Pastoral Psychology, Vol. 52, No. 3, January 2004 (© 2004)
"Seeing Conversion Whole": Testing a Model of Religious Conversion1
Peter J. Kahn23 and A. L. Greene2
Rambo (1993) theorized that religious conversion consists of a process involving seven dimensions, labeled context, crisis, quest, encounter, interaction, commitment, and consequences. To test the Rambo model, a new measure, the 97-item Adult Religious Conversion Experiences Questionnaire (ARCEQ), was developed, revised on the basis of feedback from a focus group, and administered by mail to 110 adult volunteers who self-identified as having experienced conversion. Reliability analysis of the subscales of the ARCEQ resulted in five (crisis, quest, encounter, commitment, and consequences) meeting the criterion (a > .80). A sixth (interaction) achieved a reliability of .76 and was retained, but the seventh (context) was dropped. Principal components factor analysis with Varimax rotation produced factors labeled Redemptive Love, Zealous Dedication, Dysphoric Need, Openness to Uncertainty, Extrinsic Crisis, and Experiential Learning, which could readily be interpreted in relation to Rambo's model. Subsidiary analyses showed a number of significant correlations between demographic characteristics of the sample and differences in religious conversion experience.
KEY WORDS: religious conversion; conversion process stage models; conversion types; adult religious experience; factor analysis; empirical model tests.
Western psychology has largely conceptualized religious conversion as individually experienced, intrapsychic change. Early on, William James set the standard for this perspective by defining religion as "the feelings, acts, and experiences of individual men [sic] in their solitude, so far as they apprehend themselves to stand in relation to whatever they consider the divine" (James, 1902/1994, p. 36).
1 This paper is based on a doctoral dissertation conducted by the first author, under the supervision of the second author.
2 Pacific Graduate School of Psychology, Palo Alto, CA.
3 Address correspondence to Peter J. Kahn, Ph.D., Carrufh Center for Counseling and Psychological Services, P.O. Box 6422, West Virginia University, Morgantown, WV 26506-6422; e-mail: peter.kahn@mail.wvu.edu or algreenephd@aol.com.
233
0031-2789/04/0100-0233/0 © 2004 Human Sciences Press, Inc.
234
Kahn and Greene
In the years since, conversion has most often been viewed in one of two ways: (a) as the result of the individual's passive experience of a precipitating crisis (the "classic" perspective; Zinnbauer and Pargament, 1998); or (b) as the result of the individual's more active quest for personal meaning (the "contemporary" perspective; Zinnbauer and Pargament, 1998). Despite differences in their respective construals of the individual's passivity versus activity, both "classic" and "contemporary" approaches share a common construal of religious conversion as an individual, intrapsychic phenomenon.
However, people are embedded in multiple, overlapping contexts (Bronfenbrenner, 1979) that exercise profound influences on both the process of religious conversion and its consequences. For example, Gordon (1967) has delineated the influence of the microcontext of marriage on religious conversion, Kraft (1992) has noted the importance of the cultural phenomenon of group conversion, and Rambo (1993) cites a contemporary example of mass conversion in B. R. Ambedkar's leading two million of his fellow Hindu Untouchables to convert to Buddhism on a single day, October 14, 1956. It would appear that Western notions of individualism have exerted a powerful, though largely unexamined, bias in our understanding of religious conversion (e.g., Hudson, 1992; Kraft, 1992). Such assumptions have also limited our understanding of religious conversion, as well as our appreciation of the consequences of conversion experiences for the individual.
A PROCESS MODEL OF RELIGIOUS CONVERSION
Recently, a model of religious conversion has been forwarded (Rambo, 1993) that, theoretically at least, addresses these concerns. Rambo defines religious conversion as "turning from and to new religious groups, ways of life, systems of belief, and modes of relating to a deity or the nature of reality" (1993, p. 3). The breadth of this definition reflects the fact that what constitutes conversion is defined differently by various groups and underscores the importance of adopting an inclusive, phenomenological stance. Rambo is at pains to restrict his definition to the realm of religion rather than a vaguer "spirituality" (p. 177, n. 2) and, in so doing, both clarifies and limits the boundaries of conversion as adherence to, or defection from, some organized religious body. As described below, Rambo's model consists of seven broad stages: (a) context, (b) crisis, (c) quest, (d) encounter, (e) interaction, (f) commitment, and (g) consequences.
Context
The influence of context (from the microcontext of the family to the macro-context of the society or nation) underlies the entire conversion process and forms the total environment in which conversion occurs. Rambo's emphasis on the power of the multiple contexts to inhibit or facilitate conversion is a useful corrective to
"Seeing Conversion Whole"
235
the purely intrapsychic approaches noted above. As he notes, "most scholars of conversion agree that few conversions take place in areas with well-organized, literate religions supported by the economic, political, and cultural powers of the region" (1993, p. 47).
Crisis
The psychological literature has focused extensively on the notion of crisis as the precipitant to conversion (e.g., Galanter, Babkin, Babkin, and Deutsch, 1979; James, 1902/1994; Ullman, 1982, 1988, 1989). This focus has frequently involved the concomitant, "classic" view of the convert as acted upon, or passive, in the conversion process. Psychodynamic approaches to conversion, for example, have theorized that religious conversion functions as a solution to unresolved and reemergent childhood conflicts—whether the solution is construed as an integrative and adaptive experience (Allison, 1969), or as an attempt to repress or control sexual or aggressive impulses (Ullman, 1982). Rambo's model recognizes the centrality of crisis in many conversions, but does not accord it primacy.
Quest
While not minimizing the importance of the crisis dimension in many conversions, Rambo also notes the importance of an alternate "contemporary" perspective in which potential converts are seen as active shapers of their own experience. The quest motif highlights the potential convert as actively seeking something "more" in religious life. That "more" may involve seeking new beliefs to replace ones that have come to seem inadequate, seeking new emotional experiences in a connection to God and other believers, seeking new ways of coping with life's problems, or some combination of the three.
Encounter
Studies of conversion often focus almost exclusively on the convert (e.g., Galanter et al, 1979; James, 1902/1994; Ullman, 1982,1988,1989), and, as noted above, on the intrapsychic correlates of conversion. Yet, in fact, encounters between potential converts and religious advocates (or missionaries) involve a dynamic interaction in which reciprocal needs are met. The cultural, social, personal, and theological characteristics of the advocate contribute importantly to the potential convert's experience. For example, on an individual level, demographic variables such as age, race, gender, and socioeconomic status, plus qualities that are harder to measure such as attractiveness and charisma, play an important role. On a social level, one consideration may be the relationship of the advocate to larger organizations or groups. On a cultural level, there is the possible appeal of the
236
Kahn and Greene
exotic or the forbidden. And on the theological level, there is the advocate's theory of conversion itself, of what constitutes an adequate conversion, and how much of a change the advocate's theology requires in the potential convert's beliefs, way of life, and existing social context. For the potential convert, the encounter stage represents an opportunity to consider such benefits as a new system of meaning, emotional gratifications (e.g., relief from guilt), techniques for living, and access to power (e.g., divine healing).
Interaction
In the interaction phase, the potential convert either chooses to continue contact and become more involved, or the advocate works to extend the encounter stage with the goal of persuading the person to convert. Again, there is a spectrum of activity and passivity, for as Rambo notes, some religions do not encourage conversions (e.g., ludaism, Orthodoxy). The key conceptual element of this phase in the conversion process is encapsulation. Encapsulation is the creation of an environment in which communication and social interaction are controlled in order to allow the potential convert to experience the new religious ideas free of countervailing pressures. Encapsulation can be physical, social, or ideological, and the control can range from total to partial. Once a sphere of influence is established by encapsulation, four dimensions of interaction are deployed in the conversion process: (a) relationships, which create and consolidate emotional bonds to the group; (b) rituals, which provide, through the repetition of physical actions, "embodied or holistic knowledge" as opposed to mere cognitive understanding; (c) rhetoric, which includes new uses of language, new metaphors that serve as "a vehicle for the transformation of consciousness," and new narratives that provide a reframed perspective of the self and the self's place in the larger scheme of things; as well as (d) roles, which provide new expectations of behaviors consonant with the potential convert's ascribed status within the group.
Commitment
In the commitment stage, the potential convert decides to affiliate with the new religion. At this point, the commitment decision is often dramatized and commemorated with a public demonstration of the convert's choice. Rituals are important for their symbolic enactment of the commitment decision, and typically involve a symbolic repudiation of the old self and an embracing of a new identity. Rituals of conversion represent more than just confirmatory public performances; they also embody the decision in physical ways that transcend mere cognitive assent.
It is in the commitment phase that the experience of "surrender" is found. At least in the Christian tradition, religious writers from Augustine to Eldridge Cleaver (cited in Kerr and Mulder, 1983) have maintained that an essential, and
"Seeing Conversion Whole"
237
perhaps the essential, element in the experience of surrender comes from outside the individual, that is, from God or the transcendent dimension of experience. This dramatic inbreaking of the transcendent, or "transforming moment" (Loder, 1989), is often at the heart of surrender. However, the experience of surrender is often fragile and precarious, and difficult for the convert to sustain. One of the ways some religious groups attempt to sustain commitment against the possibility of backsliding is through testimony, a narrative witness of conversion. Testimony employs the particular rhetoric or language system of the group in a process of active reconstruction and reinterpretation of personal history. Its public aspect both consolidates the person's commitment to the group and serves as a powerful reminder of the community's basic values and goals.
Consequences
The consequences phase of the conversion process requires a long temporal perspective. The implications of even a sudden surrender experience may take much processing for the convert, and a gradual conversion process may require many years to assimilate fully. Psychological studies examining the adaptive versus regressive quality of conversion experiences (e.g., Allison, 1969; Galanter et al., 1979; Ullman, 1982) have given little consideration to longer-term (theological) consequences, such as the person's evolving relationship with God, sense of mission and reason for living, and changed understanding of the nature of reality. In Erikson's terms (1968), the consequences phase involves consideration of the role of the conversion in the person's resolution of the life-course challenges of identity, intimacy, generativity, and integrity.
Conversion as Process
Although conversion is thought to occur in sequential fashion, Rambo (1993) emphasizes that the sequence is highly individual and is likely to involve a recursive aspect. A given conversion experience may include some but not all of the phases described, and individual phases may vary in intensity and duration. Moreover, there is no requirement of linearity in the process, for phases may and frequently do overlap and recur. In this way, Rambo's model differs from developmental/stage models, such as Fowler's (1981) on faith development and Kohlberg's (1986) on moral development, which build on Piagetian notions of developmental advances that reorganize and thus qualitatively alter cognitive processes.
In Rambo's model, an interaction resulting from encounters between an advocate for a particular religion and a potential convert may result in a quest on the latter's part, which may be facilitated or hindered by the context. This quest, in turn, may precipitate further encounters and interactions, the onset of a crisis, and finally commitment and its consequences. By the same token, another person's
238
Kahn and Greene
conversion process may begin quite differently, with dissatisfaction that stimulates a quest, leading to interaction with members of a new religious grouping, followed by disengagement and a further quest, an encounter with someone whose religious explanations make sense, and a decision or commitment without any felt sense of crisis at all.
Just as there is no single linear process leading to conversion, so there is no single type of conversion. Rambo identifies five conversion types, distinguished chiefly by the social and cultural distance involved in making the conversion: (a) apostasy, or defection from a person's existing religious tradition to no religious involvement (e.g., from born-again Christian to atheist); (b) intensification, or a revitalization of interest and participation in a person's existing religious tradition (e.g., from nominal to devout Catholic); (c) affiliation, or a person's movement from no religious involvement to full involvement with a particular religious group or movement (e.g., from agnostic to Buddhist); (d) institutional transition, or a person's movement from one faith community to another within a major religious tradition (e.g., from Methodist to Orthodox within Christianity); and (e) tradition transition, or a person's movement from one major religious tradition to another (e.g., from Christian to Muslim).
Integrating Rambo's typology of conversion and his process model yields a promising framework within which to better study religious conversion experiences. The model is phenomenological, incorporating earlier process models (e.g., Downton, 1980; Lofland and Stark, 1965; Tippett, 1977), and provides a broad, multidimensional theory of religious conversion. Despite recognition of its promise for bringing religious conversion into the broader stream of personality research (e.g., Paloutzian, Richardson, and Rambo, 1999), Rambo's model has been accorded little empirical attention. The purpose of our study, therefore, was to attempt an empirical test of this model; that is, to determine whether the seven dimensions of religious conversion hypothesized by the model (context, crisis, quest, encounter, interaction, commitment, consequences) emerge in the self-reported conversion experiences of adults. For purposes of our investigation, religious conversion was defined as: (a) joining a religious denomination (affiliation) or changing religious affiliation (institutional transition or tradition transition); (b) experiencing an intensification of religious feeling or of involvement within the same denomination where the participant nonetheless identified this experience as conversion (intensification); or (c) leaving a religious affiliation where the participant identified the experience as conversion apostasy.
METHOD Sample
The sample consisted of 65 female and 45 male participants (M age = 50.9 years, SD — 14.5 years, range = 25-84 years), who self-identified as having
"Seeing Conversion Whole"
239
experienced religious conversion. The sample was predominantly white (89%) and well-educated (65% with four years of college or more). Fifty-one percent of the sample were engaged in professional occupations or self-employed, 18% were in white collar occupations, 6% in blue collar occupations, 9% were home-makers, and 14% were unemployed or retired. Participants were drawn from 11 states, with the largest representations from Oklahoma (52%), Tennessee (19%), and California (16%). Thirty percent of questionnaires were distributed to people who attended a presentation and volunteered to participate; the remaining 70% were distributed by mail.
Development and Refinement of the Adult Religious Conversion Experiences Questionnaire (ARCEQ)
Individual items were developed based on Rambo's (1993) descriptions of the salient experiences characterizing six of the seven dimensions of religious conversion. The 12-item Batson and Schoenrade (1991a, 1991b) Quest scale fit the characteristics of Rambo's quest dimension and was incorporated in the ARCEQ, without alteration. This process resulted in a preliminary pool of 101 items (i.e., 89 new items, plus the 12-item Batson and Schoenrade Quest scale).
A focus group (Morgan, 1997) was convened to obtain a preliminary test of the instrument. Five adults, aged 30 and over, were recruited through a newspaper advertisement. Participants met, as a group, to complete the preliminary ARCEQ, to give feedback about how well the items covered the content of their religious conversion experiences, and to suggest wording and content changes. As a result of the focus group, several ARCEQ items were revised for greater clarity, a number of items were deleted, and five new items were added. The final version of the ARCEQ used for the main survey was a 97-item instrument.
Main Survey
For the main survey, equal numbers of female and male religious converts were sought. The lower age limit for the sample was set at 25 years old. Participation was entirely voluntary, with no remuneration provided. Recruitment relied heavily on a snowball technique, whereby religious personnel were approached and informed about the study and invited to identify and contact converts within their congregations, or to announce the research through bulletin announcements and newsletters. In addition, the study was announced at religious services, Sunday School classes, discussion forums, evening classes, and other congregational gatherings, where attendees were invited to participate directly.
Interested participants were invited to call a toll-free 800 number. Calls were returned, questions about the research were answered in a general way, and questionnaire packets were mailed to interested participants with a stamped return
240
Kahn and Greene
envelope. Follow-up calls were made after two weeks and again after four weeks if protocols had not been returned.
Measures
The research protocol contained the following: (a) an information cover page; (b) the revised, 97-item ARCEQ; (c) a brief demographic questionnaire; (d) a narrative item eliciting the individual's open-ended account of the conversion process; and (e) the Marlowe-Crowne Social Desirability Scale (Crowne and Marlowe, 1960).
Religious Conversion Experience
As described above, the 97-item ARCEQ consisted of seven subscales. All items used a 9-point Likert format (1 = disagree strongly, 9 — agree strongly). The 17-item context subscale assessed the facilitating or inhibiting effects of familial, social, cultural, and historical influences on the potential convert. The 11-item crisis subscale assessed the influence of personal problems, interpersonal difficulties, unusual or traumatic events, and other, profound experiences on a person's religious conversion. The 12-item encounter subscale assessed the varied experiences of contact between the advocate for a particular religion and the potential convert. The 17-item interaction subscale tapped socialization influences in the new religious option after the initial encounter; specifically, the degree to which potential converts were encouraged to form relationships with members of the new religion, participate in its rituals, learn its rhetoric, and adopt new roles. The 15-item commitment subscale assessed the consummation phase of conversion, as reflected in decision-making and the rituals confirming it, personal surrender, giving testimony, and responding to challenges to the convert's stated motivations for conversion. The 13-item consequences subscale assessed the ramifications of conversion, both short- and long-term.
Finally, and as noted previously, the quest dimension was measured using Batson and Schoenrade's (1991b) 12-item Quest Scale. In previous studies (Batson and Schoenrade, 1991b), internal consistency reliabilities for the Quest Scale have ranged between .75 and .82. Adequate validity of the scale was evidenced by its high correlations (.85-90) with the older 6-item Quest Scale (Batson and Ventis, 1982) which had amassed considerable empirical support for its validity, but suffered from low internal consistency (Batson and Schoenrade, 1991a, 1991b).
Demographics
Demographic items obtained information regarding participant gender, age, ethnicity, occupation, educational level, parents' religion, own current and prior
"Seeing Conversion Whole"
241
religious affiliation, age at conversion, and how the participant learned about the study.
Social Desirability
The 33-item Marlowe-Crowne Social Desirability Scale (Crowne and Marlowe, 1960) was included to control for the potential influence of social desirability in self-reported conversion experiences. Participants responded (true/false) to questions about socially approved and culturally acceptable behaviors which were, at the same time, unlikely to occur. The Marlowe-Crowne Scale has good reported validity (high correlations with other social-desirability scales and with the MMPIL and K scales) and reliability (a Kuder-Richardson internal consistency coefficient of .88) (Crowne and Marlowe, 1960).
RESULTS Preliminary Analyses
Examination of item frequencies revealed that no ARCEQ items were missing more than five percent of the cases. Therefore, linear interpolation was used to estimate and replace missing values for use in all subsequent analyses.
The next task was to determine whether internal consistency reliabilities for the newly developed subscales met the minimum criterion (.80; Tabachnick and Fidell, 1996) for inclusion in subsequent data analyses. This task was accomplished in two steps: first, alpha coefficients were calculated using the entire set of items for a given subscale; second, item analysis was used to eliminate items which contributed only minimally to or suppressed the overall subscale alpha.
Reliabilities for the crisis, encounter, and consequences subscales were in the acceptable range. However, the context, interaction, and commitment subscales did not yield acceptable alphas. No amount of item elimination produced a reliability coefficient greater than .58 for the context subscale and, accordingly, it was dropped from further analyses. An 8-item version of the interaction subscale yielded a reliability coefficient of .74. This was judged sufficiently close to .80 to warrant retaining the subscale in subsequent analyses so as to test as much of the model as possible. Selective elimination of items on the commitment sub-scale improved its alpha above the criterion. Both the Social Desirability (Crowne and Marlowe, 1960) and Quest (Batson and Schoenrade, 1991b) scales yielded acceptable reliabilities (i.e., .84 and .83, respectively). These efforts produced a final pool of 54 items, representing six of the original seven subscales. Final scale alpha coefficients, subscale means, and standard deviations are shown in Table 1.
242 Kahn and Greene
Table 1. Results of Reliability Analysis of the ARCEQ Subscales, Before and After Item Elimination
No. of
Original alpha
No. of       Final        Final Final final items    alpha    scale mean    scale SD
Subscale
original items
Context
Crisis
Quest
Encounter
Interaction
Commitment
Consequences
17
11
12
12 17 15
13
.49 .85 .83 .81 .58 .69 .86
11
8 6
8
9
12
.85 .88 .85 .74 .81 .90
33.7 44.1 28.5 35.0
40.8 83.0
19.6 16.1 14.4 11.8 15.0 19.4
Note. No amount of item elimination produced an alpha of better than .58 for the context subscale; therefore, the scale was eliminated from further analysis.
Thirty-seven different denominations were represented among participants' and parents' self-reported religious affiliations, exclusive of descriptions such as "born-again Christian" or "Protestant." To facilitate analysis, the array of denominations for participants' current religion, previous religion, mothers' religion, and fathers' religion were collapsed to eight categories (i.e., Catholic, Conservative/Evangelical Protestant, Liberal/Mainline Protestant, Unemphatic Christian/ Protestant, Other Christian, Non-Christian, Agnostic/Atheist, and None).
As shown in Table 2, each of Rambo's (1993) five conversion types was represented in the shift between participants' previous and current religious affiliation. For example, the bottom two rows show participants whose conversion was of the affiliation type, that is, the 14.5 percent of the sample who gave their previous religion as either agnostic/atheist or "none." The right-hand column shows the apostasy type, that is, the 1.8 percent of the sample who were previously Christian and reported their current adherence to agnosticism or atheism. Several cells show conversions of the tradition transition type, e.g. the 6.4 percent of the sample who reported a shift from conservative/evangelical Protestant to non-Christian and the 2.7 percent who reported a shift from non-Christian to liberal/mainline Protestant. The boldface percentages along the diagonal show conversions either of the intensification or institutional transition type (only one can be identified with certainty as intensification, the participant who gave both previous and current religious affiliations as Catholic).
Determining the Dimensions of Religious Conversion Experience
Factor analysis was used to test whether Rambo's dimensional model adequately captured participants' religious conversion experience. Because, as described previously, the context subscale was eliminated, only six factors were
Descriptive Analyses
Table 2. Participants' Current and Previous Religious Affiliations (Percent)
Current religion
		Conservative/Evangelical	Liberal/Mainline	Unemphatic	Other	Agnostic/
Previous religion	Catholic	Protestant	Protestant	Christian	Christian Non-Christian	Atheist
Catholic	0.9	2.7	4.5	0.9	1.8	
Conservative/Evangelical Protestant0	8.2	7.3	6.4		1.8 6.4	0.9
Liberal/Mainline Protestant''	4.5	4.5	11.8	0.9	5.5	
Unemphatic Christian	1.8	1.8	1.8	0.9		0.9
Other Christian''		0.9	1.8		0.9	
Non-ChristianB	0.9	0.9	2.7		0.9	
Agnostic/Atheist		0.9				
None	0.9	8.2	2.7	0.9	0.9	
a Includes those who identified themselves as "born again," "evangelical Christians," and "charsimatic Christians," as well as members of the following denominations: Assemblies of God, Baptist, Cooperative Baptist Fellowship, Church of Christ, Church of God, Evangelical Free Church, Free Methodist, Lutheran Church-Missouri Synod, Messianic Jewish, Pentecostal, and Southern Baptist.
'includes members of the following denominations: Anglican, Congregational, Disciples of Christ, Episcopal, Evangelical Lutheran (ELCA), Lutheran, Methodist, Presbyterian, United Church of Christ, and United Methodist, includes those who simply listed "Protestant" or "Christian," or said "nominal Christian."
^Includes Christian Science, Church of Jesus Christ of Latter-Day Saints, Orthodox, Greek Orthodox, Swedenborgian, and Unity. B Includes adherents of Judaism, Islam, Buddhism, and Unitarian-Universalism.
244
Kahn and Greene
specified. Alpha factor extraction was performed first, due to its potential usefulness in providing communalities that maximized coefficient alpha. Results obtained through alpha factoring were virtually indistinguishable from principal components extraction. Therefore, the more familiar and easily interpreted principal components method was chosen for the final analysis. To examine possible overlap among factors, a preliminary factor analysis was performed using principal components extraction and oblique rotation. Only a single, modest correlation was obtained, so there was no reason to prefer oblique rotation. Accordingly, orthogonal rotation was chosen for the final analysis.
The final factor analysis was performed using principal components extraction with Varimax rotation. Factor loadings less than 0.3 were suppressed, and 6 factors were specified in the solution (see Table 3). The Kaiser-Meyer-Olkin Measure of Sampling Adequacy was .72; Bartlett's Test of Sphericity = 3841.33, p < .001.
Factor 1
A total of 18 items loaded on the first factor extracted, accounting for 19.3% of the variance. The first factor received high loadings from items which appeared to tap a sense of purpose (e.g., ARCEQ 89, 93), connection with God and other people (e.g., ARCEQ 87, 88), and improved functioning (e.g., ARCEQ 97). Factor 1 also received moderate loadings from items which appeared to tap a sense of affective release (e.g., ARCEQ 72, 73, 91) and community involvement (e.g., ARCEQ 94, 95). Items loading on this factor shared a strong positive affective tone, and highlighted elements of love, joy, freedom, and new relationships with God, community, and self resulting from conversion. Factor 1, therefore, was interpreted to represent Redemptive Love.
Factor 2
Fifteen items loaded on the second factor, accounting for 10.8% of the variance. Factor 2 received moderate to high loadings from items involving a sense of total commitment of one's life and beliefs to the new religion (ARCEQ 64, 78, 83), as well as external manifestation of that commitment in the form of testimony and mission to convert others (ARCEQ 79, 80, 92). Factor 2 also received modest loadings from items reflecting encounters with people who made a personal witness, offered ways of coping, and embodied a special relationship with the divine (ARCEQ 43, 44, 50, 51), as well as items involving severed personal and family ties (ARCEQ 84), relief from the burden of sin (ARCEQ 90), and a new understanding of life (ARCEQ 96).
As compared to Factor 1, items which loaded on Factor 2 appeared to have little affective quality. However, items loading on Factor 2 shared a sense of the totality of the conversion experience, the importance of thinking and acting in
Table 3. Results of Factor Analysis Using Principal Components Extraction with Varimax Rotation
	ARCEQ Item	Factor 1	Factor 2    Factor 3	Factor 4    Factor 5    Factor 6 Commun
89.	My conversion has allowed me to love others more fully.	.83		.72
88.	Since my conversion, 1 have felt embraced by love.	.77		.62
97.	Since my conversion, 1 have better ways of coping with stress and problems.	.76		.64
93.	Since my conversion, 1 have had a renewed reason for living.	.72		.64
87.	My conversion brought me a new sense of relationship with God.	.72		.63
73.	When 1 decided to convert, 1 experienced a new feeling of freedom.	.68		.53
94.	After my conversion, 1 felt involved in a new community.	.67		.52
72.	When 1 decided to convert, 1 experienced tremendous j°y-	.66		.49
95.	My conversion gave me a new community of people to serve.	.64		.31 .53
91.	Since my conversion, 1 have felt a sense of relief from guilt.	.56		.33 .51
96.	Since my conversion, 1 have a new understanding of	.56	.31	.51
	the nature of life.			
90.	My conversion brought me a sense of relief from the burden of my sins.	.51	.44	.52
86.	1 see my conversion as the first step in a long process of transformation.	.50		.35 .50
50.	Meeting someone who embodied a special relationship with the divine contributed to my conversion.	.36	.34 .36	.41
78.	The religion to which 1 converted advocated total, complete, and sudden conversion.		.80	.68
79.	The religion to which 1 converted required personal testimony to show my commitment.		.78	.64
80.	Learning to give testimony was an important part of		.72	.64
my conversion experience.
Table 3. Continued
ARCEQ Item Factor 1
64. The religion I was considering stressed the truth of its own beliefs and the falsity of the beliefs of others in the outside world.
83. My conversion required me to make important changes .31 in many aspects of my life.
92. Since my conversion, I have had a sense of mission to .35 convert others.
77. My experience of surrender or complete inner yielding .42
of control marked a turning point in my conversion. 44. An encounter with someone who "shared the gospel" .32
with me contributed to my conversion. 76. Before my conversion, I attended ceremonies in which
other people committed themselves to the new faith. 43. A friend's or family member's personal witness about
her/his religion contributed to my conversion.
84. My conversion required me to sever many personal and family ties.
26. Difficulties in my personal relationships contributed to my conversion.
23. My unhappy childhood contributed to my conversion.
24. My troubled relationship with my mother contributed to my conversion.
25. My troubled relationship with my father contributed to my conversion.
52. Meeting someone who explained that her/his religion would accept me as I am contributed to my conversion.
22. My low self-esteem contributed to my conversion.
51. Meeting someone who explained how her/his religion could help me better cope with life contributed to my conversion.
49. Meeting someone whose religious explanations made sense contributed to my conversion.
Factor 2    Factor 3    Factor 4    Factor 5    Factor 6 Communality
.64 .39 .60
.64 .55
.58 .58
.49 .50
.43 .35 -.35 .55
.42 .28
.39 .30
.37 .22
.78 .72
.76 .67
.74 .57
.71 .58
.66 .65
.64 .40 .60
.31 .64 .59
.50 .44
27. My divorce/relationship breakup contributed to my conversion.
34. For me. doubting is an important part of what it means
to be religious. 33. It might be said that 1 value my religious doubts and
uncertainties. 38.1 am constantly questioning my religious beliefs. 40. There are many religious issues on which my views are
still changing. 37. As 1 grow and change, 1 expect my religion also to
grow and change. 31. My life experiences have led me to rethink my
religious convictions. 36. Questions are far more central to my religious
experience than are answers. 30.1 have been driven to ask religious questions out of a
growing awareness of the tensions in my world and in
my relation to my world. 19.1 had a near-death experience that contributed to my
conversion.
21.1 had experiences with alcohol and/or drags that
contributed to my conversion. 20.1 had an experience of illness and/or healing that
contributed to my conversion.
28. The effects of a crisis 1 experienced contributed to my conversion.
18.1 had a mystical experience that contributed to my conversion.
66. The religion 1 was considering stressed the importance of participating in ceremonies or rites with other members.
69. The religion 1 was considering encouraged everyone to take on certain roles and responsibilities.
.48 .43 .43
.80 .67
.79 .66
.77 .63
.73 .57
.72 .55
.69 .58
.64 .46
.37 .55 .48
.70 .51
.67 .51
.63 .60
.48 .60 .71
.60 .42
.66 .54
.64 .42
Table 3. Continued
ARCEQ Item	Factor 1	Factor 2	Factor 3	Factor 4    Factor 5	Factor 6	Communality
62. The religion 1 was considering expected me to spend a					.63	.48
lot of time in study, worship, and prayer.						
68. The religion 1 was considering stressed learning the					.61	.47
right way to talk about religious experiences.						
67.1 realized that the religion 1 was considering had a					.51	.35
special language for describing itself and its members.						
63. The religion 1 was considering expected me to wear					.42	.22
special clothing or follow certain rules about my						
personal appearance.						
65. The religion 1 was considering stressed the importance					.41	.27
of personal relationships with other members.						
Proportion of variance	19.3%	10.8%	7.1%	6.0% 5.1%	4.3%	
Proportion of covariance	24.4%	18.0%	18.0%	16.7% 11.6%	11.3%	
Note. Items appear in the order of highest to lowest loadings on each successive factor.
"Seeing Conversion Whole"
249
particular ways, and especially the importance of evangelizing. Factor 2, therefore, was interpreted to represent Zealous Dedication.
Factor 3
Thirteen items loaded on the third factor, accounting for 7.1% of the variance. This factor received high loadings from items reflecting relational difficulties of long standing (ARCEQ 23,24,25,26), and moderate loadings from items reflecting themes of low self-esteem and the new religion's promise of acceptance and help (ARCEQ 22, 51, 52). Factor 3, therefore, was interpreted to represent Dysphoric Need.
Factor 4
Ten items loaded on this factor, accounting for 6.0% of the variance. Factor 4 received high loadings from items emphasizing the positive aspects of religious doubt, questioning, and openness to religious change (ARCEQ 34, 35, 37, 38, and 40). Items with moderate loadings on this factor reflected themes of religious questioning as a result of life experiences (ARCEQ 30, 31) and the centrality of questioning in religious experience (ARCEQ 36). Taken together, items loading on the fourth factor appeared to reflect a common theme of openness to religious uncertainty in the form of doubts, questions, exploration, and change. Factor 4, therefore, was interpreted to represent Openness to Uncertainty.
Factor 5
Eight items loaded on the fifth factor, accounting for 5.1% of the variance. No item contributed a high loading, but moderate loadings came from items involving unusual, extreme, or life-threatening experiences (ARCEQ 18, 19, 20, 21). Also loading moderately on this factor was the "general crisis" item (ARCEQ 28). The external origin or extrinsic nature of precipitating events in each instance led this factor to be interpreted as representing Extrinsic Crisis.
Factor 6
Nine items loaded on the sixth factor, accounting for 4.3% of the variance. Factor 6 received moderate loadings from items reflecting active participation in ceremonies, rites, and worship preparatory to conversion, as well as seriousness about education in the faith (ARCEQ 62, 66, 68, 69). This factor was therefore interpreted to represent Experiential Learning.
250
Kahn and Greene
Subsidiary Analyses
Two sets of subsidiary analyses were conducted to contribute to the interpretation of factor analysis results. Factor scores were computed using the Anderson-Rubin method and used as dependent and criterion variables, as described below.
Group Differences in Religious Conversion
In the first set of analyses, group differences were examined through separate MANOVAs (multivariate analysis of variance) of religious conversion factor scores with gender, state of residence, type of data collection (in-person distribution versus mail), occupation, social desirability, parents' religious affiliation, and participants' previous and current religious affiliation as independent variables in each analysis.4'5 Significant differences in religious conversion were obtained as a function of: (a) type of data collection [F(6, 103) = 3.53, p < .01, rj1 — .17, power = .94]; (b) state of residence [F(18, 286.16) = 2.81, p < .001, rj2 = .37, power = 1.00]; and (c) current religious affiliation [F(36, 433.11) = 3.68, p < .001, rj1 — .69, power = 1.00]. Gender, occupation, social desirability (high/low), previous religious affiliation, and mothers' and fathers' religious affiliation were not significant.
Post hoc examination of the type of data collection effect revealed that participants who were mailed the research protocol were higher in Dysphoric Need than participants who received the research protocol following in-person presentation (p < .001). Post hoc examination of the residence effect revealed that Californians were higher than Oklahomans and Tennesseans in Dysphoric Need (p < .05), and higher than Tennesseans in Extrinsic Crisis (p < .05). Finally, post hoc examination of the current religious affiliation effect revealed that: (a) converts to Catholicism, conservative/evangelical Protestantism, liberal/mainline Protestantism, and other Christian denominations were all higher than agnostics and atheists in their experience of Redemptive Love (p < .05 for all contrasts); (b) converts to conservative/evangelical Protestantism were higher than Catholics, liberal/mainline Protestants, and non-Christians in their experience of Zealous Dedication (p < .05 for all contrasts); and (c) converts to non-Christian religions were higher than conservative/evangelical Protestants in their experience of Openness to Uncertainty (p < .05).
4State of residence was reduced to four levels (i.e., California, Oklahoma, Tennessee, and "Other"), reflecting the fact that 96 of the 110 participants lived in one of the first three states. Occupation was collapsed to three levels: blue collar/manual workers, white collar/office workers, and professional/self-employed.
5Scores on the Marlowe-Crowne Social Desirability Scale were dichotomized at the median, forming two groups (high/low) for analysis.
"Seeing Conversion Whole"
251
Predictors of Religious Conversion Experience
In the second set of subsidiary analyses, separate hierarchical multiple regressions of the six religious conversion factor scores were performed, with education, age, age at last conversion, and years since conversion as predictors in each analysis.6 A significant proportion of the variance in three of the six dimensions of religious experience was captured: (a) in Zealous Dedication (R2 — .09, F[l, 108] = 10.79, p < .01), (b) in Openness to Uncertainty (R2 = .09, F[l, 108] = 10.11, p < .01), and (c) in Extrinsic Crisis (R2 change = .05, F[l, 108] = 6.27, p < .05).
Examination of the corresponding beta weights revealed three points of interest. First, younger ages at conversion predicted higher Zealous Dedication scores (p < .01). By contrast, older ages at conversion were associated with higher Extrinsic Crisis scores (p < .05). Finally, higher education predicted greater Openness to Uncertainty (p < .01). Regression analyses of the Redemptive Love, Dysphoric Need, and Experiential Learning factor scores were nonsignificant.
DISCUSSION
Factor analysis results indicated that the ARCEQ performed quite well in capturing dimensions of religious conversion experience. An important external measure of its validity is the fact that the ARCEQ was well-received by ministers of various denominations and research participants themselves, who both commented favorably on its clarity and comprehensiveness. The fact that no items on the ARCEQ subscales had more than 5 percent missing data also suggests that the participants were readily able to appreciate the significance of the experiences described for religious conversion, whether or not they endorsed individual items as reflecting their own conversion experience. The ARCEQ's ready acceptance by converts as a reasonable, objective measure of their own conversion experience is an important finding in its own right, and suggests that continued refinement and validation of the ARCEQ would be of value to other researchers, as well.
The adequacy of our sample size also deserves consideration as authors (e.g., Comrey and Lee, 1992) typically recommend 300-500 cases at a minimum. Yet, recent findings (e.g., MacCallum, Widaman, Zhang, and Hong, 1999) indicate that good recovery of population factors can be achieved with sample sizes as small as 60 cases when the variable-to-factor ratio is at least 10:3 or 20:3 and commu-
6For all analyses, predictors were entered in separate blocks to control for the high intercorrelations observed: (a) between current age and years since conversion, r = .57, p < .001; (b) between age at conversion and years since conversion, r = —.57, p < .001; and (c) between current age and age at conversion, r = .35, p < .001. Educational level and current age were entered, simultaneously, in the first block, and years since conversion and age at conversion were entered, separately, in the second and third blocks. When significant Rs emerged, subsidiary analyses, varying order of block entry, were performed.
252
Kahn and Greene
nalities7 range between .2 and .8. As shown in Table 3, all of the communalities obtained in the present study were between .2 and .8, and the variable-to-factor ratio was 54:6. Thus, the present sample (n — 110) appears to have been adequate to test the model.
Finally, a number of considerations support our conclusion that the factor solution obtained is a robust one. First, there was minimal difference between the solution resulting from alpha factor extraction and principal components extraction. Second, there was only minimal overlap among the factors, as assessed through the preliminary analysis using oblique rotation. Third, there was minimal variable overlap among factors, since 38 of the 54 items in the variable set loaded uniquely on one factor. Fourth, interpretability of the factors was relatively easy because of the thematic clarity of clustered items, even when such items derived from different subscales.
The critical issue in any factor analytic investigation is the degree to which the obtained factors (here religious conversion experience) accord with the dimensions hypothesized by the theoretical model under study. To answer this question, we turn again to the results shown previously in Table 3. As shown there, it seems clear that the large first factor, Redemptive Love, aligns closely with Rambo's consequences phase. All 12 items for the final consequences subscale loaded on this factor, with 11 of those 12 items having loadings of .50 or higher. Although consequences is the last phase Rambo (1993) discussed, it makes sense that it should be the first factor to emerge, include the most variables, and account for the most variance in the solution, for in a sense the consequences phase is both the goal toward which the conversion experience is oriented as well as the dimension of religious experience in which the participants are currently living. Moreover, as Rambo (1993) noted, there are both immediate and long-term consequences of conversion, and thus it is possible that both recent and older converts continue to reap the benefits of conversion. The lack of correlation between participants' experience of Redemptive Love and the number of years since their conversion also indicates that this dimension is not subject to a recency effect, i.e., is not an immediate post-conversion pink cloud that dissipates with time.
The second factor, Zealous Dedication, drew its largest loadings from items on the commitment subscale of the ARCEQ and appears clearly related to Rambo's commitment phase in the conversion process. The sense of total commitment of one's life and beliefs, reflected in the items loading on this factor, highlights the distinction Nock (1933) made in his study of Christian conversion in the ancient world between "conversion" and mere "adhesion." Conversion in this absolute sense has certainly marked Christianity historically, but, in the United States, Zealous Dedication is most associated in the popular mind with "born-again" or
7"Communality" is the amount of variance an original variable shares with all other variables included in the analysis (Tabachnick & Fidell, 1996).
"Seeing Conversion Whole"
253
evangelical Christians. It is not surprising, then, to find that converts to the conservative/evangelical Protestant grouping, in our study, experienced higher levels of Zealous Dedication than did converts to Catholicism, liberal/mainline Protestant denominations, and non-Christian religions.
The third factor, Dysphoric Need, and the fifth factor, Extrinsic Crisis, are discussed together as, combined, they accounted for all the loadings on the 11 crisis subscale items. It seems clear that the dimension Rambo labels crisis has two distinct components among adults in this sample, reflecting the source of the difficulties driving their religious conversion. Dysphoric Need represents an intrinsic dimension, of early origin and relatively long duration. By contrast, Extrinsic Crisis appears situational, with the source of the crisis being chiefly external in origin, as the label indicates.
Dysphoric Need also received loadings from one of the quest items (ARCEQ 30: "I have been driven to ask religious questions out of a growing awareness of the tensions in my world and in my relation to my world"), the content of which, not surprisingly, involves dysphoric feelings. It also received loadings from five of the six encounter items, which together share a theme of finding a solution, whether in a particular person or in explanations, acceptance, or tools for coping. This connection between Dysphoric Need and a solution-offering encounter is at least suggestive of a particular conversion dynamic, involving psychological vulnerability and openness to influence that may make the potential convert vulnerable to manipulation, as in the case of recruitment to cults (Singer and Lalich, 1995). Earlier research on religious conversion that included cult adherents (e.g., members of the Unification Church in Galanter et al., 1979; Hare Krishnas in Ullman, 1982, 1988, 1989) may have inadvertently included high numbers of participants whose conversion experiences involved high levels of Dysphoric Need. This may in turn have contributed to findings of high levels of psychopathology among some converts, and thus to the appearance of an empirically based association between religious conversion and psychopathology. Thus, our findings strongly suggest that the link between religious conversion and psychopathology may obtain for only a minority of religious converts.
The critical experiences that loaded on Extrinsic Crisis shared little of the sense of early onset and chronicity characteristic of the crisis items which loaded on Dysphoric Need. We suggest that Dysphoric Need appears to reflect trait-like personality characteristics resulting from unsatisfactory resolutions of the early psychosocial stages in Erikson's (1968) model of the life course (that is, resolutions weighted more on the side of mistrust, shame and doubt, guilt, and inferiority) in a way that Extrinsic Crisis does not. The types of experiences represented by items with high loadings on Extrinsic Crisis are also likely to occur at points spread throughout the life course, or even at its later end (e.g., ARCEQ 20: "I had an experience of illness and/or healing that contributed to my conversion"; ARCEQ 21: "I had experiences with alcohol and/or drugs that contributed to my conversion").
254
Kahn and Greene
It may be that the wide age range of the present sample and the inclusion of items relevant to experiences across the lifespan allowed Extrinsic Crisis to emerge in a way that more limited age ranges and measurement approaches of earlier studies did not. For example, Galanter et al. (1979) reported the mean age of their sample as 24.7 years, and Ullman (1982, 1988, 1989) reported limiting her sample to an age range of 20 to 40 years. By contrast, the present sample ranged in age between 25 and 84 years of age.
Factor 4, Openness to Uncertainty, was an extremely cohesive factor, receiving loadings from all of the eight quest subscale items, a negative loading from an encounter item (ARCEQ 44: "An encounter with someone who shared the gospel with me contributed to my conversion"), and a modest loading from a single consequences item (ARCEQ 86: "I see my conversion as the first step in a long process of transformation"). What the negative loading from the encounter item suggests is a refusal to be limited by dogmatic certainties. This is echoed in the consequences item by a sense of expectation of continued change. Both items suggest a receptivity to change in the face of any perceived demand for orthodoxy.
In the present sample, converts to non-Christian religions reported higher levels of Openness to Uncertainty than converts to conservative/evangelical Protestantism. This finding suggests that one way of viewing Openness to Uncertainty within the context of religious conversion experience, is as one pole along a continuum of the commitment dimension, where the other pole is represented by Zealous Dedication. That is, all of the participants reported that they had experienced religious conversion. Some reported making a Christian commitment involving complete surrender and total dedication. Others made a commitment, equally strong, to a community of seekers without creeds or requirements regarding uniformity of belief or practice (e.g., converts to Unitarianism; S. Meyer, personal communication, October 8, 1999; L. Ffallman, personal communication, November 3, 1999). The former might then be seen as lying at the Zealous Dedication end of the commitment spectrum, while the latter might be seen as lying at the Openness to Uncertainty end of the commitment spectrum.
Ours is quite a different understanding of the quest dimension than Rambo has proposed, and it may reflect in part the different origin of most of the items loading on Openness to Uncertainty. The Quest Scale (Batson and Schoenrade, 1991a, 1991b) was originally developed to address a dimension of religiosity, "religion as quest," that was not assessed by either the Extrinsic or the Intrinsic Religiosity scales (Allport and Ross, 1967). "Religion as quest" was defined as "honestly facing existential questions in their complexity, while at the same time resisting clear-cut, pat answers" (Batson and Schoenrade, 1991b, p. 430). Some of the criticisms of the original Quest Scale (Batson and Ventis, 1982) concerned its validity, including concerns that the Quest Scale was in fact a measure of agnosticism, anti-orthodoxy, "sophomoric religious doubt," or simply religious conflict (Batson and Schoenrade, 1991a).
"Seeing Conversion Whole"
255
Rambo's discussion of the quest dimension highlights the active, seeking aspect of the potential convert, with an emphasis on the quest as precursor to conversion. That is, there is an implicit assumption that the quest dimension will be passed through (or largely so) on the journey to commitment and consequences in the conversion process. This approach to conversion has been institutionalized in the "inquirers' classes" found in most religious bodies, which present information about the particular religion and attempt to answer the questions of attendees in such a way that some may choose to become converts.
It seems clear that the items from the Quest Scale formed a cohesive factor, but it is an open question whether that factor is in fact related to "quest-in-conversion." It may, in fact, more properly reflect "quest-in-religiousness," or even "quest-within-commitment," both of which are somewhat different constructs. The wording of the quest items, carried over from the Batson and Schoenrade Quest Scale, does not refer to conversion specifically and so may have contributed to this ambiguity. By contrast, most of the items developed for the ARCEQ included such markers as "contributed to my conversion," "since my conversion," "before my conversion," "when I decided to convert," or simply "my conversion." Thus, despite the attractiveness of the Quest Scale and its clear reliability as a measure, further research on the specific dimension of "quest-in-conversion" will require a modified version to address such concerns.
By contrast, items loading on Experiential Learning appeared to capture participants' engagement in a community of believers that contributed to their conversion. This appears most closely allied to Rambo's interaction dimension; it is not surprising to find that all eight items from the interaction subscale loaded on this factor, seven contributing uniquely to it. The ninth item loading on this factor came from the consequences subscale (ARCEQ 95: "My conversion gave me a new community of people to serve"), but clearly involves interaction with a community of what are now fellow-believers.
That no group differences were obtained in Experiential Learning suggests that the interaction phase is likely to be an important element in many conversions that proceed relatively slowly and deliberately, whether from the convert's side or because of the stance of the religion itself. The time element in conversion is an important facet which distinguishes denominational understandings of what conversion means and, as a result, shapes the personal conversion experience. That is, certain denominations are oriented more toward a relatively rapid decision or commitment (e.g., Baptists, Pentecostals), whereas others are more patient and emphasize education and reflection before commitment (e.g., Judaism, Catholicism, Orthodoxy). The individual's conversion experience is thus bound to be shaped by the denomination she or he has encountered and is considering.
To summarize, there appear to be close alignments between Redemptive Love and Rambo's consequences dimension, between Dysphoric Need and Extrinsic Crisis and Rambo's crisis dimension, and between Experiential Learning and
256
Kahn and Greene
Rambo's interaction dimension. Zealous Dedication and Openness to Uncertainty are clearly related to the commitment and quest dimensions of Rambo's model, but, as argued above, in complex ways. Zealous Dedication may represent only a portion of the commitment experience and Openness to Uncertainty may reflect an aspect of quest only marginally related to conversion. There is also the possibility that these two factors measure opposite poles of a commitment continuum.
Missing from the present findings was any factor reflecting Rambo's encounter dimension. Of the six items from the original 12-item encounter subscale retained for the factor analysis, all but one loaded on Dysphoric Need. Two items provided moderate loadings on that factor, while the rest contributed only modest or shared loadings. No clearly independent encounter dimension emerged.
Several reasons for this finding can be adduced. First, after dropping the context subscale, the requested number of factors for the solution was reduced from seven to six. With crisis emerging as two clear factors, a relatively weak encounter factor may simply not have been able to emerge, and the item loadings may have been scattered among other factors. Second, as the shortest final subscale, encounter may simply have had too few items to emerge as a factor in its own right. Finally, it is possible that encounter is not a distinct dimension in religious conversion experience, but, as suggested in the discussion above, a part of the larger conversion dynamic involving Dysphoric Need. Further research is needed to determine which explanation is most likely to be correct.
Given our comments at the outset about the traditional predominance of intrapsychic approaches in the study of conversion, it is noteworthy that the context subscale posed the greatest measurement problem. Its poor reliability both before and after item analysis suggests that a multidimensional approach to the measurement of this complex dimension may be required to assess this aspect of Rambo's model. It is especially unfortunate that this discussion of the dimensions of religious conversion experience must remain largely silent on contextual influences, and it thereby becomes particularly important that future research address ways to measure this central and underappreciated dimension. One promising domain for exploration would be the marriage relationship. Gordon (1967) highlighted the importance of this particular microcontext in his account of 45 conversions between Judaism, Protestantism, and Catholicism. For example, he reported that considerations of parental feelings, different levels of religious commitment on the part of the spouses, and concerns for the religious upbringing of children had led a substantial segment of his sample to convert. Rather than attempting a global assessment of the influence of context, as we did, then, future research might well begin by focusing on this domain.
The generalizability of our results will be enhanced by future research that broadens the representation of conversion experience to include denominations and religions underrepresented in our sample (e.g., Latter-Day Saints, Seventh-Day Adventists, Muslims, Buddhists). Further attention also should be given to
"Seeing Conversion Whole"
257
empirical validation of Rambo's (1993) fivefold typology of conversion used in this study; that is, affiliation, intensification, tradition transition, institutional transition, and apostasy. A particular question of interest would be whether there are specific correlates of religious conversion experience associated with or unique to each type. Finally, the emergence of two distinct subdimensions to Rambo's crisis dimension represents a finding of considerable theoretical interest and deserves replication with another sample, perhaps focused entirely on older converts. If replicated, the finding that there is a distinct, age-related Extrinsic Crisis dimension promises to enrich our understanding of conversion experiences at the upper end of the life course.
REFERENCES
Allison, J. (1969). Religious conversion: Regression and progression in an adolescent experience.
Journal for the Scientific Study of Religion 8, 23-38. Allport, G. W., &Ross, J. M. (1967). Personal religious orientation and prejudice. Journal of Personality
and Social Psychology 5, 432-A43. Batson, C. D., & Schoenrade, P. A. (1991a). Measuring religion as quest: 1) Validity concerns. Journal
for the Scientific Study of Religion 30(4), 416^129. Batson, C. D., & Schoenrade, P. A. (1991b). Measuring religion as quest: 2) Reliability concerns.
Journal for the Scientific Study of Religion 30(4), 43CM147. Batson, C. D., & Vends, W. L. (1982). The religious experience. A social-psychological perspective.
New York: Oxford University Press. Bronfenbrenner, U. (1979). The ecology of human development. Cambridge, MA: Harvard University
Press.
Comrey, A. L., & Lee, H. B. (1992). A first course in factor analysis (2nd ed.). Hillsdale, NJ: Erlbaum. Crowne, D. P., & Marlowe, D. (1960). A new scale of social desirability independent of psychopathol-
ogy. Journal of Consulting Psychology 24(4), 349-354. Downton, J. V., Jr. (1980). An evolutionary theory of spiritual conversion and commitment: The case
of Divine Light Mission. Journal for the Scientific Study of Religion 19, 381-386. Erikson, E. H. (1968). Identity: Youth and crisis. New York: Norton.
Fowler, J. W. (1981). Stages of faith: The psychology of human development and the quest for meaning.
San Francisco: Harper and Row. Galanter, M., Babkin, R., Babkin, J., & Deutsch, A. (1979). The "Moonies": A psychological study
of conversion and membership in a contemporary religious sect. American Journal of Psychiatry
136, 165-170.
Gordon, A. I. (1967). The nature of conversion: A study of forty-five men and women who changed
their religion. Boston: Beacon. Hudson, D. W. (1992). The Catholic view of conversion. In H. N. Malony & S. Southard (Eds.),
Handbook of religious conversion (pp. 108-122). Birmingham, AL: Religious Education Press. James, W. (1902/1994). The varieties of religious experience. New York: Modern Library. Kerr, H. T, & Mulder, J. M. (Eds.). (1983). Famous conversions: The Christian experience. Grand
Rapids, Mich.: Eerdmans. Kohlberg, L. (1986). The psychology of moral development. New York: Harper and Row. Kraft, C. H. (1992). Conversion in group settings. In H. N. Malony & S. Southard (Eds.), Handbook
of religious conversion (pp. 259-275). Birmingham, AL: Religious Education Press. Loder, J. E. (1989). The transforming moment (2nd ed). Colorado Springs: Helmers and Howard. Lofland, J., & Skonovd, N. (1981). Conversion motifs. Journal for the Scientific Study of Religion 20,
373-385.
Lofland, J., & Stark, R. (1965). Becoming a World-Saver: A theory of conversion to a deviant perspective. American Sociological Review 30, 862-875.
258
Kahn and Greene
MacCallum, R. C, Widaman, K. R, Zhang, S., & Hong, S. (1999). Sample size in factor analysis.
Psychological Methods 4(1), 84-91. Morgan, D. L. (1997). Focus groups as qualitative research (2nd ed). Thousand Oaks, CA: Sage. Nock, A. D. (1933). Conversion: The old and the new in religion from Alexander the Great to Augustine
of Hippo. Oxford: Clarendon Press. Paloutzian, R. R, Richardson, J. T, & Rambo. L. R. (1999). Religious conversion and personality
change. Journal of Personality 67(6), 1047-1079. Rambo, L. R. (1993). Understanding religious conversion. New Haven, CT: Yale University Press. Singer, M. T, & Lalich, J. (1995). Cults in our midst. San Prancisco: Jossey-Bass. Tabachnick, B. G., & Pidell, L. S. (1996). Using multivariate statistics (3rd ed.). New York:
HarperCollins.
Tippett. A. R. (1977). Conversion as a dynamic process in Christian mission. Missiology 2, 203-221. Ullman, C. (1982). Cognitive and emotional antecedents of religious conversion. Journal of Personality
and Social Psychology 43, 183-192. Ullman, C. (1988). Psychological well-being among converts in traditional and nontraditional religious
groups. Psychiatry 5i(Aug), 312-322. Ullman, C. (1989). The transformed self. The psychology of religious conversion. New York: Plenum
Press.
Zinnbauer, B. J., & Pargament, K. I. (1998). Spiritual conversion: A study of religious change among college students. Journal for the Scientific Study of Religion 37(1), 161-180.