faith, family, and teen dating: examining the effects of ...

3 downloads 0 Views 8MB Size Report
Faith, Family, and Teen Dating. ADOLESCENT ROMANTIC RELATIONSHIPS: CONTOURS AND CORRELATES. Dating is a significant part of adolescence for ...
FAITH, FAMILY, AND TEEN DATING: EXAMINING THE EFFECTS OF PERSONALAND HOUSEHOLD RELIGIOSITY ON ADOLESCENT ROMANTIC RELATIONSHIPS* JOHN P. BARTKOWSKI and XIAOHE XU UNIVERSITY OF TEXAS, SAN ANTONIO KRISTI M. FONDREN UNIVERSITY OF SOUTHER]^ MISSISSIPPI REVIEW OF RELIGIOUS RESERCH VOLUME 52(3), PAGES 248-265

Dating continues to be a key point of transition into American adolescence for many teens. Yet, despite a growing research literature on romantic relationships and dating in adolescence, little attention has been paid to the impact of religion on teen dating. This study examines how various forms ofpersonal religiosity and the family's religious environment influence dating. Using data from the National Study of Youth and Religion, we explore the effects of religion on dating involvement (total number of partners dated), and partner selection (number ofinterfaith versus same-faith dating relationships). Signiflcant religious effects surface for both dating involvement and partner selection, underscoring the need to examine more closely the influence of religion on the lives of American teens.

ating is a key point of transition in the lives of many American adolescents, and has been the subject of social research for many years now. Although much attention has been paid to the antecedents of adolescent romantic relationships, such as youth motivations, peer pressure, and family influences, religion has been virtually ignored. This oversight is all the more remarkable because religion has been shown to be quite influential in the social lives of teens (Smith 2005). The few studies that do treat religion and teen dating are marked by serious data limitations that include the use of convenience samples, retrospective reports, and crude one-dimensional measures of religion. Consequently, these studies have rendered insights of limited generalizability and questionable validity (Zimmer-Gembeck, Sienbenbruner, and Collins 2001). This study aims to extend research on the social sources of adolescent romance and teen dating by exploring the influence of religion on this important cultural practice. We use nationally representative data from the National Study of Youth and Religion to undertake this investigation and employ multiple measures of religion in doing so. Because these data are drawn directly from a recent nationally representative, random sample of teens 13-17 years of age, we can have considerably more confidence in the representativeness of the responses to questions about religiosity and dating.

D

248

Faith, Family, and Teen Dating

ADOLESCENT ROMANTIC RELATIONSHIPS: CONTOURS AND CORRELATES Dating is a significant part of adolescence for many young people today (Adams, Laursen, and Wilder 2001; Shulman and Kipnis 2001). Rather than following the transition of earlier eras from courtship to marriage to parenthood, many young people in the U.S. and other Westem cultures now experience a succession of premarital romantic relationships (Bailey 1989; Collins 2003; Waller 1937). Dating relationships today are complex (Glenn and Marquidt 2001; Shulman and Collins 1997; Shulman and Kipnis 2001; Zimmer-Gembeck et al. 2001 ), and thus can be ditficult to define (Shulman and Kipnis 2001 ; Zimmer-Gembeck et al. 2001 ). Nevertheless, scholars generally recognize that dating relationships are marked by emotional attachment and physical attraction (sometimes coupled with sexual expression), and are designed to address desires for companionship, self-discovery, and anticipatory socialization (e.g., Collins 2003; Shulman and Kipnis 2001). Recent studies distinguish between various facets of dating, including involvement (frequency) and partner selection (homogamy/heterogamy), key foci that guide our study (Collins 2003). Adolescent romantic relationships are affected by various structural and cultural forces (Whyte 1990). Young people are likely to date those who are similar to them in terms of social background, experiences, and interests. Youngsters who come from a home with a happy marriage are less likely to date a wide variety of partners, but those whose parents attempt to intervene in their dating relationships tend to date more partners. In addition, early relationships (caregiver-child relationships, peer relationships) play a significant role in the development and quality of adolescent romantic relationships (Adams et ai 2001; Collins Henninghausen, Schmit, and Sroufe 1997; Conger. Cui, and Bryant 2000; Shulman and Collins 1997; Shulman and Kipnis 2001 ). Teens whose parents are nurturing and involved tend to have positive interpersonal interactions with their romantic partners (Conger et al. 2000; see Shulman and Collins 1997). Adolescent romantic relationships are also influenced by peer relationships (Furman and Shaffer 1999; Gonzaga et al. 2001 ; Sippola 1999). Teens who have multiple opposite-sex friends are more likely to have a greater number of romantic relationships, date more frequently (Connolly and Johnson 1996), and have longer lasting relationships than teens with fewer friends (Feiring 1999). However, peer influence decreases as adolescents become more committed to romantic relationships (Adams et al. 2001 ; Collins 2003; Shulman and Collins 1997). Whereas companionship appears to be most important in early adolescent romantic relationships, care and commitment become important as teens age (Shulman and Kipnis 2001). Older teens in romantic relationships also report experiencing more interdependence, daily interaction, participation in diverse activities, and reciprocity in their relationships than do younger teens (Adams et al. 2001).

RELIGION AND DATING: A MISSING PIECE OF THE PUZZLE Despite the attention paid to adolescent romantic relationships, the influence of religion on this facet of young people's social lives has been largely overlooked. There are several reasons why it is important to consider the possible infiuence of religion on teen dating and romantic relationships. First, scholars have established that religion is an important predictor of various behavioral outcomes in young people's social development (Smith 2005). Given that religion affects the self-perceptions, peer networks, and social behavior of youth, 249

I

Review of Religious Research

it seems reasonable to surmise that another key facet of adolescence—namely, dating— may be subject to religious variations. Second, a great deal of research has revealed the connections between religion and family life (e.g., Bartkowski et al. 2008; Becker 2005). Dating and romantic relationships during adolescence are subject to family influences (as noted above), and can also be seen as a form of anticipatory socialization for future family formation. Thus, given the close and formidable linkages between religion and family life (e.g., parenting standards of religious caregivers, family aspirations of religious youth), distinctive patterns of dating among youth raised in religious homes deserves exploration. Third, dating has been the subject of vigorous discussion in some religious quarters. Authors of conservative Protestant advice manuals on dating have argued for a revision of conventional dating habits in favor of approaches (church-sponsored events, parental approval) that ostensibly foster more "wholesome" interactions (e.g., Harris 1999, 2000). And, while Latter-day Saints (Mormons) have long proscribed dating before the age of sixteen, they recognize that dating can serve as a valuable form of anticipatory socialization for marriage (Barlow 1977; Ingoldsby 1992; Markstrom-Adams 1989; 1991). Thus, both conservative Protestants and Latter-day Saints are not opposed to dating, but encourage it under the "right conditions." Such teens are told to date many partners to avoid getting too serious with any one person. These distinctive orientations toward dating are broadly consistent with the subcultural character of conservative faith traditions (Smith 1998). Subcultural theories of religion have revealed that conservative faith traditions express considerable suspicion of mainstream cultural values and instead reshape mainstream sensibilities to comport with conservative religious values that are rooted in theological orthodoxy (e.g., biblical inerrancy), objectivist views of truth (e.g., religious exclusivity), and familism (e.g., early marriage, valuation of family relations over career pursuits) (Smith 1998; Xu, Hudspeth, and Bartkowski 2005). Turning to the existing research on religion and dating, the flrst empirical study to examine this issue explored variations in attitudes toward interfaith dating and marital relationships (Burchinal 1960). Using data drawn from high school and college students in rural Iowa, the investigator hypothesized that negative attitudes toward interfaith dating and marriage would be more pronounced among (1) the highly religious (measured by church attendance), (2) respondents who occupied a high social class position (measured by father's occupational prestige), (3) female respondents, and (4) college students (as opposed to their younger counterparts in high school). The findings revealed only partial support (among non-farm college males) for an inverse relationship between religiosity and a negative disposition toward interfaith dating and marriage. Hypotheses concerning variations in opposition to interfaith dating and marriage by social class, gender, and age were generally supported. Given longstanding Latter-day Saint (LDS, Mormon) resistance to interfaith marriage (Barlow 1977), Mormons have received some attention in the small literature on religion and dating (Ingoldsby 1992; Markstrom-Adams 1989; 1991). However, such studies have been conducted with small, nonrandom samples and often were undertaken in predominantly Mormon areas of the country. Research on this score reveals Mormon youth's opposition to interfaith dating when compared with youth from other faith traditions. LDS youth's opposition to interfaith dating is at once rooted in theological mandates (date and marry 250

Faith, Family, and Teen Dating

\

only within the faith), the cultural distinctiveness of this particular tradition (modem-day revelation, extra-biblical scripture, the postponement of dating until age sixteen, proscriptions against premarital sexual activity), and the wide availability of Latter-day Saint dating partners in predominantly Mormon communities (Utah and the Mountain West). Mormon dating pattems might differ in areas of the country where this religious tradition is in the minority (see Bamett 1962). . . , i

I

STUDY FOCI AND KEY HYPOTHESES

This study is designed to augment the existing literature on adolescent romantic relationships by examining religion as a cultural antecedent of dating pattems. It also aims to extend the burgeoning body of research on religion and youth. Although this scholarship has underscored the profound influence of religion in the lives of American teens (e.g.. Smith 2005; Smith and Dentón 2003), the place of dating among youth has not been investigated in such research. Given the foregoing literature review and the subcultural character of conservative religious groups (conservative Protestants, Lattet-day Saints), out study addresses the following hypotheses conceming the overall number of partners dated (designated as "a" hypotheses) and the number ofinterfaith dates reported (designated as "b" hypotheses). Justifications are featured after each hypothesis is introduced. H1 a : Conservative Protestant and Latter-day Saint teens will date more partners on average than their peers in other denominations. Dating continues to function as a gateway to marriage. Previous research on religious differences in marriage timing reveals that Latter-day Saints and conservative Protestants marry at a considerably younger age than their peers in other religious denominations and unchurched youth (Xu et al. 2005). Because the entry of Latter-day Saints and conservative Protestants into first marriage is accelerated, it is anticipated that youngsters from these denominations will have dated more partners than their mainline peers between ages 13 and 17. Additionally, conservative Protestant (Schwadel and Smith 2005) and LDS teens (Smith 2005) attend worship services and participate in youth group activities at a higher rate than teens affiliated with other denominations. Previous research demonstrates that teens embedded in expansive social networks have more dating partners and date more frequently than those with fewer friends (Connolly and Johnson 1996; Feiring 1999). Thus, it is anticipated that conservative Protestant youth will date most frequently of all religious groups, followed by Latter-day Saint teens. H1 b: Conservative Protestant, Latter-day Saint, and Catholic teem will date fewer interfaith partners on average than their peers in other denominations. Conservative Protestants and Latter-day Saints are more sectarian faith traditions. Each of these traditions places a greater value on maintaining cultural distinctiveness from secular American social life (Smith 1998; Mauss 1994), and both stress the importance of marrying within the faith—as does the Roman Catholic Chutch (Sherkat 2004). Hence, it is expected that teens affiliated with these traditions will be less likely to date outside their faith. H2a/b: Teens from homes with a more robust family religious environment will be less likely to (a) date frequently and (b) date interfaith partners across denominations. Dating remains a rite of passage to adulthood for most American teens. We hypothesize that parents in homes with a robust family religious environment exercise more control 251

Review of Religious Research over their children's activities and greater infiuence on their children's lifestyle choices, in much the same way that parental monitoring reinforces parental authority. Consequently, we expect both dating frequency and interfaith dating will be curtailed in homes with a highly religious environment. H3a^: Teens embracing an exclusive view of religious truth will be least likely to (a) date frequently and (b) date interfaith partners across denominations. Dating can be considered a marker for a youngster's entry into the mainstream of secular American culture. Inasmuch as the belief that one's own religion is the true faith (religious exclusivity) creates a strong cultural bulwark against secular American social life and could severely limit one's access to a pool of potential dating partners, we expect the lowest likelihoods of overall dating and interfaith dating to be exhibited by those who adopt an exclusive orientation toward religious truth. The highest levels of overall dating and interfaith dating will be exhibited by teens who are skeptical of religious truth altogether, because faith will not be a selection criterion such teens use in evaluating potential dating partners. Teens with an inclusive view of religious truth (many religions are true) will be situated between these two poles. H4a/b: Teens who attend religious services and participate in a religious youth I group (a) will be more likely to date overall but (b) less likely to date interfaith partners than their nonattending peers across denominations. As previously noted, youth involved in congregational activities are more socially integrated and thus will be more likely to date partners frequently. However, these same youth will be more strongly invested in their own tradition, and thus unlikely to date outside the faith. H5a^: Religious salience, experience, commitment, and devotion will all have (a) no effect on overall dating involvement, and (b) a depressive effect interfaith dating across denominations. Subjective dimensions of religiosity (that is, religious salience, experience, commitment, and devotion) indicate an intemalization of religious norms and a privileging of faith in a teen's life. More robust forms of subjective religiosity are not expected to influence dating in general, but are likely to serve as a counterbalance against dating outside of one's faith. This latter expectation is congruent with pattems of dating homogamy. Therefore, higher levels of subjective religiosity will not affect the frequency of dating, but will deter interfaith dating.

METHOD DATA SOURCE The data we utilize are drawn from the National Study of Youth and Religion (NSYR), a nationally representative telephone survey of 3,290 U. S. English and Spanish speaking teenagers between the ages of 13 and 17, and of their parents. The NSYR also included 80 over-sampled Jewish households using the listed surnames sampling method, bringing the total number of completed NSYR cases to 3,370. The NSYR was conducted from July, 2002 to April, 2003 by researchers at the University of North Carolina at Chapel Hill using a random-digit-dial (RDD) telephone survey method, employing a sample of randomly generated telephone numbers representative of all household telephones in the fifty United States (Smith and Dentón 2003). 252

Faith, Family, and Teen Dating Unlike other data sets collected from American youth, the NSYR includes a wide range of measures of religion, ranging from denominational affiliation, worship service attendance, and youth group attendance to the salience of religious beliefs and devotional activities such as frequency of prayer. In addition, the NSYR also incorporates a large number of outcome measures for the youth respondents. This study is restricted to two sub-samples of the NSYR based on the two dependent variables, key independent variables, and statistical controls. The restricted sample sizes are 2,932 and 1,897, respectively, for the total number of dates and the number ofinterfaith dates since the age of 13. As advised by the NSYR research team, our study uses the simple selection probability weight (weight 1) to be nationally representative, and the Jewish oversample (a non-probability sample) dummy variable to remove potential sampling bias (Smith and Dentón 2003).

MEASURES Dependent variables. Two dependent variables are used in this study to capture pertinent teen dating patterns in contemporary America, namely dating involvement (frequency) and partner selection (interfaith dating). The flrst dependent variable is measured by a survey item tapping the number of people dated in which respondents were asked: "How many total different people, if any, have you dated since you turned 13 years old?" The second dependent variable, the number ofinterfaith dates, explores the role of faith, not simply denomination, in fostering selectivity in dating partnerships by asking: "How many, if any, of these people you dated held religious beliefs that were different from yours?" Both variables range from 0 to 100 with 100 = 100 or more. Because of severe skewness in these variables, they are top-coded ranging from 0 to 20 with 20 = 20 or above. Key independent variables. The National Study of Youth and Religion contains a wide range of religion measures, thereby permitting the examination of religion as a complex, multidimensional construct. As constructed by the NSYR team (Smith and Dentón 2003), flve groups of religion variables in the parent and teen surveys are utilized. The key variable used is the religion or congregation where the respondent attends religious services most often. Consistent with standard coding schemes (Steensland et al. 2000), respondents' denominational affiliation is recoded into a new variable with 1 = Conservative Protestant, 2 = mainline Protestant, 3 = Black Protestant, 4 = Catholic, 5 = Jewish, 6 = Latter-day Saint, 7 = other religion, and 8 = none (no religious affiliation). This new variable is then dummy-coded with the last category (no religious affiliation) as the reference category. While adolescent dating or romantic involvement is by and large a part of youth centered culture, it is by no means free of familial and parental influence. In fact, youth dating behaviors are often subject to parental guidance and approval (Bailey 1989; Conger et al. 2000; Whyte 1990). For such reasons, we also use an item tapping ihe family religious environment. The responses to how frequently the family talks about religious topics or spiritual matters together are reverse-coded so lower scores indicate a less religious family environment and higher scores indicate a more religious family environment. The response categories range from I = the family never talks about God, scripture, prayer, or other religious or spiritual things together to 7 = the family talks about God, scripture, prayer, or other religious or .spiritual things together every day. Religious participation is measured by two different survey items. Teens' religious service attendance at the respondents' congregation (the one they flrst identifled) ranges from 253

Review of Religious Research 0 for "«ever" to 6 for "wore than once a week," and is treated as a continuous variable. Another important participation measure is drawn from a survey question that asked respondents: "Are you currently involved in any religious youth group?" Response categories for religious youth group participation are 0 = no and 1 =yes. Given prior research on religious subcultural variations in marriage/family values, beliefs, and norms (Gay, Ellison, and Powers 1996; Xu et al. 2005), we choose a number of measures that reflects the respondent's commitment to their religious beliefs. First, levels of commitment to religious exclusivity, the idea that there is one tme religion, are captured by a survey question that asked respondents, "Which of the following statements comes closest to your own views about religion: ( 1 ) only one religion is true, or (2) many religions may be true, or (3) there is little tmth in any religion?" Responses are dummy-coded to include the following orientations: exclusive (one true religion),/^/wrafcr (many true religions), and skeptical (little tmth in religion), with the last response category as the reference. Next, religious salience (a = .88) is created as an index variable from two questions designed to gauge the influence of religious beliefs in the lives of respondents. Respondents were asked, "How important or unimportant is religious faith in shaping (a) how you live your daily life and (b) your major life decisions?" The response categories are coded so that 1 = not important at all to 5 = extremely important. Respondents were also asked about religious experiences: "Have you ever, or not, (a) had an experience of spiritual worship that was very moving and powerflil; (b) experienced a definite answer to prayer or specific guidance from God; and (c) witnessed or experienced what you believe was a miracle from God. We combined these items to create a religious experience index (a = .63). (We recognize that these items could be conceptualized separately, as such experiences need not co-occur. However, our alpha measure meets an acceptable threshold of intemal reliability.) Respondents were also asked if they had made a personal religious commitment to live their life for God, with responses dichotomized as 0 = no and 1 = yes. Finally, a religious devotion measure (a = .67) is derived from two questionnaire items. Respondents were asked, "How often, if ever, do you pray by yourself alone?" and "How often, if ever, do you read from scripture to yourself alone?" Responses range from 1 for "«ever" to 7 for "many times a day." These two items are averaged, resulting in an index variable with higher scores indicating more frequent private devotional activities. Control variables. Guided by previous research on teen dating/romantic involvement, a number of statistical controls are included in our multivariate statistical analyses. Respondents 'age (continuous variable), race-ethnicity (dummy-coded with White as reference)', and gender (with male as the reference) are included in this study. Family characteristics such us parent s marital status used as a proxy of family structure (dummy-coded with 1 = married parents), family income (required by the survey design which is treated as a continuous variable), and parental supervision (an index variable [a = .60] with higher scores indicating higher levels of parental supervision in terms of monitoring teens' music, television, movie viewing, peer relationships, and whereabouts) are controlled as well. (Here again, we recognize that parental supervision may be more vigilant in some domains than others, but our alpha score is acceptable.) Finally, school type (dummy-coded as 1 = public school) and census region (required by the survey design which is dummy-coded with the South as the reference) is included in statistical modeling.

254

Faith, Family, and Teen Dating

STATISTICAL MODELING The two dependent variables, dating involvement (the number of people dated since age 13) and partner selection (the number of interfaith dates since age 13), are highly skewed count-like variables, which roughly measure the number of occurrences of a particular event. Whereas Poisson regression is typically used to analyze count-like variables, the over-dispersion of our variables {i.e. the standard deviations are larger than the means) lead us to use negative binomial regression (Long and Freese 2001 ). (Zero-inflated count models would also be appropriate, but are not as commonly utilized or readily understood as negative binomial regression.) Unlike its logit counterpart where the log odds or odds of certain events occurring are estimated, negative binomial regression estimates the expected number of certain events. For intuitive interpretations, the negative binomial regression coefficients are often converted into percentage change(s), namely, { exp (B) - 1} %, such that percentage increase or decrease in the expected mean number of people dated or interfaith dates since age 13 can be stated. As suggested by the NSYR research team (Smith and Dentón 2003), this study used the simple selection probability weight and included the Jewish oversample dummy variable 10 conduct a series of negative binomial regression analyses, which allowed us to analyze nationally representative samples and at the same time to remove potential bias due to the oversampled non-probability Jewish cases. Unless stated otherwise, the complex survey procedure in Stata was utilized to weight and analyze the data.

RESULTS Descriptive statistics for all variables used in this study are presented in Table 1. On average, teens reported dating four partners. To be sure, the average numbers of dates reported by respondents is, in part, a function of the age range of the sample (13-17 years old), .lust under half of all dates ( 1.96), on average, occurred with a partner of another faith. The mean age of all respondents is 15.04 years old. Among the descriptive statistics for religion variables, the vast majority of youth claim a denominational affiliation, and most believe that many religions are true (pluralism), followed respectively by those with exclusive and skeptical orientations toward religious truth. Rates of worship service attendance are fairly robust, and over one third of respondents participate in a religious youth group. Religion is quite salient for youth in the sample, and a sizable proportion report having had a religious experience. Over 55% have personally committed to live their lives for God, and devotional activities (prayer and scripture study) are fairly frequent in this nationally representative sample. Table 2 reports negative binomial regression results for dating involvement, that is, the number of dates reported by teen respondents. Model 1 of Table 2 displays the effects of covariates on the dependent variable in the absence of key independent variables of interest (namely, religion measures). On average, older teens are signiflcantly more likely to have more dates, as are youth in public schools. Lower expected frequencies of dating are observed among teen girls, teens living in homes with a higher family income, youth supervised more closely by their parents, and teens residing in regions other than the South. Control variables remain in all subsequent models to examine the net effects of religion variables on dating involvement among American teens.

255

Review of Religious Research Table 1. Listwise Descriptive Statistics by Dependent Variables Number of People Interfaith ] Dated Dating n % n % Number of people dated 4.00" 4.66" 5.62' 4.69'' Number of interfaith dates 1.96' 3.47'' Respondent's age 15.04' 1.40*" 15.37" 1.31" Respondent's race White 1,954 66.7 1,284 67.6 Black 489 16.7 326 17.2 Hispanic 333 11.3 193 10.2 Other 156 5.3 94 5.0 Respondent's gender Male 1,468 50.1 963 50.8 Female 1,464 49.9 934 49.2 Parent's marital status Married 1,981 67.6 1,276 67.3 Other 951 32.4 621 32.7 Family income 5.94" 2.92'' 5.92' 2.92" Parental supervision (index) 3.59' 0.89" 3.50" 0.88" School type Public school 2,538 86.6 1,664 87.7 Other 394 13.4 233 12.3 Census region Northeast 420 14.3 259 13.7 Midwest 654 22.3 426 22.5 West 588 20.1 351 18.5 South 1,270 43.3 861 45.3 Oversamp dummy variable Oversampled Jewish youth 64 2.2 33 1.7 Other 2,868 97.8 1,864 98.3 Denominational affiliation Conservative Protestant 940 32.1 638 33.6 Mainline Protestant 306 10.4 198 10.4 Black Protestant 341 11.6 228 12 Catholic 712 24.3 428 22.6 Jewish 91 3.1 50 2.6 Latter-day Saints 62 2.1 39 2.1 Other religion 138 4.7 94 5.0 None 342 11.7 222 11.7 Family religious environment 3.3r 1.72" 3.24" 1.69"

256

Faith, Family, and Teen Dating Table 1 (continued) Views of religious truth Exclusive (one true religion) Pluralist (many true religions) Skeptical (little truth in religion) Religious service attendance Religious youth group participation Yes No Religious salience (index) Religious experience (index) Personal religious commitment Yes No Religious devotion (index) .V M. h SD.

1,825 263 3.13'

28.8 62.2 9.0 2.19''

518 1,201 178 3.12"

27.3 63.3 9.4 2.17''

1,117 1,815 3.43" 0.50"

38.1 61.9 1.08'' 0.38''

738 1,158 3.42" 0.52"

38.9 61.1 1.07'' 0.38"

1,626 1,306 3.45' 2,932

55.5 44.5 1.62''

1,072 825 3.46" 1,897

56.5 43.5 1.6"

OAA

Model 2 in Table 2 features the net effects of denominational variations on teen dating involvement. While none of the regression coefficients is statistically significant in Model 2, the expected number of dates is significantly greater for conservative Protestant and Latter-day Saint youth than for unaffiliated teens (the reference category) in Model 10 after all independent variables are entered into the regression model. This finding confirms Hypothesis la, which anticipated higher likelihoods of dating involvement among these religious groups given that their members marry at younger ages and may be more strongly integrated into teen religious networks. In Model 3, we find that the family religious environment, by itself, has a significant effect on dating involvement. This result confirms Hypothesis 2a, in which a depressive eifect on dating involvement was anticipated. This variable remains statistically significant in the hypothesized direction in the full model, as discussed below. Hypothesis 3a, empirically tested in Model 4, stated that teens who believe in one true religion (religious exclusivity) would be less likely to date than those not holding this view, especially those who believe there is little truth in any religion (skeptical orientation toward religion). The religious pluralism variable (many religions may be true) is negatively and statistically predictive of dating involvement in Model 4 but borders on significance in Model 10. Hypothesis 3a is confirmed, such that teens who embrace religious exclusivity are less likely to date. Hypothesis 4a anticipated that teens who attend religious services more often and who belong to a religious youth group would be more inclined to date than their nonattending peers. Hypothesis 4a is not supported. As Model 5 reveals, worship service attendance and youth group participation have no significant effect on dating involvement. Hypothesis 5a anticipated that subjective religiosity (religious salience, experience, commitment, and devotion) would have no effect on the number of partners dated, but would influence interfaith dating. This hypothesis is generally supported with respect to the 257

Review of Religious Research

o

•g

d d o

o

d

o

o

d

o

o' o'

(NfNOOOO^

o

d

ó

ó

o



d

f^Wl

a

o

q

o

o © o

sss

q

o d d

o oo •8

o

d

d o o

Ss

o d d

d —

d d o

o

d o d

d o o

d

d o d

o

8pq

i

I

d

â ö ö

d

g

S

o d d

d

"^ ö ö

o d d

d

ö à ö

d

S3

o

— —

d

S S —

rs

3

dod

d

(^od

d

do'o

o

d o d

o

o d d

d

d o ' d

d

^

;*^(N

n

u-i(*iw-i

00

oa>r-

q q — ; d o d