Religious Dynamics and Marital Dissolution: A Latent Class Approach

Abstract

Past research on religious homogamy has struggled to distinguish whether religiosity or homogamy has a stronger impact on preventing a marital dissolution. In order to rectify this problem, I use a latent class approach to compare couples with various forms of partner religiosity and similarity. Based on 707 newlywed couples from the Marriage Matters survey (1998)(1999)(2000)(2001)(2002)(2003)(2004), I discovered four latent classes: "holy" couples (both partners are highly religious), "non-attending" couples (both partners identify as religious, but don't regularly attend services), "unbalanced" couples (the wife is religious, but the husband is not), and "secular" couples (both partners are not religious). Findings indicate that holy, non-attending, and unbalanced couples experience less odds of divorce compared to secular couples, suggesting that religiosity in a variety of forms is more important than partner similarity in avoiding divorce.

In sickness and in health: the longitudinal associations between marital dissatisfaction, depression and spousal health

“…Nor does covenant marriage predict decline in marital satisfaction over time for couples in this sample (DeMaris et al ., ). Further, the MMPS sample is similar in terms of race, education, religious affiliation and income to recently married couples across the US at the time of initial data collection (Gurrentz, ). Therefore, participants were included in the present analyses regardless of type of marriage, and regardless of whether they experienced a disruption in their marriage during the course of data collection.…”

Couple Religiosity and Well-Being Among Older Adults in the United States

“…Religious belief salience was coded where 1 = "agree" and "strongly agree" that religious beliefs were carried into daily life, and 0 = all else. Finally, difference in religious affiliation is a binary variable coded as "1" if each partner selected a different religious and "0" if they selected the same option (see Gurrentz, 2017). If couples were both not religiously affiliated, they received a score of 0 here.…”

“…Following Gurrentz (2017), we included a measure of husband's attendance, wife's attendance, husband's religious importance, wife's religious importance, and whether there were differences in religious affiliation for each couple (0 = no and 1 = yes). Each member of the couple is assigned a score of "1" if the couple reports different affiliations in the latent-class analysis.…”

“…Alternatively, is the religiousness of one partner sufficient to improve their own and their partner’s well-being? We build on previous work by Gurrentz (2017), who examined these questions in the context of predicting marital dissolution, by assessing aspects of physical and mental well-being.…”

“…We include control variables for each partner in our models, which were identical to those included for each individual (e.g., age, race, employment and education). This was done to adjust for potential biases from non-independence between members of an intimate relationship, a common pitfall in individual-level analyses of marital relationships (see Gurrentz, 2017 for a similar approach). Partner controls are not shown in our main statistical tables due to space constraints (available upon request).…”