Discordance in couples’ reporting of courtship stages: Implications for measurement and marital quality

Social Science Research 42 (2013) 1143–1155

Contents lists available at SciVerse ScienceDirect

Social Science Research journal homepage: www.elsevier.com/locate/ssresearch

Discordance in couples’ reporting of courtship stages: Implications for measurement and marital quality Sarah Halpern-Meekin a,⇑, Laura Tach b a b

Department of Sociology, University of Wisconsin – Milwaukee, PO Box 413, Milwaukee, WI 53201, USA Department of Policy Analysis & Management, Cornell University, 247 Martha van Rensselaer Hall, Ithaca, NY 14853, USA

a r t i c l e

i n f o

Article history: Received 13 June 2012 Revised 25 January 2013 Accepted 28 January 2013 Available online 15 February 2013 Keywords: Cohabitation Dating Marital quality Marriage Stayovers

a b s t r a c t The symbolism of rituals creates a shared understanding of events among group members. In the context of romantic relationships, a shared understanding of relationship status transitions may be associated with greater commitment and higher quality relationships. We argue that couples with differing retrospective accounts of their premarital courtship may not have had clear discussions or rituals marking relationship turning points. We test the association between discordance in couples’ reports of premarital courtship stages and marital quality using data from married couples in a national online survey (n = 1504). We find that couple discordance is common, particularly among former premarital cohabitors and for the less institutionalized relationship stages of dating and stayovers, and is associated with lower marital quality. Implications for relationship measurement and the meaning of couple discordance are discussed. Ó 2013 Elsevier Inc. All rights reserved.

1. Introduction Sociology has long recognized the symbolic power of rituals for communicating shared meaning in social interactions (for a review, see Gusfield and Michalowicz, 1984) even as they are a way of sharing of true emotion (Grimes, 2000; Hochschild, 1998; Schweingruber et al., 2004). In the romantic relationship context specifically, relationship rituals such as engagement may play a protective role in relationship development because of the clarity, mutuality, and publicness of the information carried about commitment (Stanley et al., 2010b). Couples who ‘‘slide’’ into more serious union stages, rather than ‘‘deciding’’ to do so, may be at greater risk for poor relationship outcomes because of how such courtship patterns affect the development of commitment (Rhoades et al., 2009; Stanley et al., 2010a, 2006; Vennum and Fincham, 2011). If clear decisions and signals about commitment provide some protection as relationships progress toward marriage, discordance in partners’ retrospective reports of the progression of their relationship may be a potential indicator that a couple ‘‘slid’’ into a more serious relationship rather than having ‘‘decided’’ to do so. Turning points in the relationship that were more clearly demarcated should be more salient, and therefore more memorable. Differing reports of a relationship’s history may be a consequence of couples progressing without clear signals in the relationship, such as particular conversations about what was happening in the relationship and what it meant; this may have implications for the quality of their relationship. We examine whether discordance in couples’ retrospective reports of whether and when various relationship stages took place is associated with current marital quality. The present study builds on previous research by (1) investigating the frequency of discordance in couples’ retrospective reporting of premarital relationship stages, (2) examining premarital relationship stages beyond pre-engagement

⇑ Corresponding author. Fax: +1 414 229 4266. E-mail addresses: [email protected] (S. Halpern-Meekin), [email protected] (L. Tach). 0049-089X/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ssresearch.2013.01.009

1144

S. Halpern-Meekin, L. Tach / Social Science Research 42 (2013) 1143–1155

cohabitation as potentially meaningful areas of discordance, and (3) considering whether such intra-couple discordance is a meaningful risk factor for poorer marital quality. Given that those who cohabit before marriage have experienced an additional relationship stage, compared to those who enter marriage directly, and given indications in the existing literature that premarital cohabitors are more likely than those who enter marriage directly to have slid into a more serious relationship without proactively deciding to do so (as one can more easily slide into cohabitation than marriage), we pay particular attention to potential differences between premarital cohabitors and noncohabitors. We use original survey data collected in a special topics module of the Knowledge Networks online research panel. Our data are representative of currently married couples and include responses from both members of the couple about the timing of their premarital courtship stages. 2. Background 2.1. ‘‘Sliding’’ versus ‘‘Deciding’’ and relationship quality Spouses who cohabited without being engaged report lower dedication and marital satisfaction as well as a greater likelihood of divorce than those who entered marriage directly or who only cohabited after engagement (Kline et al., 2004; Rhoades et al., 2009). In a similar vein, the quality of cohabiting relationships differs based on whether or not there is a commitment to marry (Brown and Booth, 1996). This follows from the idea that the symbolism of rituals fosters a shared understanding of events and adjustment to new social roles (Young, 1965); specifically, the ritual of engagement prior to cohabitation strengthens relationship quality and reduces marital dissolution by increasing the likelihood that the couple agreed about the meaning and direction of the relationship. Stanley et al. (2006) argued that part of the risk of pre-engagement cohabitation was that couples may have ‘‘slid’’ into an increasingly serious relationship without consciously making the decision to do so. The authors suggested that some cohabiting couples may end up following a path into marriage that is not based primarily on the quality of their relationships, but rather on the inertia that develops when they already share a home, possessions, and sometimes children. Getting engaged or explicitly deciding on future marriage plans prior to cohabitation, therefore, can be a marker that ‘‘deciding’’ rather than ‘‘sliding’’ led a couple down the aisle. Further, such clear and public markers should decrease the likelihood that partners misperceive one another’s commitment levels (Stanley et al., 2010b); that is, engaging in the ritual of engagement – the proposal, the ring, the engagement party – ensures the couple has a shared understanding of one another’s intentions for the relationship. This is supported empirically, with couples who cohabit prior to engagement more likely to show asymmetrical levels of dedication in the relationship; these differences do not abate after the transition into marriage (Rhoades et al., 2006). Other relationship transitions—beyond engagement—may be more or less ritualized for couples; for example, starting to spend the night together on a regular basis may have a different meaning if having ‘‘the talk’’ accompanies partners’ exchanging house keys or leaving a toothbrush in the bathroom. If a girlfriend gives her boyfriend a house key, saying it is easier than coordinating their schedules, the meaning and salience of this step in their relationship is far different than if she explains she is giving him the house key because she trusts him and imagines they will live together some day. In the first instance, the event of starting to spend the night together on a regular basis is not marked in any particular way that would make the couple likely to remember it; in the second instance, the expressed symbolism of the event raises its salience, more likely demarcating it as a memorable event (and therefore one on which they could give similar retrospective reports). We argue that discordance in couples’ retrospective reporting of relationship stages could be a potential marker for ‘‘sliding.’’ Couples’ discordance about the progression and timing of their relationship stages may signal the lack of an explicit decision-making process or weaker commitment (Manning and Smock, 2005; Rhoades et al., 2006, 2009). A range of behaviors, including both more institutionalized rituals (like engagement) and less institutionalized rituals (like having ‘‘the talk’’ or changing your Facebook status), may create a more conscious decision to advance the relationship and increase the memorability of the event for both partners. Therefore, as a potential indicator of sliding, couple discordance about the occurrence and start date of various relationship stages may be associated with lower relationship quality. 2.2. Couple discordance on relationship stages Previous research has examined the quality of current versus retrospective reports of marriage and cohabitation start dates, and the level of agreement between partners in their reporting of these dates (Hayford and Morgan, 2008; Lillard and Waite, 1989; Peters, 1988; Teitler and Reichman, 2001; Thomson and Colella, 1992). Researchers examining the quality of reports of marriage start dates generally find high levels of concordance between partners and for the same partner across multiple survey waves. For example, Lillard and Waite (1989) found that 83% of husbands and wives reported the same marriage year in the Panel Study of Income Dynamics (PSID), and Peters (1988) found that only 4% of women in the 1968 National Longitudinal Survey (NLS) reported marriage dates that differed by more than a year across survey waves. Discordance over the occurrence and date of marital disruption was more common. Discordance also increased as the length of time between the event and the survey increased, and couples with simpler relationship histories (such as having only one marriage versus multiple marriages) were also more likely to agree (Lillard and Waite, 1989; Peters, 1988).

S. Halpern-Meekin, L. Tach / Social Science Research 42 (2013) 1143–1155

1145

There is also discordance between partners in reporting on cohabitation. Teitler and Reichman (2001) found that 11% of the parents of young children differed in their reports of whether they were currently cohabiting at the time of the interview, and Thomson and Colella (1992) found that 12% of currently married couples gave inconsistent reports of whether they cohabited before marrying. Only half of currently cohabiting couples in the National Survey of Families and Households (NSFH) reported the same month and year of starting cohabitation, and partners in just 70% of couples reported a start date within 3 months of each other (Manning and Smock, 2005). Individual reports of cohabitation experiences also vary over time (Hayford and Morgan, 2008; Teitler et al., 2006). Researchers have argued that greater discordance in reporting on cohabiting unions, compared to marriages, reflects the fact that cohabitation is less institutionalized (Manning and Smock, 2005; Nock, 1995). Drawing on qualitative data from 115 couples with recent cohabitation experience, Manning and Smock (2005) found that the boundaries between cohabitation and dating were blurry and the process of moving in together unfolded over weeks or even months. Cohabitation did not require a formalized agreement or ceremony, raising the possibility that it was a less conscious decision; that is, cohabitation was less likely to involve the symbolism of rituals that can demarcate a transition in relationship status. Manning and Smock characterized this as a ‘‘slide’’ into cohabitation. Reports also revealed a considerable amount of disagreement between cohabiting partners over expectations for future marriage (Waller and McLanahan, 2005). 2.3. Correlates of couple discordance and marital quality In order to better isolate the relationship between couple discordance and marital quality, we control for an array of personal and household characteristics; personal characteristics include age, race, educational attainment, and employment status, and household characteristics include household income, the presence of children under eighteen in the household, and marital duration. Age serves as a proxy for cohort-based cultural experiences, like changing family behavior norms (Casper and Cohen, 2000). Race is also associated with variation in marital quality and stability, with particularly notable differences between African Americans and Whites (Bulanda and Brown, 2007). Because economic stress and education are tied to marital quality and stability (Teachman and Polonko, 1990; White and Rogers, 2000), we account for educational attainment, household income, and employment status. We also control for the presence of children in the household as parenthood is related to marital satisfaction and divorce risk (Twenge et al., 2003; Waite and Lillard, 1991). Marital duration is associated with declining marital quality (Tach and Halpern-Meekin, 2012; Van Laningham et al., 2001) and also measures how long ago respondents experienced their premarital relationship stages, as we expect recall bias to increase over time. This robust set of control variables gives us greater leverage to isolate the association between couple discordance and marital quality. 2.4. Contributions of the present study We build on previous research by examining the consistency of couples’ retrospective reports of the relationship stages in their courtship process. The events we consider include the start of dating and stayovers, as well as cohabitation and marriage. Of these, stayovers is the least standard measure in studies of premarital courtship and is likely the least institutionalized as well; however, research on a sample of young adults indicates that stayovers—spending the night on a regular basis before moving in together—is the typical pathway into cohabitation (Pollard and Harris, 2007) and thus is worthy of consideration as a relationship stage. We also include couples’ reports of any instances of relationship disruptions (being on-again/off-again) during the respective relationship stages, as research has revealed this to be a fairly common experience (Halpern-Meekin et al., 2013). With novel data from a national sample of married couples, we use reports from both members of a couple of the occurrence and start date of each relationship stage to understand how each partner saw the relationship unfold and how frequently partners disagreed in their retrospective reporting of these relationship events. The present study, therefore, contributes to our understanding of the accuracy of estimates of relationship stages and durations when we rely only on reports from only one member of a couple. While previous studies have focused only on pre-engagement cohabitation as a marker of the ‘‘sliding’’ that is associated with greater marital risk, we argue that discordance in couples’ retrospective accounts of other premarital relationship stages is also a potential indicator of ‘‘sliding’’ because a stronger, decisional process would make the timing and nature of transitions more salient and memorable. We test whether discordance in couples’ reporting on premarital courtship stages is more than a measurement problem and can be leveraged as a predictor of marital quality. While our data are for currently married couples and exclude relationships that have dissolved, this bias should minimize marital quality differences between couples with concordant versus discordant retrospective reports, thereby providing a conservative test of our hypothesis. We draw several hypotheses from previous literature about the extent and consequences of couple discordance over past relationship stages. First, based on previous research and the logic of our argument, we expect that discordance will be more common among premarital cohabitors compared to those who entered marriage directly, as they are more likely to have ‘‘slid’’ into marriage. Second, we hypothesize that discordance is more common for less institutionalized relationship stages, such as dating and stayover relationships, while it is less frequent for more institutionalized relationship stages, such as marriage. Third, based on our argument that discordance is a marker for ‘‘sliding’’ into a relationship stage without an explicit decision or commitment, we hypothesize that it is associated with lower marital quality. Finally, we hypothesize that

1146

S. Halpern-Meekin, L. Tach / Social Science Research 42 (2013) 1143–1155

discordance is most strongly associated with marital quality for the more institutionalized relationship stages, such as marriage, because it is a stronger indication of less commitment and certainty about the relationship. 3. Data and method 3.1. Data Our analysis uses data from an internet survey conducted by Knowledge Networks (KN), in conjunction with the National Center for Family and Marriage Research (NCFMR) at Bowling Green State University, between July and October 2010. In 1999, KN established the first online research panel (Knowledge Panel [KP]) that is representative of the U.S. population ages 18–64, using probability-based sampling methods that include computer users and non-users. If the panel members did not have access to the Internet, they were provided computer equipment to participate in the study. A recent evaluation found that a survey using the KP was comparable to a nationally representative random digit dial (RDD) telephone survey sample, and the data obtained from the KP internet component were superior to the RDD sample with respect to reliability and validity (Chang and Krosnick, 2009). Knowledge Networks assigned the NCFMR survey to 1500 men in the panel who indicated that they were currently married to an opposite sex spouse, of whom 1060 completed the survey. The wives of men who completed the survey were assigned the survey as well, and 752 of them completed the survey. The survey team elected to sample men first because men tend to have lower response rates than women; sampling men first is expected to lead to fewer cases in which surveyors received responses from only one member of a couple. The full sample includes data from each partner in 752 couples, resulting in 1504 survey responses. Data are weighted to adjust for sample design and survey non-response. Knowledge Networks constructed post-stratification weights based on respondent gender, age, race/ethnicity, education, census region, metropolitan area, and internet access (Knowledge Networks, 2010). We further divide the sample into two groups: married couples who cohabited before marriage (N = 380) and married couples who did not cohabit before marriage (N = 369). 3.2. Measures We designed survey questions to measure the onset and length of several relationship stages, including dating, stayovers, cohabitation, and marriage. Due to space limitations in the survey, we did not collect complete relationship histories, but rather focused on the path of the current relationship with particular attention to the beginning and duration of each relationship stage. For all relationship stage questions, respondents were not given a ‘don’t know’ option, but were asked to provide their best guess. 3.2.1. Dating relationship stage First, we asked all respondents for the date (month/year) they started dating their current partners. We also asked respondents whether they had ever separated and gotten back together while dating. We calculate the duration of the dating relationship as the number of months between the date the respondents reported they started dating and the date they started the stayover relationship stage. 3.2.2. Stayover relationship stage We define the stayover relationship stage as the period of time when couples transition into full-time co-residence, occasionally spending the night with each other but not viewing themselves as cohabiting. We asked respondents, ‘‘How long before you were officially living together did you and your partner start spending the night at one another’s homes?’’ Respondents could report a number of weeks or months, or they could respond that they never spent the night at one another’s homes before officially living together. If they provided a non-zero response to this question, they were asked about how many times in a typical week they spent the night at one another’s homes, ranging from one to seven. Previous research indicates that respondents can recall the process of starting to spend nights together before officially cohabiting (Manning and Smock, 2005; Pollard and Harris, 2007; Sassler, 2004). We calculate the duration of the stayover relationship stage as the number of months the respondents reported they spent the night at one another’s homes before officially living together. We exclude couples who agreed that they did not have stayover relationships in analyses of the duration of the stayover relationship stage and the number of nights spent in stayovers during a typical week. 3.2.3. Cohabitation relationship stage We asked respondents about marital intentions: ‘‘Before you were officially living together, had you and your partner decided to get married in the future?’’ We then asked respondents whether they lived together before marrying and asked them to report the date (month/year) they officially started living together. Those who reported that they had lived together before marriage were then asked whether they ever separated and got back together while cohabiting. We use the phrase ‘‘‘‘officially’’ living together’’ and do not describe ‘‘officially’’ to allow respondents to self-define the start of their cohabiting relationship. Providing a definition assumes a common understanding of the start of cohabitation that qualitative research indicates is not the same across all cohabitors (Manning and Smock, 2005). Previous studies, both qualitative and

S. Halpern-Meekin, L. Tach / Social Science Research 42 (2013) 1143–1155

1147

quantitative, have found that respondents are able to report the beginning dates of dating and cohabitation and the length of time between the start of the relationship and cohabitation (Rhoades et al., 2006, 2009; Sassler, 2004). We calculate the duration of the cohabitation relationship stage as the number of months between the date the respondents reported they officially started living together and the date they got married. 3.2.4. Marital relationship stage Finally, we asked respondents to report on the date (month/year) that they got married. We also asked them whether they ever separated and got back together while married. We calculate the duration of the marital relationship stage as the number of months between the date the respondent reported getting married and the survey date. 3.2.5. Relationship stage discordance We selected a 1 month discrepancy in couple reports as the benchmark to gauge discordance because we did not expect couples to recall the exact dates they began new relationship stages (with perhaps the exception of a wedding date). Not surprisingly, the level of discordance declines if we extend the allowable distance between husbands’ and wives’ reports to 2 months. Levels of discordance were roughly 5–10 percentage points lower if we used a 2-month threshold for discordance instead of a 1-month threshold. 3.2.6. Marital quality The survey includes several measures of marital quality. A measure of relationship satisfaction asks respondents, ‘‘Taking all things together, how satisfied are you with your relationship?’’ A measure of listening satisfaction asks ‘‘How satisfied are you with how well your spouse listens to you?’’ For both measures, respondents answered on a five-point scale ranging from very dissatisfied (1) to very satisfied (5). A measure of partner supportiveness takes the mean of the following items: ‘‘My spouse shows love and affection toward me’’; ‘‘My spouse encourages me to do things that are important to me’’; ‘‘My spouse will not cheat on me’’; ‘‘My spouse listens when I need someone to talk to’’; and the reverse code of ‘‘My spouse and I avoid discussing unpleasant or difficult topics.’’ Responses were on a five-point scale from strongly disagree (1) to strongly agree (5). A measure of relationship happiness asked respondents ‘‘How would you rate your relationship with your current spouse?’’ with responses ranging from completely unhappy (1) to completely happy (10). Finally, respondents were asked ‘‘What are the chances you and your spouse will break up in the future?’’ with responses ranging from no chance (1) to almost certain chance (5). Because these marital quality measures were highly correlated with each other, we reverse coded negative measures, standardized each (M = 0, SD = 1), and created a composite measure of marital quality that is the mean of these five standardized measures.1 3.2.7. Background characteristics We include measures of respondents’ demographic and economic characteristics as control variables. We measure a respondent’s age (<30, 30–44, 45–59, or 60+), educational attainment (less than high school, high school graduate, some college, or college graduate), race (non-Hispanic black, non-Hispanic white, non-Hispanic other, or Hispanic), whether there are children under the age of 18 in the household, household income (<$20,000, $20–39,999, $40–59,999, $60–99,999, or $100,000+), and employment status (unemployed, employed, or retired/disabled). 3.3. Method The following analysis proceeds in two steps. First, we provide descriptive data on each partner’s report of the duration of the following relationship stages: dating to first stayover, first stayover to cohabitation, and cohabitation to marriage. Couples who cohabited before marriage experience all three stages, while couples who entered marriage without cohabiting only experience the transitions into stayovers and marriage. We also provide descriptive evidence from each partner on whether they broke up and got back together during each of these relationship stages. We then provide descriptive evidence on the frequency of discordance in couples’ retrospective reporting of their relationship stages, and run random effects logistic regressions to predict the likelihood of discordance based on couples’ background characteristics. Next, we examine the associations between intra-couple discordance and marital quality. Using hierarchical linear models (HLM) that account for non-independence of observations within couples, we regress marital quality on indicators of couple discordance over each relationship stage, controlling for the background characteristics described above:

Marital Qualityij ¼ b0j þ b1 ðWifeÞij þ

N X

bn X nij þ rij

ð1Þ

n¼2

The first level of the model regresses marital quality for respondent i in couple j on a vector of n individual-level covariates and a dummy variable indicator for whether the respondent is the wife (and the omitted reference is husband). The intercept and regression slopes are allowed to vary across couples. In level two of the model, the couple level, the intercept boj is predicted by a couple-level indicator of discordance to test whether discordance is a risk factor for lower marital quality: 1

We also replicated our analyses separately for each measure of marital quality, and results were substantively and statistically similar.

1148

S. Halpern-Meekin, L. Tach / Social Science Research 42 (2013) 1143–1155

b0j ¼ k0 þ k1 ðDiscordantÞj þ lj We also test whether the association between discordance and marital quality differs by husbands’ and wives’ reports of marital quality by estimating a cross-level interaction between discordance and the dummy variable for wife. 4. Results 4.1. Descriptive results Table 1 provides the descriptive characteristics of the sample separately for husbands and wives. Couples who cohabited before marriage report slightly lower quality marriages than couples who did not cohabit before marriage. Similarly, husbands report significantly more positive marriages on average than do wives. On average, premarital cohabitors are younger, have lower education and household income, and are more likely to be unemployed than non-cohabitors. They have also been married for fewer years than non-cohabitors. There are few gender differences in reports of background characteristics, except that wives have lower educational attainment and are less likely to be employed than husbands. Table 2 presents detailed descriptive statistics for husbands’ and wives’ reports of the mean starting dates and durations of each premarital relationship stage. On average, premarital cohabitors spent less time dating than non-cohabiting married couples, about 11 months compared to about 24 months, indicating that cohabitors transitioned faster from dating to stayover relationships than non-cohabitors. Premarital cohabitors were also more likely than non-cohabitors to spend the night together in stayover relationships before officially living together. Over 70% of premarital cohabitors reported stayovers, Table 1 Descriptive characteristics of premarital cohabitors and non-cohabiting married couples. Premarital cohabitors Husband Relationship quality

0.086

Non-cohabiting marrieds Wife 0.079**

Husband 0.101

Wife 0.010a,*

Background characteristics Age <30 30–44 45–59 60+

10.59 46.09 37.39 5.92

14.54* 38.62 38.10 8.74

9.22 32.57a 42.24 15.96a

12.45 33.33a 41.33a 12.88a,*

Education Less than high school High school graduate Some college College graduate

18.51 30.01 23.12 28.27

13.08+ 28.49+ 30.69* 27.74

5.07a 27.95 26.82 40.14a

7.21 27.62+ 25.69+ 39.47a

Race Non-Hispanic white Non-Hispanic black Hispanic Non-Hispanic other

73.23 6.32 14.55 5.90

76.81 4.97 12.22 5.99

75.02 2.76 14.67 7.54

74.94 3.61 14.59 6.86

Children <18 in household

50.57

49.09

49.62

49.10

Income <$20,000 $20–$39,999 $40–$59,999 $60–$99,999 $100,000+

11.19 15.07 21.36 29.33 23.04

11.39 14.30 22.91 29.46 21.94

3.93a 11.23 16.09 43.21a 25.53

3.14a 10.02 16.87 45.59a 24.38

Employment Employed Unemployed/not in labor force Other (retired/disabled)

75.73 11.39 12.87

57.80*** 31.27*** 10.92

82.67 6.76a 10.56

54.29*** 30.54*** 15.17

Marital duration in years Unweighted N

14.75 380

16.66 380

Notes: Descriptive statistics are weighted using national sampling weights. Asterisks indicate significance tests for difference between male and female means. + p < 0.1. * p < 0.05. ** p < 0.01. *** p < 0.001. a Significant difference between same-sex premarital cohabitors and non-cohabitors at p < 0.05.

22.75a 369

24.58a 369

1149

S. Halpern-Meekin, L. Tach / Social Science Research 42 (2013) 1143–1155 Table 2 Descriptive statistics for relationship stage measures. Premarital cohabitors

Non-cohabiting marrieds

Husband

Wife

Husband

Wife

20-January92

2-January-92

27-March-85a

22-June-85a

# Months between dating and first spending the night together Separated, got back together while dating

10.58 17.98

11.38 18.95

24.11a 10.96a

23.62a 14.08a

Spending the night Spent the night together before officially living together # Weeks between first spending the night together and officially living togetherb # Nights per week spent the night together before officially living togetherb

73.76 10.69 2.59

80.01 11.75 2.94

29.37a 19.01a 0.09a

27.34a 23.43a 0.27a

Cohabiting Date started living together

10-May-93

7-February93

–

–

Dating Date started dating

# Months between officially living together and getting married

23.67

21.61

–

–

Separated, got back together while living together Decided to marry before living together

6.58 42.33

6.84 48.43c

– –

– –

Married Date married

6-July-95

2-August-95

Separated, got back together while married

5.38

5.21

13-August87a 4.67

13-March87a 4.17

Notes: Descriptive statistics are weighted means and percentages. a Same-sex difference in means between premarital cohabitors and non-cohabiting marrieds is significant at the p < 0.05 level. b Excludes couples who agreed that they did not spend the night together before officially living together. c Difference between husbands and wives is significant at the p < 0.05 level.

compared to less than 30% of non-cohabiting married couples. Of those who reported stayovers, premarital cohabitors had stayover relationships for an average of 2–3 months before officially living together while non-cohabitors reported slightly longer average stayover relationships of 4–5 months. Cohabitors reported more intensive stayover relationships on average, however, spending more nights per week together during the stayover stage than non-cohabitors – about 2 nights per week, compared to less than 1 night per week on average for non-cohabitors. For non-cohabitors, this relationship stage is followed by marriage. For cohabitors, this relationship stage is followed by cohabitation, which lasted an average of 22–24 months before marriage. Couples were also asked whether they had separated and gotten back together during each relationship stage. Premarital cohabitors were significantly more likely to report that they had separated and gotten back together while dating than noncohabitors (18–19% versus 11–14%). Between 6% and 7% of premarital cohabitors reported that they had separated while cohabiting. Fewer couples in both groups reported that they had separated and gotten back together while they were married (5% for premarital cohabitors versus 4% for non-cohabitors). We also compared husbands’ and wives’ mean reports of premarital relationship stage durations and found that they did not differ significantly. 4.2. Intra-couple discordance over timing and duration of relationship stages The mean starting dates and durations of the relationship stages presented in Table 2 mask a great deal of heterogeneity in couples’ reports. Table 3 directly examines the amount of couple discordance over each relationship stage. During the dating stage, we find that 43% of premarital cohabitors and 37% of non-cohabitors report starting dates for their relationships that differ by more than 1 month. Over two-thirds of cohabitors (68%) and over one-half of non-cohabitors (54%) differ by more than 1 month in the number of months they report dating before starting stayovers. Additionally, 10% of cohabitors and 8% of non-cohabitors differ in their reports of whether they broke up and got back together while dating. Premarital cohabitors and non-cohabitors report similar amounts of discordance over the stayover stage of their relationships. About 17% of couples in both groups disagree whether they spent the night together before officially living together. Among couples in which at least one member reported a stayover relationship, over 70% differed by more than 1 month in their reports of how long they had stayovers before officially living together. 43% of premarital cohabitors differed by more than 1 day in their reports of how many nights on average they had stayovers in a typical week, compared to just 10% of non-cohabitors. Discordance is also a modal experience among premarital cohabitors regarding the cohabitation stage of their relationship. About half (49%) differ by more than 1 month on the date they started officially living together, and 57% differ by more than 1 month on how long they report cohabiting before getting married. Interestingly, 22% of premarital cohabitors also disagreed over whether they had decided to get married before living together. Couple concordance was considerably higher concerning marriage for both premarital cohabitors and non-cohabitors. Just 14% of premarital cohabitors and 10% of non-cohabitors differed by more than 1 month in their reports of their marriage

1150

S. Halpern-Meekin, L. Tach / Social Science Research 42 (2013) 1143–1155

Table 3 Couple discordance on relationship stage measures. Premarital cohabitors

Non-cohabiting marrieds

% Disagree

Correlation

% Disagree

Correlation

Dating Date started datinga # Months between dating and first spending the night togethera Separated, got back together while dating

42.97 67.96 9.74

0.749 0.193 –

37.39c 54.29c 8.44

0.825c 0.239c –

Spending the night Spent the night before officially living together # Months between first spending the night together and officially living togethera,b # Nights per week spent the night together before officially living togethera,b

16.84 72.24 43.08

– 0.271 0.308

17.10 75.56c 9.92c

– 0.472c 0.171c

Cohabiting Date started living togethera # Months between officially living together and getting marrieda Separated, got back together while living together Decided to marry before living together

49.32 57.18 5.00 22.37

0.766 0.382 – –

– – – –

– – – –

Married Date marrieda Separated, got back together while married

13.98 2.64

0.792 –

9.56 3.27

0.807 –

Notes: Couple (dis)cordance measured by % discordance for all measures and by correlations (r) for continuous and ordinal measures. a Discordance measured by >1 month (31 day) difference. b Excludes couples who agreed that they did not spend the night together before officially living together. c Significantly different between premarital cohabitors and non-cohabitors at p < 0.05 level.

dates. Similarly, few couples disagreed over whether they had separated or gotten back together while married – just 3% of cohabitors and non-cohabitors. Taken together, the results in Table 3 indicate discordance over the timing and duration of premarital relationship stages. Discordance was positively correlated across relationship stages, indicating that couples who were discordant on one relationship stage tended to be discordant on other relationship stages. Discordance was higher among premarital cohabitors than non-cohabitors, and it was highest for the relationship stages that are least institutionalized, such as stayovers, and lowest for marriage, the most institutionalized relationship stage. From these results, it is clear that many couples are not on the same page about the timing of their relationship progression, at least based on their retrospective reports. We next predict the likelihood of discordance based on couples’ demographic and economic characteristics. Because the pattern of results for premarital cohabitors and non-cohabitors was similar in predicting discordance (results not shown, available upon request), we pool the results for these two groups in the same model and include a dummy variable indicator for premarital cohabitors. In these analyses, we pool husbands and wives in the same model and estimate a random effects model that accounts for the non-independence of observations within couples. Table 4a reports the odds ratios of logistic regressions of couple discordance on husbands’ and wives’ background characteristics. The results in Table 4a show that, even net of controls for age, education, race, income, employment, and marital duration, premarital cohabitors remain significantly more likely to disagree than non-cohabitors for most relationship stages. Older couples were more likely to disagree than younger couples for most relationship stages, probably reflecting the fact that the accuracy of recall declines as more time passes. There were few consistent differences by income, race, employment, or education, suggesting that more disadvantaged couples are no more likely to disagree than more advantaged ones. Table 4b presents results for the cohabitation relationship stage for the premarital cohabitor sample alone. We find a similar pattern of results here, with older couples being more likely to disagree on most measures, and few other consistent predictors among background characteristics. The pseudo-R2 values for the models in Tables 4a and 4b are relatively low, however, indicating that the background characteristics included in our model do not do a very good job of explaining which couples disagree. By far the strongest correlate of discordance is premarital cohabitation.

4.3. Associations between couple discordance and marital quality Next, we ask whether discordance over relationship stages is associated with marital quality. Table 5 presents the results of regressions of the composite measure of marital quality on couple discordance separately for each relationship stage, controlling for background characteristics. We again pool husbands’ and wives’ reports and estimate HLM models to include both couple-level and individual-level covariates. We include a dummy variable indicator for wives that tests whether wives report significantly different marital quality than husbands. We also estimate a cross-level interaction between the discordance and wife variables, which tells us whether the association between discordance and marital quality is different for wives than it is for husbands. Each couple discordance measure was regressed in a separate model with the full set of controls, so each set of coefficients in Table 5 reports the results of a separate regression. Because results were substantively

1151

S. Halpern-Meekin, L. Tach / Social Science Research 42 (2013) 1143–1155 Table 4a Random effects regression of couple discordance on background characteristics – all respondents. Date started datinga

Months between dating and spending the nighta

Separated while dating

Spent the night before officially living together

Months between spending the night and officially living togethera

Nights/week spent the night before officially lived togetherb

Date marrieda

Separated while married

Premarital cohabitor

1.42*

2.06***

1.47

1.05

1.07***

2.82***

1.41

0.93

Age 30–44 45–59 60+

1.87+ 2.85* 4.63*

2.13* 2.16* 2.85*

0.95 1.08 2.09

1.03 0.85 0.68

0.66 1.15 2.13+

0.77 0.62 1.09

1.48 2.58 3.14

0.37 0.12+ 0.03*

Education Less than high school Some college College graduate

1.32 1.06 1.08

1.21 1.07 1.43+

2.25 0.82 0.89

0.44 0.97 0.93

0.86 1.07 3.26***

0.74 0.83 1.76*

2.76* 1.16 1.42

– 0.53* 0.09*

Race Non-Hispanic black Hispanic Non-Hispanic other Children <18 in household

1.44* 1.50* 3.22* 0.91

0.82 0.95 1.03 1.12

0.58 1.80 0.47 0.79

1.74 0.66 1.36 0.94

1.69 0.94 3.72+ 1.26

4.74* 1.11 0.84 1.16

2.65 2.18* 1.75 1.12

– 1.52 3.32 0.67

Income $20–$39,999 $40–$59,999 $60–$99,999 $100,000+

2.15 2.31+ 2.75* 2.15

1.78 1.56 1.61 1.55

0.87 0.82 1.18 0.72

1.12 0.81 1.21 1.22

1.41 0.94 1.36 1.18

1.10 0.84 0.49 0.50

1.89 1.31 2.36 2.34

0.27 0.06* 0.34 0.49

Employment Unemployed Other (retired/disabled)

1.28 1.05

0.65 0.96

0.32 1.14

2.01* 1.49

0.91 0.87

0.77 1.26

1.70 0.86

1.38 3.28+

Marital duration in months Pseudo R2

0.99 0.04

1.00 0.03

1.01 0.05

1.02+ 0.02

0.99 0.06

0.99 0.13

1.00 0.04

1.01 0.19

Notes: Odds ratios. Omitted reference categories are age < 30, high school graduate, non-Hispanic white, no children, income < $20,000, and employed.  p < 0.01. + p < 0.1. * p < 0.05. *** p < 0.001. a Discordance measured by >1 month (31 day) difference. b Discordance measured by >1 day difference.

similar for premarital cohabitors and non-cohabitors, we pooled them into a single sample and included a dummy variable for premarital cohabitation.2 During the dating stage of a relationship, discordance over when a couple started dating was associated with significantly lower marital quality. Couples who disagreed by at least 1 month over when they started dating reported that their marital quality was about one tenth of a standard deviation lower than couples who agreed on the start of their dating relationships. Discordance over whether a couple separated while dating was also associated with lower marital quality. Wives reported consistently lower marital quality than husbands. In Model 2, we add an interaction between respondent’s gender and discordance. We find that, for two of the three dating discordance measures, the negative association between discordance and marital quality was significantly weaker for wives than it was for husbands. Unlike the dating stage, discordance over how long couples were in the stayover relationship stage, and how many nights per week they had stayovers, were only weakly associated with lower marital quality. Among premarital cohabitors, discordance over the cohabitation stage was inconsistently associated with lower marital quality. Both husbands and wives who disagreed over whether they had separated and gotten back together while cohabiting reported lower quality marriages than their same-sex counterparts who were concordant. Discordance over the date cohabitation began and the duration of cohabitation was associated with husbands’ reports of lower marital quality, but not wives’ reports. Notably, discordance over whether the couple had agreed to marry before living together was not significantly associated with marital quality. At the marital stage, discordance over the date couples married was not associated with marital quality. However, discordance over whether the couple had separated and gotten back together while married was strongly and consistently asso-

2 We also reran our analyses excluding couples in which at least one respondent was age 60 or older to test whether our results were sensitive to the inclusion of this older cohort. The results were statistically and substantively the same, so we retained the full age distribution in our analyses.

1152

S. Halpern-Meekin, L. Tach / Social Science Research 42 (2013) 1143–1155

Table 4b Random effects regression of couple discordance on background characteristics – premarital cohabitors only. Date started living togethera

# Months between officially living together and getting marrieda

Separate while living together

Decided to marry before living together

1.21 1.89 4.18+

1.23 1.85 3.81+

1.99 3.04 –

2.50* 2.94* 2.12*

1.17

1.32

3.21+

2.13

1.21 1.37

1.20 1.19

0.51 0.72

0.74 0.73

0.80

1.54

1.10

1.26

2.48 2.31+

1.35 2.06

3.03 1.42

1.03 1.97

0.90

0.66+

0.78

0.87

2.50+ 2.53+ 2.89+ 3.33*

2.42+ 1.93 2.99* 3.28*

0.82 1.08 0.23 0.88

0.68 0.95 1.68 1.34

Employment Unemployed Other (retired/ disabled)

1.09 0.77

1.44 0.78

0.36 0.90

0.91 1.14

Marital duration Pseudo R2

1.00 0.05

1.00 0.04

1.00 0.11

0.99 0.06

Age 30–44 45–59 60+ Education Less than high school Some college College graduate Race Non-Hispanic black Hispanic Non-Hispanic other Children <18 in household Income $20–$40,000 $40–$60,000 $60–$100,000 $100,000+

*

Notes: Odds ratios. Omitted categories are age < 30, HS graduate, non-Hispanic white, no children, income < $20,000, and employed.  p < 0.01. p < 0.001. + p < 0.1. * p < 0.05. a Discordance measured by >1 month (31 day) difference.



ciated with lower quality marriages. In fact, this was the single strongest predictor of poor marital quality among all our measures of couple discordance. Taken together, the results in Table 5 suggest that discordance over the progression of couples’ relationships is significantly associated with lower quality marriages. This is particularly true when couples disagree over the progression of the earliest relationship stages and is stronger for husbands’ reports of marital quality than it is for wives’ reports. Discordance over the dates of cohabitation and marriage are less consequential for marital quality; what seems to matter during these relationship stages is whether the couple disagrees about past breakups. 5. Discussion and conclusions 5.1. Discussion Discordance in couples’ recounting of premarital courtship stages is quite common, particularly among premarital cohabitors and for the less institutionalized courtship stages of dating and stayovers. The higher rates of discordance among premarital cohabitors are not accounted for by standard demographic controls. Discordance, in turn, is associated with lower marital quality, although the magnitude of this association differs across relationship stages and may be stronger for husbands than for wives. We argue that discordance is an indicator that partners ‘‘slid’’ into a more serious relationship, as it may be more likely when a couple lacks explicit markers or rituals marking turning points in the relationship (like having ‘‘the talk’’ about the relationship’s future). Along these lines, our findings support the contention in previous theory and research that sliding is associated with more negative marital outcomes (Stanley et al., 2006, 2010a). We maintain that discordance in retrospective relationship accounts may reflect less clarity in the process of how couples’ relationships developed – an indication that they may not have shared clear signals of commitment along the way. Notably, our results indicate that sliding during the less institutionalized stages of a relationship – dating and stayovers – is common and consequential for marital quality. Further,

1153

S. Halpern-Meekin, L. Tach / Social Science Research 42 (2013) 1143–1155 Table 5 HLM regression of marital quality on couple discordance about relationship stages. Marital quality Model 1

Model 2

Discordance on Dating Date started datinga Female Female  Discordant

0.087* 0.123*** –

0.123** 0.152*** 0.074+

# Months between dating and first spending the nighta Female Female  Discordant

0.056 0.122*** –

0.094* 0.167*** 0.076+

Separated, got back together while dating Female Female  Discordant

0.138* 0.121*** –

0.152* 0.123*** 0.029

Spending the night # Months between first spending the night and officially living togethera Female Female  Discordant

0.039 0.132*** –

0.083 0.180*** 0.076

# Nights per week spent the night together before officially living togetherb Female Female  Discordant

0.039 0.119*** –

0.069 0.138*** 0.059

Cohabiting Date started living togethera Female Female  Discordant

0.072 0.140*** –

0.183** 0.249*** 0.226***

# Months between officially living together and getting marrieda Female Female  Discordant

0.038 0.147*** –

0.127+ 0.249*** 0.179**

Separated, got back together while living together Female Female  Discordant

0.404** 0.142*** –

0.206 0.124*** 0.396**

Decided to marry before living together Female Female  Discordant

0.062 0.143*** –

0.038 0.133** 0.049

Marriage Date marrieda Female Female  Discordant

0.017 0.119*** –

0.039 0.131*** 0.115+

Separated, got back together while married Female Female  Discordant

0.449*** 0.121*** –

0.548*** 0.128*** 0.196

Notes: Cohabiting results are for premarital cohabitors only. Regressions control for premarital cohabitation, age, education, race, children in household, household income, employment status, and marital duration. Dependent variable is the average of five standardized measures of marital quality. + p < 0.1. * p < 0.05. ** p < 0.01. *** p < 0.001. a Discordance measured by >1 month (31 day) difference. b Discordance measured by >1 day difference.

discordance appears to be more common among those entering marriage through cohabitation. Future research should develop and test theory on the basis of the present results, investigating why couple concordance on certain premarital relationship stages or transitions is more closely tied to marital quality than others, and why this may be more true for husbands than wives; in addition, the association between discordance and divorce should be explored. Our findings have implications for researchers interested in the measurement of relationship stages. First, premarital cohabitors are more likely than non-cohabitors to disagree in recounting their courtship stages. Therefore, studies that rely on measures from only one partner must be particularly cautious when comparing results for premarital cohabitors and those who entered marriage directly, as the higher likelihood of divergent accounts among the cohabitors means we may draw less reliable conclusions about them.

1154

S. Halpern-Meekin, L. Tach / Social Science Research 42 (2013) 1143–1155

Second, couple discordance on whether they had agreed to get married before beginning cohabitation is not uncommon – occurring among nearly one-quarter of the couples. Those examining the presence or absence of engagement prior to cohabitation and relying only on the reports of one partner for their measures need to be cautious as these may contain a good deal of measurement error. Relatedly, discordance on whether the couple had agreed to get married before cohabiting was not significantly associated with marital quality; however, in supplemental analyses we found that the couple agreeing that they had not decided to get married before cohabiting was associated with lower marital quality (results not shown), which is consistent with studies showing that pre-engagement cohabitation is associated with poorer marital outcomes (Rhoades et al., 2009; Stanley et al., 2010a). While we ask whether couples had plans to marry before living together, we do not ask specifically about engagement; the public nature of an engagement as a relationship stage marker (Stanley et al., 2010b)—versus potentially private plans to marry—might offer different results if captured in a future study. 5.2. Limitations The present study contributes to our understanding of the courtship process by using a unique dataset that includes couple-level data drawn from a national panel. However, the married couples in the sample have been married for a long time – 15 years on average for premarital cohabitors and 23 years for those who entered marriage directly. This means we have a sample that is slanted toward longer-lasting marriages; for example, of marriages occurring between 1980 and 1984, one-third of couples were divorced by their 15th anniversary (Kreider, 2005). In terms of testing our hypotheses, this sample creates a bias in a conservative direction, as we should be less likely to find differences based on premarital courtship experiences and between premarital cohabitors and those who entered married directly among these longer-lasting marriages. Therefore, although the prevalence of couple discordance found here is not representative of all marriages, it does offer a more conservative test of our hypotheses. Future studies should examine the courtship processes explored in the present study in the context of ‘‘younger’’ marriages in order to capture the characteristics of those that will later end in divorce. Future research should also focus more specifically on whether there are consistent patterns of ‘‘his’’ and ‘‘hers’’ retrospective courtship stage reporting and whether such gender differences are tied to variation in marital quality. In addition, because we do not collect complete relationship histories, we cannot account for the potential issue that people with more relationships prior to the current union may have more difficulty in accurately recounting their relationship stages. This issue should be explored in future studies because previous research indicated that having multiple previous partners predicted more negative marital outcomes (Teachman, 2003). We are also unable to include all potentially important control variables in our models, such as religiosity, given the limited number of questions asked in the Knowledge Panel survey. We theorize that rituals marking relationship transitions can make passage through relationship stages more salient, forcing a more active decision to advance the relationship and making this transition more memorable to both partners. However, we only measure spouses’ recall of these events; we do not directly measure why some couples share more similar memories than others. While we speculate that the occurrence of rituals is a priming event, more apt to create shared memories, there are other possible mechanisms at work. Stanley and colleagues’ theory about sliding versus deciding is about how commitment develops in a relationship. This commitment may come from enacting rituals (a form of deciding leading to increased dedication) or from inertia (a form of sliding leading to increased constraint). Related processes could include the couple’s development of mutual commitment (Drigotas et al., 1999) or a shared identity (Acitelli et al., 1999; Agnew et al., 1998) as factors mediating the association between retrospective reporting concordance and marital quality. Because we are unable to adjudicate between these possibilities using our cross-sectional data, exploring these additional mediating processes is an important area for future research. 5.3. Conclusions By taking advantage of the relatively rare opportunity to examine couple-level data on multiple relationship stages prior to, and including, cohabitation and marriage, we found that discordance in couples’ retrospective accounts of premarital courtship is both common and meaningful. There are important implications for survey research on relationship histories, particularly given the higher couple discordance rates among premarital cohabitors relative to non-cohabitors. However, this is not just a measurement issue; rather, discordance is a substantively meaningful marker associated with poorer marital quality. That discordance over the less institutionalized stages of the courtship process is common and associated with marital quality indicates that these may be stages in the premarital relationship that couples more easily ‘‘slide’’ through, with negative implications for their marital experiences. References Acitelli, L.K., Rogers, R., Knee, C.R., 1999. The role of identity in the link between relationship thinking and relationship satisfaction. Journal of Social and Personal Relationships 16, 591–618. Agnew, C.R., Van Lange, P.A.M., Rusbult, C.E., Langston, C.A., 1998. Cognitive interdependence: commitment and the cognitive representation of close relationships. Journal of Personality and Social Psychology 74, 939–954. Brown, S.L., Booth, A., 1996. Cohabitation versus marriage: a comparison of relationship quality. Journal of Marriage and Family 58, 668–678. Bulanda, J.R., Brown, S.L., 2007. Race-ethnic differences in marital quality and divorce. Social Science Research 36, 945–967. Casper, L.M., Cohen, P.N., 2000. How does POSSLQ measure up? Historical estimates of cohabitation. Demography 37, 237–245.

S. Halpern-Meekin, L. Tach / Social Science Research 42 (2013) 1143–1155

1155

Chang, L., Krosnick, J.A., 2009. National surveys via RDD telephone interviewing versus the internet: comparing sample representativeness and response quality. Public Opinion Quarterly 73, 641–678. Drigotas, S.M., Rusbult, C.E., Verette, J., 1999. Level of commitment, mutuality of commitment, and couple well-being. Personal Relationships 6, 389–409. Grimes, R.L., 2000. Deeply into the Bone: Re-Investing Rites of Passage. University of California, Berkeley. Gusfield, J.R., Michalowicz, J., 1984. Secular symbolism: studies of ritual, ceremony, and the symbolic order in modern life. Annual Review of Sociology 10, 417–435. Halpern-Meekin, S., Manning, W.D., Giordano, P.C., Longmore, M.A., 2013. Relationship churning in young adulthood: on/off relationships and sex with an ex. Journal of Adolescent Research 28, 166–188. Hayford, S.R., Morgan, S.P., 2008. The quality of retrospective data on cohabitation. Demography 45, 129–141. Hochschild, A.R. 1998. The sociology of emotion as a way of seeing. In: Bendlow, G., & William, S.J., Emotions in Social Life: Critical Themes and Contemporary Issues. Routledge, London, pp. 3–15. Kline, G.H., Stanley, S.M., Markman, H.J., Olmos-Gallo, P.A., St. Peters, M., Whitton, S.W., Prado, L.M., 2004. Timing is everything: pre-engagement cohabitation and increased risk for poor marital outcomes. Journal of Family Psychology 18, 311–318. Knowledge Networks, 2010. Field Report to National Center for Family and Marriage Research. Unpublished Report. Kreider, R.M., 2005. Number, Timing, and Duration of Marriages and Divorces: 2001. Current Population Reports, P70-97. U.S. Census Bureau, Washington, DC. Lillard, L.A., & Waite, L.J., 1989. Panel versus retrospective data on marital histories: lessons from the PSID. In: Individuals and Families in Transition: Understanding Change through Longitudinal Data. Social Science Research Council, pp. 243–253. Manning, W.D., Smock, P.J., 2005. Measuring and modeling cohabitation: new perspectives from qualitative data. Journal of Marriage and Family 67, 989– 1002. Nock, S.L., 1995. A comparison of marriages and cohabiting relationships. Journal of Family Issues 16, 53–76. Peters, E., 1988. Retrospective versus panel data in analyzing lifecycle events. The Journal of Human Resources 23, 488–513. Pollard, M.S., Harris, K.M., 2007. Measuring cohabitation in add health. In: Hofferth, S.L., Casper, L.M. (Eds.), Handbook of Measurement Issues in Family Research. Lawrence Erlbaum, Mahwah, NJ, pp. 35–52. Rhoades, G.K., Stanley, S.M., Markman, H.J., 2006. Pre-engagement cohabitation and gender asymmetry in marital commitment. Journal of Family Psychology 20, 553–560. Rhoades, G.K., Stanley, S.M., Markman, H.J., 2009. The pre-engagement cohabitation effect: a replication and extension of previous findings. Journal of Family Psychology 23, 107–111. Sassler, S., 2004. The process of entering into cohabiting unions. Journal of Marriage and Family 66, 491–505. Schweingruber, D., Anahita, S., Berns, N., 2004. When the answer is known: the engagement proposal as performance. Sociological Focus 37, 143–161. Stanley, S.M., Rhoades, G.K., Markman, H.J., 2006. Sliding versus deciding: inertia and the premarital cohabitation effect. Family Relations 55, 499–509. Stanley, S.M., Rhoades, G.K., Amato, P.R., Markman, H.J., Johnson, C.A., 2010a. The timing of cohabitation and engagement: impact on first and second marriages. Journal of Marriage and Family 72, 906–918. Stanley, S.M., Rhoades, G.K., Whitton, S.W., 2010b. Commitment: functions, formation, and the securing of romantic attachment. Journal of Family Theory and Review 2, 243–257. Tach, L., Halpern-Meekin, S., 2012. Premarital cohabitation, nonmarital childbearing, and divorce decisions: how does marital quality matter? Family Relations 61, 575–581. Teachman, J., 2003. Premarital sex, premarital cohabitation, and the risk of subsequent marital dissolution among women. Journal of Marriage and the Family 65, 444–455. Teachman, J.D., Polonko, K.A., 1990. Cohabitation and marital stability in the United States. Social Forces 69, 207–220. Teitler, J.O., & Reichman, N.E., 2001. Cohabitation: An Elusive Concept. Social Indicators Survey Center Working Paper #01-4. Teitler, J.O., Reichman, N.E., Koball, H., 2006. Contemporaneous versus retrospective reports of cohabitation in the Fragile Families Survey. Journal of Marriage and Family 68, 469–477. Thomson, E., Colella, U., 1992. Cohabitation and marital stability: quality or commitment? Journal of Marriage and the Family 54, 259–267. Twenge, J.M., Campbell, W.K., Foster, C.A., 2003. Parenthood and marital satisfaction: a meta-analytic review. Journal of Marriage and Family 65, 574–583. Van Laningham, J., Johnson, D.R., Amato, P.R., 2001. Marital happiness, marital duration, and the U-shaped curve: evidence from a five-wave panel study. Social Forces 79, 1313–1341. Vennum, A., Fincham, F.D., 2011. Assessing decision making in young adult romantic realtionships. Psychological Assessment 23, 739–751. Waite, L.J., Lillard, L.A., 1991. Children and marital disruption. American Journal of Sociology 96, 930–953. Waller, M.R., McLanahan, S.S., 2005. ‘‘His’’ and ‘‘her’’ marriage expectations: determinants and consequences. Journal of Marriage and Family 67, 53–67. White, L.K., Rogers, S.J., 2000. Economic circumstances and family outcomes: a review of the 1990s. Journal of Marriage and the Family 62, 1035–1051. Young, F., 1965. Initiation Ceremonies. Bobbs-Merrill, Indianapolis, IN.