enhanced their ability to build social capital with their children, which correlated positively with academic achievement (Cooksey & Fondell, 1996; Flouri & Buchanan, 2002). Similarly, Goodsell and Meldrum (2010) have demonstrated that fathers can be both a playmate and a nurturer. They found that fathers could fill a nurturing void in the event of a mother's lack of ability or willingness to nurture. That is, fathers are important both as nurtures and as financial providers.

Since the original research in the area of attachment, researchers have (a) applied attachment process and theory to romantic relationships (Freeney & Noller, 1990; Hazan & Shaver, 1987), (b) utilized Ainsworth et al.'s (1978) classifications to develop measures, inventories, interviews, and models to assess adult bonding styles (Bartholomew & Horowitz, 1991; Collins & Read, 1990; van Ijzendoorn, 1995), and (c) explicated the link between individuals and God in terms of attachment bonds (e.g., Beck, 2006a, 2006c; Beck & McDonald, 2004; Cicirelli, 2004; McDonald, Beck, Allison, & Norsworthy, 2005; Miner 2009).

Attachment to God

Beck and McDonald's (2004) formulation of the Attachment to God Inventory (AGI) provided a means for direct assessment of individuals' attachment bonding with God. Beck and McDonald based their formulation on the Experiences in Close Relationships scale's (ECR; Brennan, Clark, & Shaver, 1998) two subscales of avoidance and anxiety. Beck and McDonald's use of the ECR as a basis for the development of the AGI suggested a parallel between individuals' attachment with others and individuals' attachment to God. Indeed, research has indicated that the two ECR subscales (avoidance of intimacy and anxiety about abandonment) underpin individuals' affiliative bonds both to others and to God (Beck, 2006a, 2006c; Beck & McDonald, 2004). However, despite the suggested parallel between interpersonal affiliations and individuals' affiliation with God, other researchers have made distinctions between person-to-person attachment bonding and person-to-God attachments; that is, God, unlike humans, has been considered as infallible and not subject to human foibles (Kirkpatrick & Shaver, 1990). Accordingly, person-to-God relationships are not thought to be influenced by the same kinds of constraints that inform interpersonal relationships (e.g., personal agendas that ultimately shape interpersonal affiliations).

Recent research has also found a link between matters of faith development and attachment; specifically, this research has examined how individuals' faith develops relative to the types of romantic attachment bonds they exhibit with others (Hart et al., 2010). This research found that anxiety over abandonment (but not avoidance of intimacy) predicted the development of individuals' faith. These results were congruent with previous research on bonding with God, suggesting that a perception of God as a secure and dependable figure (i.e., indicated by individuals' low levels of anxiety) is imperative for fostering bonds with God (i.e., the development of faith).

Similarly, God has been conceptualized as a transcendental attachment personality (e.g., Beck, 2006c; Beck & McDonald, 2004; Cicirelli, 2004; McDonald et ah, 2005; Miner, 2009). Researchers using religious samples have shown that the monotheistic Christian God is a secure, safe, and dependable foundation from which individuals can make forays into theological explorations (cf. Beck, 2006c). Further, Beck (2006c) found that individuals who perceived God as a secure and dependable base were more likely to be tolerant of Christian religious views that were different from their own. Additionally, such individuals were found to manifest less distress and were more likely to fully embrace the core Christian teachings about God. In contrast, the individuals who manifested attachment avoidance tended to be uninterested in theological explorations; furthermore, their religious views seemed inflexible and were non-inclusive of divergent religious views.

Expanding research on attachment to God, Miner (2009) found that the extent to which individuals were in secure attachments with God influenced their psychological adjustment beyond the emotional bonds they had in their attachment to primary caregivers. Individuals who exhibited secure bonds with God and with parents manifested an existential well-being and were also less anxious than individuals who had a negative perspective of relationships with parents. These findings intimated that secure attachment to God was a buffer to stress.

In this vein, research on attachment to God has proposed two salient hypotheses; compensation and correspondence (Beck, 2006c; Beck & McDonald, 2004; Kirkpatrick & Shaver, 1990). The compensatory hypothesis has proposed that persons who experienced untoward relationships with primary caregivers and lovers sought reparative attachment bonds with God. According to Kirkpatrick and Shaver (1990), such individuals see God as an alternative, personal, and relational attachment figure. These individuals were more likely to report sudden conversion experiences if they were from nonreligious backgrounds than if they were from religious ones originally, especially following emotionally tumultuous events (e.g., after a divorce, demise of a loved one, a miscarriage, or if they were in midst of marital discord).

Conversely, McDonald et al. (2005) found supportive evidence for the correspondence hypothesis. According to this hypothesis, individuals' attachment to God consistently mirrored the attachment bonds they manifested with lovers and caregivers. Although individuals who experienced unfulfilling interpersonal affiliations with lovers and parents might have sought a compensatory bond with God, they eventually settled down and reverted to the roles they played in their interpersonal affiliations. That is, individuals tended to transfer the attachment styles they manifested in personal affiliations to their relationship with God. Thus individuals' attachment bonds with God corresponded their attachment bonds with others.

Among older adults, the perception of God as a substitutionary attachment personality has been prevalent (Cicirelli, 2004). Many older adults tended to have no other attachment figures available to them. Many seemed to associate their substitutionary attachment to God to their fear of death; that is, the stronger their fear of death, the stronger their attachment to God was. However, as older adults advanced in years, their fear of death ironically decreased. As a result, Cicirelli suggested that although many older adults considered themselves religious, they did not have a strong affiliation to God. Therefore, it seemed that older adults sought a relationship with God for substitutionary reasons because they had no other attachment personalities available to them.

Other research has surveyed the role of religious orientation in attachment to God (Beck 2004,2006b). For example, Beck (2004) has differentiated between two religious orientations: existential believers and defensive believers. Defensive believers use their attachment to God as a buffer against existential threats (e.g., death), whereas existential believers embrace the ambiguity of life even when confronted with existential threats. Beck (2006b) later found that existential believers manifested less in-group bias than defensive believers following a Terror Management Theory mortality salience experimental procedure (e.g., Rosenblatt, Greenberg, Solomon, Pyszczynski, & Lyon, 1989; Greenberg, Pyszczynski, Solomon, Rosenblatt, Veeder, Kirkland, & Lyon, 1990). Furthermore, defensive believers scored higher on the Defensive Theological Scale (an assessment designed to measure individuals' religious defensiveness) than existential believers, suggesting that defensive believers had a sense of specialness in relation to God (e.g., "I believe God protects me from illness and misfortune," and "When making a choice or tough decision, God gives me clear answers and directions;" Beck, 2006b).

To summarize, recent research on affiliation with God has conceptualized God as a secure base and substitutionary attachment personality and has perceived the relationship with God as mirroring individuals' attachment bonds with parental and other primary caregivers. However, research on attachment to God has not focused on whether differences exist in patterns of attachment to God by family structure or whether attachment to God could buffer the effects of a religiously-inspired threat.

Parental Marital Status

The rise of divorce rates since the 1970s, and the proliferation of step-families, have diversified the structure of the American family (DeMuth & Brown, 2004; Griffin, Botvin, Scheier, Diaz, & Miller, 2000; Moore, Jekielek, & Emig, 2002; Sampson & Laub, 1994). Although some research has suggested that marriage dissolution is associated with poor academic performance and emotional health problems among teenagers (Moore et ah, 2002), other research has argued that the changes in family time, attachment, and family formation (remarriage following divorce), rather than the process of marital dissolution itself, predicts juvenile delinquency (Schroeder, Osgood, & Oghia, 2010).

Family attachment seems to matter more than family structure in predicting adolescent alcohol and marijuana use (Barfield-Cottledge, 2015). Among post-divorce couples, co-parental relationships are linked to attachment styles (Roberson, Sabo, & Wickel, 2011). That is, divorced parents with a stable co-parenting relationship tend to have children with securely attached styles, whereas divorced parents with conflicted co-parenting relationships have children with anxious ambivalent styles, and divorced parents with disengaged co-parenting relationships have children with avoidant attachment styles. Co-parenting arrangements are linked to continued presence of fathers, which has implications for children as well. Father absence has been linked to substance abuse and delinquency, early sexual activity among teenagers, mating rather than parenting, and low parental supervision (DeMuth & Brown, 2004; Griffin et al., 2000).

To summarize, parental divorce impacts children, although attachment to caregivers, co-parenting relationships, and continued presence of fathers moderate these detrimental effects.

Current Research

Recent research has shown individuals' affiliation with God as attachment-based; furthermore, this research has conceptualized God as a secure base (Beck, 2006c; Beck & McDonald, 2004; Cassiba et al., 2008). Moreover, in Christian samples, attachment to father, but not mother, predicts attachment to God, supporting the correspondence hypothesis of attachment to God (i.e., that relationships with fathers are projected onto relationships with God). However, to date, research has not examined differences in the link between attachment to parents and attachment to God by parents' marital status. It is possible that when fathers are absent, the ability or necessity to project images of or relationships with fathers onto relationships with God diminish, increasing the complexity of the link between attachment to parents and attachment to God. Therefore, the current study examined differences in the link between attachment to mothers, fathers, and God by parents' marital status.

Method

Participants

We recruited 297 undergraduate students from the University of Central Oklahoma (UCO) and 24 students from undergraduate programs at Southern Nazarene University (SNU) to participate in a study entitled "Attachment to God--Attitudes and Family Relationships." Whereas the 297 participants from UCO participated for partial fulfillment of a course research requirement and were recruited through a research registration system, the 24 participants from SNU were approached via email to participate. Both UCO and SNU participants were provided with a link to the survey website.

Of the 321 participants who responded, nine from UCO and five from SNU were excluded from the analyses because of failure to complete the survey. We ran a multivariate analysis of variance (MANOVA) to compare the remaining 288 UCO participants' and 19 SNU participants' samples. Whereas the UCO sample scored higher on Defensive Theology (M = 93.52, SD = 30.60) than the SNU sample [M = 78.32, SD =23.78), the SNU sample had higher scores than the UCO sample on God Anxiety ([M.sub.snu] = 53.89, SD - 15.51 vs. [M.sub.uco] = 40.35, SD = 15.93), Father Avoidance ([M.sub.snu] = 48.84, SD = 16.36 vs. [M.sub.uco] = 40.47, SD = 14.71), and Father Anxiety ([M.sub.snu] = 46.16, SD = 19.24 vs. AC, = 35.28, SD = 14.85), Fs(1, 305) > 4.50,ps < .04. Due to these differences and the fact that the SNU sample was too small to be broken down further, the entire SNU sample was excluded from subsequent analyses.

The remaining 288 participants were between the ages of 18 and 55 (M = 21.37, SD = 5.23). Of the 288 participants, 34.7% were male, 64.9% were female, and 0.3% self-reported as "other/neither." Further, 68.4% were White (non-Hispanic), 12.2% were Black or African American, 3.8% were American Indian or Alaska Native, 0.7% were Native Hawaiian or Pacific Islander, 5.2% were Asian, 7.6% were Hispanic or Latino/a, 1.0% were Middle Eastern, and 0.7% self-reported as "Other." Table 1 provides additional demographic information about the participants.

Materials

Attachment to parents. We used both the mother and father versions of the Experiences in Parental Relationship Scales (EPR; Limke & Mayfield, 2011) to assess participants' attachment to parents. The EPR Scales measure the attachment avoidance and anxiety about attachment that individuals have towards parents. The EPR is a 22-item instrument modified from 22 items of the ECR scale (Brennan et al., 1998). The EPR's avoidance subscale consists of such items as "Just when my father/mother started to get close to me, I found myself pulling away," "I preferred not to be too close to my father/mother," "I did not feel comfortable opening up to my father/mother," and "I found it difficult to allow myself to depend on my father/mother." Some of the anxiety over abandonment items includes, "When my father/mother disapproved of me, I felt really badly about myself," "I worried about being abandoned by my father/mother," "I worried a lot about my relationship with my father/mother," and "I got frustrated when my father/mother was not around as much as I would have liked." Each item is rated on a 7-point Likert scale (1 = disagree strongly, 4 = neutral / mixed; 7 = agree strongly). Internal consistency was high ([alpha]s > .84) across both subscales during initial validation as well as during the current study ([alpha]s > .83).

Attachment to God. We used the AGI (Beck & McDonald, 2004) to assess participants' attachment bonds to God. The AGI is based on two subscales that measure anxiety about abandonment by God and avoidance of intimacy with God. The AGI has 28 items--14 items each on both the anxiety and the avoidance subscales. The anxiety subscale consists of such items as "I often worry about whether God is pleased with me," "I fear God does not accept me when I do wrong," and "I worry a lot about my relationship with God." The avoidance subscale consists of such items as "I prefer not to depend too much on God," "I am uncomfortable allowing God to control every aspect of my life," and "I just don't feel a deep need to be close to God." The AGI had good internal consistency coefficients during validation (Avoidance = .86, Anxiety = .87; cf. Beck & McDonald, 2004) as well as during the current study (Avoidance = .92, Anxiety = .88).

Demographic and background characteristics.

Participants also completed a demographic questionnaire regarding gender, age, ethnicity, religious affiliation, marital status, family structure (individuals with whom individuals lived between the ages of 0 and 16), and the current marital status of their parents.

Procedure

Participants registered for the study using an online registration and tracking system. Only participants who were at least 18 years of age were allowed to participate. Once registered, participants were directed to an online survey session lasting approximately 60 minutes. Participants completed measures assessing their attachment relationships to parents, attachment to God, and a demographics' questionnaire. Participants also completed measures not used in these analyses. These measures were counterbalanced by the online survey system (see www.surveymonkey.com) to eliminate contamination effects.

Results

Due to limited samples of participants with "married and separated parents" (N = 4), "widowed parents--fathers deceased" (N = 15), "widowed parents--mothers deceased" (N = 3), "never married and currently living together" parents (N =1), "never married and currently not together" parents (N = 16), and "no relationship" between participants' father figure and mother figure (N = 7), we limited our analysis to test whether differences exist in attachment to God by family structure to participants with currently married parents (N = 160) and with divorced parents (N = 82). (Note: Analyses excluding an additional five participants whose age identified them as outliers did not differ from analyses including them, so the participants were not permanently removed from subsequent analyses.)

First, we used a multivariate analysis of variance (MANOVA) to examine differences in avoidance and anxiety towards fathers, mothers, and God by parents' marital status (married or divorced). There were differences by parents' marital status in attachment anxiety towards mothers, F(1, 240) = 5.28, p = .02, attachment avoidance towards fathers, F(1, 240) = 33.95, p < .001, and attachment anxiety towards fathers, F(1, 240) = 18.67, p < .001. Specifically, individuals with divorced parents reported higher attachment anxiety towards mothers (M = 34.81), attachment avoidance towards fathers (M = 47.29), and attachment anxiety towards fathers (M = 40.34) than individuals with married parents (M = 30.86, M = 36.49, and M = 31.86, respectively). A multiple correlation also revealed that among individuals with divorced parents, time spent with fathers--indicated by the average number of days per month individuals saw their fathers following the divorce--was associated with avoidance towards fathers, r(72) = -.38, p = .001, such that the greater the number of days with their fathers, the lower the attachment avoidance towards fathers individuals reported. Similarly, time spent with mothers was linked to the avoidance towards mothers, r(72) = -.40, p < .001, such that the greater the number of days with their mothers, the lower the attachment avoidance towards mothers individuals reported.

Next, we used hierarchical multiple regressions to examine the relationship between attachment to fathers, attachment to mothers, and attachment to God (see Table 2). On Step 1, we entered the main effect predictors for parents' marital status (married or divorced), father attachment, and mother attachment. On Step 2, we entered two-way interactions involving parents' marital status (i.e., those effects in which parental marital status would be a moderator of the relationship between parental attachment and attachment to God). On Step 3, we entered three-way interactions involving both types of attachment to parents and parents' marital status (e.g., mother avoidance, mother anxiety, and marital status; i.e., those effects in which parental marital status moderated the potential interaction between attachment avoidance and attachment anxiety representing secure or fearful attachment). All variables were centered for the purpose of interpreting interactions (Aiken & West, 1991). The first model (main effects only) was significant for attachment avoidance towards God ([R.sup.2] =.06, p = 02). That is, parents' marital status and parent attachment accounted for 6% of the variance in attachment avoidance towards God. Specifically, attachment avoidance towards fathers predicted attachment avoidance towards God, [beta] = .20, t(236) = 2.69, p = .008, such that the higher the attachment avoidance towards fathers, the higher the attachment avoidance towards God. There were no other significant effects found.

The first (main effects only) and second (two-way interaction effects added) models were significant for attachment anxiety towards God (second model: [R.sup.2] = .17, p < .001). That is, parents' marital status and parent attachment accounted for 17% of the variance in attachment anxiety towards God. Specifically, both attachment anxiety towards fathers and attachment anxiety towards mothers predicted attachment anxiety towards God, [beta]s > .19, ts(236) > 2.40, ps < .02, such that the higher the attachment anxiety towards fathers and mothers, the higher the attachment anxiety towards God. However, two interactions qualified these main effects. Specifically, there was an interaction between attachment avoidance towards fathers and parents' marital status (see Figure 1), [beta] = -.19, t(232) = -2.74, p = .007. That is, high attachment avoidance towards fathers only predicts attachment anxiety among children of married (but not divorced) parents. Similarly, there was an interaction between attachment anxiety towards fathers and parents' marital status for attachment anxiety towards God (see Figure 2), [beta] = .15, t(232) = 2.05, p = .04. That is, among those low in attachment anxiety towards fathers, children of married parents reported higher levels of attachment anxiety towards God than did children of divorced parents.

Discussion

The current study extends previous research on the link between attachment to parents and attachment to God. Similar to previous research (e.g., Beck & McDonald, 2004; Limke & Mayfield, 2011; McDonald et al., 2005), this study suggests support for the correspondence hypothesis; that is, children of married parents seemed to project attachment to their fathers (and to a lesser extent, their attachment to their mothers) onto their attachment to God. In contrast, however, children of divorced parents did not engage in this correspondence. Instead, at high levels of avoidance to fathers, children of divorced parents manifested low levels of anxiety towards God, suggesting that they sought attachment with God to compensate for the lack of relationships they never had with absent fathers (i.e., on average, participants reported seeing their fathers eight days per month post-divorce). Thus, this study also found support for the compensation hypothesis for these participants. Because the reported median age of these children at divorce was 8.67 years (SD = 5.57), it is possible that children of divorced parents wrote off the absent parent from their memories. In fact, results indicated that the number of days participants saw their parents following divorce was associated with avoidance towards those parents; however, this avoidance did not translate to avoidance towards God, perhaps because those parents were no longer seen as central figures in the participants' lives.

This finding is congruent with the results of McDonald et al. (2005), suggesting that individuals from unstable or inconsistent backgrounds might seek relationships with God that compensate for insecure and lacking relationships with parents. Although these individuals may initially exhibit secure bonding with God, they may experience sudden and radical conversions--especially if they are from nonreligious backgrounds--once they settle down in their relationship with God; thus, their relationship with God would come to correspond to the anxiety that characterized their inadequate and uncertain relationship with primary caretakers (cf. Kirkpatrick & Shaver, 1990).

Strengths and Limitations

Limitations of this study may include the sample characteristics (e.g., age and religious affiliation). That is, because the average age of this sample was 21.37 (SD = 5.23) and over 75% of the sample reported some type of Christian affiliation, the study might contain issues with generalization. College samples are convenient samples that may generally tend not to mirror general society. For example, college samples tend to have a higher percentage of educated participants than the general public. Exposure to education and educational material about religious beliefs sets this sample apart from the general population. Also, older samples might tend to be more conservative in their religious views than this sample. It is possible that older samples would not abandon their religious beliefs in the way that this college sample did.

Similarly, the somewhat high percentage of the sample that reported some type of religious affiliation and an average of 2.74 (SD = 3.29) monthly church attendance might limit this study and may suggest that the sample was influenced by their religious beliefs in responding to the surveys on attachment to God and defensive theology. In addition, the surveys assessed participants' retrospective attitudes towards parents but did not assess current attitudes and experiences towards parents. It is possible that current participants' attitudes and experiences towards parents might be more important in accounting for participants' recollections of experiences with parents during the first 16 years of life. Finally, a pertinent question is whether a comprehensive understanding of attachment to God is feasible. Although the role of individuals in attachment to God may be sufficiently described and understood, the finite nature of human knowledge severely limits attempts to explore the role of God in this relationship because God cannot be reduced to a variable to be quantified. At best, researchers can only theorize about this subject.

Strengths of the current study include a large sample, diversity in the characteristics of the sample, and established data collection materials. The large UCO sample (N = 288) increased the power of the research, thereby facilitating the evaluation of the research questions. That is, the large sample facilitated the finding of differences. Also, unlike previous studies on attachment to God that have utilized overly religious samples (e.g., Beck, 2006c; Beck & McDonald, 2004; Cassiba et al., 2008), this study recruited participants from a non-religious university who had not been exposed to theological classes that might have influenced their responses to survey items.

Future Directions

This research provides many avenues for future research. First, there may be questions regarding the AGI's ability to distinguish between individuals who are securely attached to God and individuals who are merely indifferent to God altogether. That is, the scale itself (although useful in determining secure attachment as the interaction of the two subscales and low scores on each) may not be able to make this differentiation. Specifically, although preliminary analyses of the current data suggest that there is no overall difference in attachment avoidance towards God or anxiety about abandonment by God, Fs (1, 240) [less than or equal to] 2.05, ps [greater than or equal to] .16, and religious affiliation and family structure were independent of each other, [chi square] %2 (2, N = 242) = 4.27, p = .12, participants whose parents were married did report currently attending religious services more often each month (M = 3.24, SD = 3.71) than individuals whose parents were divorced (M = 1.96, SD = 2.58), t(238) = 2.79,p = .006. It may be premature, however, to suggest that individuals who are indifferent toward God appear as securely attached individuals. Specifically, preliminary analyses suggest that (when grouped into three religious affiliation categories: none, Christian, and other) although individuals with no religious affiliation (i.e., those who identify themselves as atheist or agnostic) report less anxiety about abandonment (M = 32.43, SD = 2.35) than individuals who are Christian (M = 42.11, SD = 1.11), F(2, 239) = 7.23, p = .001, they also report more avoidance toward God (M = 75.83, SD = 2.50) than both individuals who are Christian (M = 43.25,SD = 1.81) and individuals who selected other religious affiliations (M = 57.91, SD = 3.29), F(2, 239) = 72.65, p < 001. The use of a less religious sample might be helpful in addressing this issue as well as allowing researchers to investigate to what/ whom non-religious individuals are referring when completing the instrument.

Because previous research has tended to examine the relationship between defensive theology and existential threats (e.g., mortality salience processes), further research is needed to investigate the relationship between attachment to God and defensive theology. For example, would securely attached individuals from nonreligious backgrounds respond to a religiously-inspired threat by embracing a defensive perspective or an existential perspective? Because the current study found evidence for both correspondence and compensation hypotheses, the answer to this question might clarify the relationship between attachment to God and defensive theology among nonreligious individuals. Another area of interest for the future should be the growing phenomenon of same-gender families. This new family genre should become a focus of future research of the association between family structure vis-a-vis attachment to God.

Conclusions

In conclusion, this research found differences in the link between attachment to parents and attachment to God between individuals with married parents and individuals with divorced parents. Specifically, among participants with married parents, attachment to God seems to work as a correspondence with attachment to father (i.e., attachment to father is projected onto attachment to God). In contrast, among participants with divorced parents, attachment to God may be a compensation for an absent father. Therefore, these findings align with attachment theory's proposition that children are pre-programmed to seek proximity with caretakers and thus develop attachment styles that compliment the responsiveness of their caretakers to maintain that attachment. For clinicians, targeting interventions towards ameliorating the home environment and parental psychological presence (i.e., secure bases) could address the fears of abandonment or avoidance of intimacy. Furthermore, clinicians and marriage advocacy groups could design interventions targeted at lowering acrimony in cases of separation, divorce, and custody battles while promoting parental access to children. In summary, the current research extends previous research and opens up new avenues for future research on attachment to God.

References

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage.

Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of the Strange Situation. Hillsdale, NJ: Erlbaum.

Barfield-Cottledge, T. (2015). The triangulation effects of family structure and attachment on adolescent substance use. Crime & Delinquency, 61, 297-320.

Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology, 61, 226-244.

Beck, R. (2004). The function of religious belief: Defensive versus existential religion .Journal of Psychology and Christianity, 23, 208-218.

Beck, R. (2006a). Communion and complaint: Attachment, object-relations, and triangular love perspective on relationship with God. Journal of Psychology and Theology, 34, 43-52.

Beck, R. (2006b). Defensive versus existential religion: Is religious defensiveness predictive of worldview defense? Journal of Psychology and Theology, 34, 142-152.

Beck, R. (2006c). God as a secure base: Attachment to God and theological exploration. Journal of Psychology and Theology, 34, 125-- 132.

Beck, R., & McDonald, A. (2004). Attachment to God: The Attachment to God Inventory, tests of working model correspondence, and an exploration of faith group differences. Journal of Psychology and Theology, 32, 92-103.

Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. New York, NY: Basic Books.

Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation: Anxiety and anger. New York, NY: Basic Books.

Bowlby, J. (1980). Attachment and loss: Vol. 3. Loss. New York, NY: Basic Books.

Bowlby, J. (1988). A secure base: Parent-child attachment and healthy human development. London: Routledge.

Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measures of adult romantic attachment. An integrative overview. In J. A. Simpson & W. L. Rholes (Eds.), Attachment theory and close relationships. New York, NY: Guilford Press.

Cassiba, R., Granqvist, P., Costanini, A., & Gatto, S. (2008). Attachment and God representations among lay Catholics, priests, and religious: A matched comparison study based on Adult Attachment Interview. Developmental Psychology, 44, 1753-1763.

Cicirelli, V. G. (2004). God as the ultimate attachment figure for older adults. Attachment & Human Development, 6, 371-388.

Cooksey, E. C., & Fondell, M. M. (1996). Spending time with his kids: Effects of family structure on fathers' and children's lives. Journal of Marriage and Family, 58, 693-707.

Collins, N. L., & Read, S. J. (1990). Adult attachment, working models, and relationship quality in dating couples. Journal of Personality and Social Psychology, 58, 644-663.

DeMuth, S., & Brown, S. L. (2004). Family structures, family processes, and adolescent delinquency: The significance of parental absence versus parental gender .Journal of Research in Crime and Delinquency, 41, 58-81.

Flouri, E., & Buchanan, A. (2002). Life satisfaction in teenage boys: The moderating role of father involvement and bullying. Aggressive Behavior, 28, 126-133.

Freeney, J. A., & Noller, P. (1990). Attachment style as a predictor of adult romantic relationships .Journal of Personality and Social Psychology, 58,281-291.

Goodsell, T. L., & Meldrum, J. T. (2010). Nurturing fathers: A qualitative examination of child-father attachment. Early Child Development and Care, 180, 249-262.

Greenberg, J., Pyszczynski, T., Solomon, S., Rosenblatt, A., Veeder, M., Kirkland, S., & Lyon, D. (1990). Evidence for terror management theory II: The effects of mortality salience on reactions to those who threaten or bolster the cultural -worldview. Journal of Personality and Social Psychology, 58, 308-318.

Griffin, K. W., Botvin, G. J., Scheier, L. W., Diaz, T., & Miller, N. L. (2000). Parenting practices as predictors of substance abuse and aggression among urban minority youth: Moderating effects of family structure and gender. Psychology of Addictive Behaviors, 14, 174-184.

Hart, J. T., Limke, A., & Budd, P. R. (2010). Attachment and faith development. Journal of Psychology and Theology, 38, 122-128.

Hazan, C., & Shaver, P. R. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52, 511-524.

Kirkpatrick, L. A., & Shaver, P. R. (1990). Attachment theory and religion: Childhood attachments, religious beliefs, and conversion. Journal for the Scientific Study of Religion, 18, 266-275.

Limke, A., Amin, S., Poojar, R., Kamble, S. V. (2015, May). Attachment to God across three major world religions: Fathers and mothers matter differently. Poster session presented at the annual conference of the Association for Psychological Science, New York, NY.

Limke, A., & Mayfield, P. B. (2011). Attachment to God: Differentiating the contributions of fathers and mothers using the Experiences in Parental Relationships Scale. Journal of Psychology and Theology, 39, 122-129.

Main, M., & Solomon, J. (1990). Procedures for identifying infants as disorganized/disoriented during the Ainsworth Strange Situation. In M. T. Greenberg, D. Ciccheti, & E. M. Cummings (Eds.), Attachment in the preschool years: Theory, research, and intervention (pp. 121-160). Chicago, IL: University of Chicago Press.

McDonald, A., Beck, R., Allison, S., & Norsworthy, L. (2005). Attachment to God and parents: Testing the correspondence vs. compensation hypotheses. Journal of Psychology and Christianity, 24, 21-28.

Miner, M. (2009). The impact of child-parent attachment, attachment to God and religious orientation on psychological adjustment. Journal of Psychology and Theology, 37, 114-124.

Moore, K. A., Jekielek, S. M., &: Emig, C. (2002). Marriage from a child's perspective: How doesfamily structure affect children, and what can we do about it? A research brief submitted to the United States District Court Northern District of California. Retrieved from http://www.casamariposa.org/Prop8/Attachments/DIX0026.pdf

Roberson, P. N. E., Sabo, M., & Wickel, K. (2011). Internal working models of attachment and postdivorce coparent relationships. Journal of Divorce & Remarriage, 52, 187-201.

Rosenblatt, A., Greenberg, J., Solomon, S., Pyszczynski, T., & Lyon, D. (1989). Evidence for terror management theory: I. The effects of mortality salience on reactions to those who violate or uphold cultural values. Journal of Personality and Social Psycholog)/, 57, 681-690.

Sampson, R.J., & Laub, J. H. (1994). Urban poverty and family context of delinquency: A new look at structure and process in a classic study. Child Development, 65, 523-540.

Schroeder, R. D., Osgood, A. K., & Oghia, M. J. (2010). Family transitions and juvenile delinquency. Sociological Inquiry, 80, 579-604.

van Ijzendoorn, M. H. (1995). Adult attachment representations, parental responsiveness, and infant attachment: A meta-analysis on the predictive validity of the Adult Attachment Interview. Psychological Bulletin, 117, 387-403.

Wong, M. S., Mangelsdorf, S. C., Brown, G. L., Neff, C., & Schoppe-Sullivan, S. J. (2009). Parental beliefs, infant temperament, and marital quality: Associations with infant-mother and infant-father attachment. Journal of Family Psychology, 23, 828-838.

Maurice S. Murunga

Southern Nazarene University

Alicia Limke-McLean

University of Central Oklahoma

Ronald W. Wright

Southern Nazarene University

Author Note: Maurice K. Murunga, Graduate Program in Counseling, Southern Nazarene University; Alicia Limke-McLean, Department of Psychology, University of Central Oklahoma; Ronald W. Wright, Department of Psychology and Counseling Southern Nazarene University.

Some of these data were collected in partial fulfillment of the requirements for Maurice S. Murunga's thesis project for the Master of Science degree.

Correspondence concerning this article should be addressed to Alicia Limke-McLean, Department of Psychology, University of Central Oklahoma, 100 N. University Drive, Edmond, OK 73034. E-mail: alimke@uco.edu

Author Information

MURUNGA, MAURICE S. Address: 4130 N. Lincoln Boulevard, Oklahoma City, OK 73105. Email: mmurunga@mail.snu.edu Title: Psychotherapist. Degrees: BA (Political Science & Literature in English), University of Nairobi; MA (Political Science--International Affairs), University of Central Oklahoma; MS (Counseling Psychology), Southern Nazarene University. Specializations: Psychological well-being, particularly including aspects of culture and faith.

LIMKE-McLEAN, ALICIA. Address: Department of Psychology, University of Central Oklahoma, 100 N. University Drive, Edmond, OK 73034. Email: alimke@uco.edu Title: Associate Professor, Department of Psychology. Degrees: BS (Psychology), Southern Nazarene University; MS (Psychology), University of Oklahoma; PhD (Psychology), University of Oklahoma. Specializations: Social/ personality psychology, particularly relationship processes.

WRIGHT, RONALD W. Address: Department of Psychology and Counseling, Southern Nazarene University, 6729 NW 39th Expressway, Bethany, OK 73008. Email: rwright@snu.edu Title: Clinical Psychologist and Chair/Professor of Psychology. Degrees: BA (Psychology), Mount Vernon Nazarene College; MA (Theology), Fuller Theological Seminary; PhD (Clinical Psychology), Fuller Theological Seminary. Specializations: Integration of psychoanalysis and Wesleyan theology as well as empirical research investigating the relationship of death anxiety and attachment states of mind to spiritual development and the functioning of faith.

Caption: FIGURE 1

Adjusted predicted values for individuals' attachment anxiety towards God, illustrating the interaction between individuals' attachment avoidance towards fathers and parents' marital status.

Caption: FIGURE 2

Adjusted predicted values for individuals' attachment anxiety towards God, illustrating the interaction between individuals' attachment anxiety towards fathers and parents' marital status. TABLE 1 Descriptive Statistics for Demographic Characteristics Variable Percent Marital Status Single, never been married, not living with 80.6% significant other Single, never been married, living with 8.7% significant other Divorced 2.8% Married 7.3% Other 0.7% Household Structure (Before Age 16) Both biological/adoptive parents 68.1% One biological parent (single parent) 14.9% One biological parent and stepparent 12.8% (or parent's significant other) One biological parent and another relative 2.4% No biological parent present (e.g., lived 1.7% with relative) Religious Affiliation Other 2.1% Atheist 5.6% Agnostic 9.7% Pagan 1.7% Unitarian/Universalist 0.3% Deist 0.3% Taoist 0.3% Other Indian Affiliation 0.3% Hindu 0.3% Buddhist 1.4% Jewish 0.7% Muslim 1.0% Christian--Roman Catholic 8.7% Christian--Orthodox (including Greek, 0.3% Eastern, and Oriental) Christian--Baptist (including Southern 27.8% Baptist and Free Will) Christian--Methodist (including United 6.9% Methodist and Nazarene) Christian--Lutheran (including Presbyterian 3.1% and Anglican) Christian--Pentecostal (including Assembly 4.2% of God and New Age) Christian--Non-denominational evangelical 19.4% Christian--Restorationist (including Church 3.1% of Christ) Christian--Mormon 0.7% Christian--Jehovah's Witness 0.3% Christian--Church of Christ, Scientist 1.0% TABLE 2 Hierarchical Regressions of Attachment to God onto Measures of Attachment to Father and Mother Avoidance towards God Cumulative Increase Predictors [R.sup.2] in [sr.sup.2] sr [R.sup.2] Step 1 .0 6** .06 ** Parental marital .00 .03 status (PMS) Father--attachment .03 ** .17 ** avoidance Father--attachment .00 -.05 anxiety Mother--attachment .01 .11 avoidance Mother--attachment .00 -.01 anxiety Step 2 .08 ** .03 PMS x Father .00 -.06 attachment avoidance PMS x Father .01 -.09 attachment anxiety PMS x Mother .01 -.08 attachment avoidance PMS x Mother .00 .05 attachment anxiety Step 3 .10 ** .02 PMS x Father avoid x .00 -.07 Father anxiety PMS x Mother avoid x .01 -.09 Mother anxiety Anxiety about Abandonment by God Cumulative Increase Predictors [R.sup.2] in [sr.sup.2] sr [R.sup.2] Step 1 .13 *** .13 *** Parental marital .01 -.10 status (PMS) Father--attachment .00 .04 avoidance Father--attachment .02 * .15 * anxiety Mother--attachment .00 .00 avoidance Mother--attachment .03 ** .17 ** anxiety Step 2 .17 *** .04 * PMS x Father .03 ** -.16 ** attachment avoidance PMS x Father .02 ** .12 * attachment anxiety PMS x Mother .00 -.02 attachment avoidance PMS x Mother .00 .03 attachment anxiety Step 3 .18 * .01 PMS x Father avoid x .00 -.07 Father anxiety PMS x Mother avoid x .00 -.04 Mother anxiety