Abstract

Family researchers interested in the links between religion and family life have been challenged to move beyond a simplistic linear conceptualisation of religion (how much religion). Studies in the psychology of religion suggest that the way in which religious beliefs are internalised and the degree of openness with which religious notions are held (what kind of religion) are associated with issues of mental health and interpersonal relationships (Batson, Schoenrade, & Ventis, 1993). Thus, the present study hypothesised that what kind of religion measures would be more strongly associated with the nature of family relationships than how much religion measures. Study participants comprised a national sample of 996 Seventh-day Adventists aged 19-93 years. The results supported the view that how an individual engages in religious practice (what kind of religion) is more strongly linked to the quality of family relationships than how much religion that individual reports. Participants reporting rigid and literalistic or guilt driven approaches to religion reported an increased emphasis on control and difficulty in expressiveness in relationship with children, and lower levels of satisfaction, difficulties in communication, and support for traditional gender roles in relationship to marital partners. In contrast, participants who identified with and maintained an open approach to religious sentiment tended to promote independence in children and were more likely to engage children in affectionate and warm relationships and experience increased marital satisfaction. The findings suggest a parallel in kind between the internal dynamics of religious motivation and way in which relationships are constructed with family members.

The field of studies loosely known as psychology of religion has a long and rich heritage of approaching religion from this what kind of religion approach (see Batson, Schoenrade, & Ventis, 1993; Schumaker, 1993). Generally, the weight of evidence suggests that how an individual goes about being religious is of more significance for the quality of social interaction than whether he or she is religious or not. Gordon Allport's work stands as the classic example of approaching the conceptualisation and measurement of religion in this way (Allport, 1950; Allport & Ross, 1967). Stimulated by findings which indicated that racial prejudice increased with more frequent church attendance, Allport set about to explain this counter-intuitive result by proposing differing motivational orientations to religious belief and practice. His resulting typology of intrinsic and extrinsic religious orientations has dominated research in the field of religious studies for nearly three decades. An intrinsic religious orientation was described as driven by an internalisation of religious values and sentiment which was associated with a more mature and tolerant self in relation to other. In contrast, Allport commented that the extrinsic religious orientation was typical of those individuals who had turned to God but had not turned away from themselves. Within an extrinsic orientation, religion was used as a vehicle for furthering self interests and, not surprisingly, was shown to be associated with prejudicial attitudes.

Although Allport's theoretical work on religious orientation has been tremendously beneficial in helping to define some of the more essential ingredients of the relation between religion and social relationships, the intrinsic-extrinsic dichotomy has not been without its criticisms. A number of researchers have found difficulties with measurement issues, while others call for the psychology of religion field to move beyond the intrinsic-extrinsic dichotomy (Batson, Schoenrade, & Ventis, 1993; Gorsuch, 1994; Kirkpatrick & Hood, 1990). For example, Batson and his colleagues maintain that the intrinsic orientation to religion (as it is currently measured) has omitted three important ingredients of Allport's description of mature faith‹a capacity for facing the complex problems of life without reducing their complexity, a readiness to doubt and to be self-critical, and a perspective of openness and tentativeness (Batson, Schoenrade, & Ventis, 1993). In support of these views Batson reports studies linking a more open 'quest' orientation to religion with cognitive complexity, whereas intrinsic religious orientation was negatively linked to cognitive complexity (Batson & Raynor-Prince, 1983). The major difficulty with the intrinsic-extrinsic approach has been the repeated finding that high scores on the intrinsic scale do not distinguish between a rigid dogmatic approach to religion and a more authentic and flexible approach to religion (Batson & Ventis, 1982).

In spite of its shortcomings the intrinsic/extrinsic dichotomy was able to establish the importance of an internalised religious faith. A more recent attempt to describe the ambiguous territory of religious motivation has been made by Richard Ryan (Ryan, Rigby, & King, 1993). Ryan approaches the religious terrain from the perspective of his earlier description of the internalisation process (Deci & Ryan, 1985) to describe two forms of religious internalisation. Identification describes a form of internalisation where religious beliefs and practices have been assimilated to the self so that these religious values are experienced as having arisen from one's own volition.

In contrast religious values can also be internalised in the form of an introjection. In introjected regulation religious practices are performed because one has a sense of 'should' or 'oughtness' about the behaviour. Not to perform the behaviour would result in anxiety or guilt or loss of self-esteem. Inevitably, there is a sense of pressure and inner conflict that comes with an introjected religious orientation. Deci and Ryan (1985) describe introjection as an internally controlling state in which affective and self-esteem contingencies are applied to enforce compliance with a specific set of regulations. Ryan et al (1993) found that the identification with religious belief and practices was associated with higher levels of self-actualisation and self-esteem and lower levels of depression and anxiety than was the introjection of religious beliefs.

A second important issue in mapping the terrain of religious motivation is the particular degree of openness to experience that an individual might bring to their pursuit of religious practice. For example, religious practice can be construed in terms of a particular dogma and held to rigidly, or alternatively, religious sentiments can be held in such a way where they are open to revision and understood as those values and beliefs that make life purposeful and relationships meaningful. Batson attempted to capture this aspect of religious sentiment by the use of his 'quest' dimension (see review by Batson, et al, 1993). Alternatively, a number of researchers have utilised a measure of fundamentalism to capture a sense of individuals being closed to new experiences (Kirkpatrick, 1993; McFarland, 1989). Not surprisingly these researchers found that literalistic and fundamentalist approaches to religion were associated with discriminative attitudes towards various social minority groups.

In several studies at Avondale College with student and community samples we have consistently found negative correlations between McFarland's fundamentalism scale and measures of tolerance for ambiguity. We have also found robust correlations between the fundamentalism scale and defensiveness (as measured by Eysenck's lie scale). Typically, we have found strong negative correlations between fundamentalism and Batson's quest scale. In this sense McFarland's fundamentalism scale appears to capture a closed, rigid approach to religion that is associated with discriminatory attitudes, defensiveness, and difficulty in tolerating ambiguity or open-endedness.

In the present study we opted to adopt Ryan's measures of identification and introjection as measures of different styles of internalising religious sentiment. In addition, we adopted the measure of fundamentalism developed by McFarland (1989) in an attempt to capture the dimension of openness or closedness in approaching religious belief and practice. McFarland's measure of fundamentalism contains a high proportion of items describing a literalistic approach to biblical literature and seemed particularly appropriate for our sample.

In the present study we have referred to the two measures of religious internalisation and the measure of literalism as the 'what kind of religion' measures, as opposed to indicators of 'how much religion' (church attendance, proportion of income donated to the church, and frequency of family worship). The major hypothesis of the present study was that the what kind of measures would be more strongly linked to the quality of family relationships than the how much religion measures. We further conjectured that the pattern of links between religion and family relationships would represent a parallel in kind between the internal organisation of religious motivation and the external world of family relationships.

The sample consisted of 996 subjects; 431 (45%) were males and 546 (55%) were females, 19 subjects did not complete the gender item. Subjects ranged in age from 19 to 93 years (M=51.2, SD=15.6). All participants were members of the Seventh-day Adventist church in Australia.

Instrument

The 134-item questionnaire included the religion measures of the identification with or introjection of religious beliefs and practice (Ryan et al, 1993), religious literalism (McFarland, 1989), frequency of church attendance, proportion of finance donated to churches, and frequency of family worship. Each of the religious orientation scales comprised six items rated on a 6-point likert scale. The items are listed in appendix A, and described more fully in Strahan and Craig (1995).

Measures used to assess the nature of family relationships included 32 items from the Child-Rearing Practices Report (McNally, Eisenberg, & Harris, 1991). These 32 items contained eight scales that assessed the following aspects of the parent-child relationship: control, independence, achievement orientation, punishment, enjoyment of child, negative affect, expressiveness, and rational guidance. A factor analysis supported the structure of the scales and reliability analyses provided strong support for the internal consistency of the scales with alpha coefficients ranging from .74 to .95.

The nature of the marital relationship was assessed by the Enrich Marital Satisfaction scale (EMS, Fowers & Olson, 1993). The EMS included 10 items assessing marital satisfaction and a 5-item realistic distortion scale. In addition, two items were included to assess the degree to which individuals went out of their way to avoid conflict in the marriage relationship, and three items which assessed attitudes towards gender roles in marriage. Each of the above scales was shown to have adequate reliability according to estimates of internal reliability.

Procedure

A random sample of all SDA churches in Australia was taken. From this subset of Australian churches, the names and addresses of married church members were obtained from church rolls. Questionnaires were then direct mailed to randomly selected individuals from these church rolls. Of the 2,294 questionnaires mailed, 996 were returned by mail. An additional 81 questionnaires were returned marked 'Return to Sender', producing an effective response rate of 45%.

Relations between the religious orientation scales

Table 1 presents correlations between the religious orientation scales for men and women in the sample. The results indicated that a literalistic approach to religion was strongly associated with identifying with religious belief and practice. Literalistic approaches to religion were also moderately associated with a guilt driven approach to religious experience. The two styles of internalising religious belief and practice were not related to each other indicating that these approaches to religion operated quite independently of each other. The pattern of correlations was similar for men and women.

The links between how much religion and what kind of religion

Both bi-variate and partial correlations were computed between the 'how much religion' and the 'what kind of religion' measures. Partial correlations were included for several reasons. First, all of the religion measures, except introjection, were significantly correlated with a measure of distortion. Secondly, the religious orientation measures were related to each other (see Table 1), and thirdly, religion was influenced by age and gender. Therefore, partial correlations were calculated with variations brought about by distortion, age, and the interrelatedness of the orientation scales controlled. These correlations are presented in Table 2.

Under each of the religious orientation measures, two coefficients are presented. The coefficient on the left represents the bi-variate correlation (a measure of the relation between the two measures, with no consideration given to the relation each variable has with other variables), and the coefficient on the right represents the partial correlation (a measure of the relation between the two measures once the influence of other variables had been controlled).

The robust bi-variate correlations between literalism and the how much measures were dramatically reduced when distortion, age, and the internalisation variance was held constant. This trend suggested that the bi-variate correlations between literalism and the 'how much measures' were artifacts of the relation between literalism and the other measures, particularly identification. It appears that a literalistic approach to religion in and of itself was not strongly predictive of participation in religious activities. Two exceptions occurred, although the correlations were weak. For males, literalism was associated with increased church attendance; for females, literalism was associated with frequent family worship.

In contrast, the strong connection between personal identification with religious belief and practice and the actual performance of religious activities and the donating of financial resources, that was evident for both males and females, remained significant and was only marginally reduced when literalism, distortion and age were held constant. The robust partial correlations between identification and the 'how much measures' suggest that it is the personal identification with religious belief and practice that is most likely to motivate actual church attendance, the giving of finances to the church and other religious causes, and the practice of family worship.

The introjection scale was not strongly related to the 'how much measures'. The correlations tended to be negative and none were significant for males. For females, a guilt driven religious experience was related to lower rates of participation in family worship and less willingness to donate financial resources to the church or other religious cause. These results indicated that it is impossible to predict the degree to which male's religious experience is motivated by feelings of personal guilt or fear of reprisal by observing the frequency with which they attend church, conduct family worship in the home, or by even knowing what proportion of income they donate to the church. For women, the guilt driven approach was actually negatively related to participation in religious ritual and giving.

The links between religion and family relationships

The critical results for the present study are presented below in two tables. Table 3 presents partial and bi-variate correlations between what kind of religion and the measures of family relationships. Table 4 presents partial and bi-variate correlations between the how much religion measures and the family measures. Again, partial correlations were used because of the inter-correlations between the religion measures and the effects of age and gender on both the religion and family measures.

An overview of the results presented in Table 3 suggested that variations in religious orientation was associated in meaningful ways with the nature of family relationships. For example, a literalistic approach to religion was associated with an increased emphasis on the control of children, the inhibition of children's independence in parent-child relationships, and reports of difficulty in enjoying laughter and play with children. In the marital relationship the literalistic approach to religion was associated with lower levels of marital satisfaction, support for traditional gender roles, and difficulties in openly dealing with conflict.

In contrast, the identification with religious belief and practice was associated with a turning away from an emphasis on control towards fostering independence in children. In addition, identification was associated with a sense of openness and expressiveness in parent- child relationships and an increased ability to engage in conversation about children's successes and failures. In the marriage relationship identification with religious belief and practice was associated with higher levels of marital satisfaction.

Similar to the literalistic approach to religion, an introjection of religious belief and practice was associated with an emphasis on control and the inhibition of children's independence in parent-child relationships. However, introjection was also associated with pushing children towards public performance and achievement, an increased likelihood of interactions with children characterised by anger and conflict, and an increased willingness to punish children. In the marriage relationship the introjection of religious beliefs was associated with lower levels of marital satisfaction and support for traditional gender roles.

Table 4 presents the results from the correlations between the indicators of how much religion and the measures of family relationships. The first impression gained from this table of results is the scarcity of significant coefficients. In the parent-child relationship the correlations are low and mixed, with little support for the notion that religiosity is related to an increased benefit in the parent-child relationship. In marriage relationships 'more religion' was associated with increased marital satisfaction, but also with support for traditional gender roles. A simple comparison of Tables 3 and 4 indicates that what kind of religion participants adopted was more strongly predictive of the nature of family relationships than how much religion they reported.

The above results present several issues that warrant further discussion. Firstly, although the identification and literalism measures were strongly correlated, these two measures were associated with very different approaches to family relationships. A literalistic approach to religion seems to be associated with relationships that are oriented to control and stereotypical roles rather than the spontaneous enjoyment of children and partners. It seems that this approach to religion is associated with the construction of family relationships along the lines of rules and specific roles rather than any real open engagement of another individual in conversation.

Such an interpretation of the world of literalism has been confirmed in a recent study where we found very strong correlations between literalism and a normative approach to self-construction (Strahan, 1996). Given that literalism has also been shown to correlate with measures of an intolerance of ambiguity, defensiveness, and prejudicial attitudes, it does seem that the inner world of literalism could be characterised by a rigid preoccupation with control and the construction of self in terms of social stereotypes. If this is the way the literalist constructs a sense of self, it is not surprising that the self-other relationships of the literalist is constructed along parallel lines.

In a similar way, the inner world of introjection could be characterised by the self in conflict and the drive to gain approval by compliance to powerful introjected others. Introjected religiosity has been associated with a diffuse approach to self construction, where the diffuse approach is thought to grow out of a lack of stable internal structures and a reliance on the contingencies of the external world for decision making (see Berzonsky, 1989; Strahan, 1996). It is then not surprising, that a diffuse sense of self and a coercive approach to participating in religious practice is associated with coercive and conflictive relationships with children and stereotypical relationships with marital partners. It is no wonder that Ryan et al (1993) reports that introjected religion was associated with higher levels of anxiety and increased vulnerability to symptoms of depression, and lower levels of self-actualisation.

Both the literalism and introjection scales suggest a control orientation to religious belief and practice. Literalism, it seems, is associated with a control-of-other orientation to religion and to family relationships that limits the individual's capacity for emotional expressiveness and tolerance of individuality and engages in self-other interaction in terms of social stereotypes rather than any sense of openness and acceptance. The inner world of introjection seems to be preoccupied with a 'control-of-self in order to gain approval from other' orientation to religious experience and family relationships. Perhaps a control-of-self orientation in the face of somewhat overwhelming inner chaos.

Such an interpretation of the results is supported from studies that have found that USA Christians supporting a literalistic interpretation of biblical passages are more prone to rely on corporal punishment as a primary method of discipline (Ellison & Bartkowski, 1994; Ellison & Sherkat, 1993, 1994). In fact, McNamara (1985) suggests that the emphasis on authority and obedience in the homes of many conservatively religious people mirrors a broader preoccupation with authority and control that prevails within the institutions of conservative Christians.

In contrast to a control orientation to religion and self-other relationships, the identification with religious beliefs and practices was associated with family relationships that gave clear evidence of openness, emotional warmth, and the capacity for fostering individuality in others. It seemed that the identification with religious belief and practice fostered a growth orientation to both religion and family relationships and could be understood as growing out of the essential congruence of self, possible because of the advanced internalisation of beliefs and values. Assuming that the inner world of identification is characterised by a strong sense of congruence of self and self-determination, then it is not difficult to imagine why the interaction between self and other is marked by openness, warmth and an acceptance of other. Presumably such an organisation of the inner self would gives rise to religious identifications and warm growth promoting family relationships.

The results of the present study has a number of implications for religious families and communities, and secondly, for the research community. Firstly, the results suggest that there are significant connections between the way religion is experienced and the nature of family relationships. Secondly, religious families would do well to explore the kind of family interactions that might foster the internalisation of religious values. There is ample evidence from the research literature that suggests that the family is a vital social context that can either promote or inhibit the internalisation of values (Deci, Eghrari, Patrick, & Leone, 1994; Hauser, Powers, Noam, Jacobson, 1984; Strahan, 1994). The pattern of results in the present study suggests that a growth orientation to religion might foster the internalisation of religious values whereas a control orientation to religion might result in the inhibited development of children. The results also raise significant questions about rigid, literalistic religious experience. It is somewhat ironic that the supporters of fundamentalist religion often take the high moral ground and posture as protector of family values, yet deliver so little of substantial value to families.

The present study has an interesting set of implications for religious communities. Firstly, religious communities might want to think about the nature of social relationships that is most likely to foster the internalisation of religious values as opposed to the introjection of religious values. For example, Baard (1993) reports that church members' identification with religion was associated with perceptions of the congregation as promoting individuality and autonomy among its members, fostering opportunities for members to experience themselves as competent, and a sense of relatedness with other members of the congregation. Secondly, the results from the present study has some implications for our understanding of healthy religion. Religious experience that is characterised by rigidity and needs for control, or guilt and need for approval has been clearly associated with family relationships that could be recognised as dysfunctional.

In addition to the above, there are several implications for family professionals and researchers interested in mapping the terrain connecting religion and the nature of family relationships. First, the results show considerable support for the view that religion is a complex construct and that simple how much religion measures do not really capture the importance of the religion construct. There is a need for a more careful appraisal of the measures of religion that are used. Secondly, measures of strength of commitment to religious values or intrinsic religious motivation do not distinguish between a reflective and open approach to religion and a more rigid closed approach. The issue of openness to experience does seem to be a vital dimension in discriminating between various approaches to religion. Finally, the results of the present study suggest that religious belief and practice is important for the quality of life that many Australians enjoy or endure. To ignore the religion variable in our attempts to better understand family relationships seems to me to be somewhat petty. Religion is an important influence in the lives of many Australians, and we would do well to offer ourselves and others a more informed view about the costs and benefits of religious belief and practice.

Finally, the present study set out to enhance our understanding of the territory in the connection between religion and family relationships. Although the study is limited by its specific denominational focus and its cross-sectional design the results have suggested that how individuals go about being religious is of more importance for the quality of their family relationships than how much religion they report. Religion that is characterised by openness and the encouragement of personal volition is most likely to foster the quality of family relationships that most of us recognise as healthy. Finally, the results have been interpreted as being representative of the intriguing parallels between the inner world of religious motivation and the outer world of family relationships.

Allport, G.W., & Ross, J.M. (1967). Personal religious orientations and prejudice. Journal of Social Psychology and Personality, 5(4), 432-443.

Baard, P.P. (1994). A motivational model for consulting with not-for-profit organizations: A study of church growth and participation. Consulting Psychology Journal, 46(3), 19-31.

Batson, C.D., Schoenrade, P., & Ventis, W.L. (1993). Religion and the individual: A social- psychological perspective. Oxford: Oxford University Press.

Berzonsky, M. (1989). The self as theorist: Individual differences in identity formation. International Journal of Personal Construct Psychology, 2, 363-376.

Deci, E.L., Eghrari, H., Patrick, B.C., & Leone, D.R. (1994). Facilitating internalization: The self-determination theory perspective. Journal of Personality, 62, 119-142.

Deci, E.L., & Ryan, R.M. (1985). Intrinsic motivation and self-determination in human behaviour. New York: Plenum Press.

Ellison, C.G., & Bartokowski, J.P. (1994). Religion and legitimation of violence: The case of conservative Protestantism and corporal punishment. In The web of violence: From interpersonal to global. L. Kurtz, & J. Turpin, (eds). Urbana: University of Illinois Press.

Ellison, C.G., & Bartkowski, J.P., & Segal, M.L. (1995). Conservative Protestantism and the corporal punishment of toddlers and preadolescents. University of Texas at Austin. Manuscript under review.

Ellison, C.G., & Sherkat, D.E. (1993a). Conservative Protestantism and support for corporal punishment. American Sociological Review, 58, 131-144.

Ellison, C.G. & Sherkat, D.E. (1993b). Obedience and autonomy: Religion and parental values reconsidered. Journal for the Scientific Study of Religion, 32, 313-329.

Fowers, B.J., & Olson, D.H. (1993). ENRICH marital satisfction scale: A brief research and clinical tool. Journal of Family Psychology, 7, 176-185.

Grolnick, W.S., & Ryan, R.M. (1989). Parent styles associated with children's self-regulation and competence in school. Journal of Educational Psychology, 81, 143-154.

Hauser, S.T., Powers, S.I., Noam, G.G., Jacobson, A.M. (1984). Familial contexts of adolscent ego development. Child Development, 55, 195-213.

Johnnson, W.G. (1995). The real issue at stake. Record, August 12, 7.

Kirkpatrick, L. (1994). Fundamentalism, Christian orthodoxy, and intrinsic religious orientation as predictors of discriminatory attitudes. Journal for the Scientific Study of Religion, 32, 256-268.

McFarland, S.G. (1989). Religious orientations and the targets of discrimination. Journal for the Scientific Study of Religion, 28, 324-336.

McNally, S., Eisenberg, N., & Harris, J.D. (1991). Consistency and change in maternal child-rearing practices and values: A longitudinal study. Child Development, 62, 190-198.

McNamara, P.H. (1985). The new Christian right' view of the family and its social science critics: A study in differing presuppositions. Journal of Marriage and the Family, 47, 449-458.

Ryan, R.M., & Rigby, S., & King, K. (1993). Two types of religious internalization and their relations to religious orientations and mental health. Journal of Personality and Social Psychology, 65(3), 586- 596.

Schumaker, J.F. (Ed.) (1992). Religion and mental health. Oxford: Oxford University Press.

Sorenson, E.S. (1989). Religion and family health: A need for study. Family Science Review, 2, 303-316.

Strahan, B.J. (1994a). Parents, adolescents and religion. Cooranbong, NSW: Avondale Academic Press.

Strahan, B.J. (1994b). The relationships between Christian orthodoxy and fundamentalism and personality: An empirical investigation among Seventh-day Adventists. Unpublished paper, Avondale College.

Strahan, B.J. (1995). Religious orientation and tolerance for ambiguity: An empirical study among Seventh- day Adventists students. Unpublished paper, Avondale College.

Strahan, B.J., & Craig, B. (1995). Marriage family and religion. Sydney: Adventist Institute of Family Relations.

Strahan, B.J. (1996). Parallels in religious orientation and approaches to self-theorising: A constructivist perspective. Unpublished paper, Avondale College.

Thomas, D.L., & Cornwall, M. (1990). Religion and family in the 1980s: Discovery and development. Journal of Marriage and Family, 52, 983-992.

Literalism Scale Items, McFarland (1989)

I am sure the Bible contains no errors or contradictions
It is very important for true Christians to believe that the Bible is the infallible word of God
The Bible is the final and complete guide to morality; it contains God's answers to all important questions about right and wrong.
Christians must try hard to know and defend the true teachings of God's word.
Christians should not let themselves be influenced by worldly ideas.
The best education for a Christian child is in a school with Christian teachers.

Identification Scale, Ryan et al (1993)

I share my faith because God is important to me and I'd like others to know him too
I pray because I find it satisfying
I turn to God because I enjoy spending time with Him
I attend church because by going I learn new things
I turn to God because it is satisfying
I often experience the joy and peace which comes from knowing I am right with God

Introjection Scale

I pray because God will disapprove if I don't
I turn to God because I'd feel guilty if I didn't
I share my faith because I want other Christians to approve of me
I attend church because others would disapprove if I didn't
I attend church because one is supposed to go
I actively share my faith because I'd feel bad if I didn't