Understanding the Relationship between Religiosity and Marital Commitment to Marital Stability: An Investigation on Indonesian Female Journalists

: The complete understanding of marital stability is hindered by limitations of theory and method, especially investigation on female journalists. The purpose of the current study was to test the effect of religiosity and marital commitment on the marital stability, by assessing Indonesian female journalists. This research used a quantitative approach with multiple regression analysis methods. The sample of this study involved 200 married female journalists residing in Jakarta and were taken using non-probability sampling techniques, speciﬁcally purposive sampling. The measurements used in this study were adaptations of the (1) Marital Stability Scale; (2) Centrality of Religiosity Scale (CRS-15); and (3) Inventory of Marital Commitments. Conﬁrmatory factor analysis (CFA) was used to test the validity of each scale. The results of the F -test showed a p -value = 0.000 (signiﬁcant), and a total variance explained ( R 2 value) of 0.224. This ﬁnding indicated that religiosity and marital commitment have a signiﬁcant effect on marital stability (sig < 0.05). The direction of the coefﬁcient regression of the religiosity variable and marital commitment is positive, indicating that the higher the religiosity and marital commitment, the higher the marital stability.


Introduction
Prior to the 21st century, changes in the modern era revolutionized the social roles of both males and females. According to Livingston and Judge (2008), historically, the realms of work and household have been gendered, whereby men dominate in workplaces and women dominate in nurturing the lives of families and fulfilling household's chores. However, nowadays many women take parts in the male role, namely working and earning a living for the family. The data from the BPS (Indonesian Central Statistic Agency) in Indonesia for the last four years show the number of working women in Jakarta has increased from year by year, resulting in common encounters with working wives. This shows the increasing contribution of wives in workplaces in Jakarta who are actively working for the family from year to year. According to Duffy and Atwater (2007), work may result in conflict between husband and wife, consequently reducing harmony and stability of the marriage.
One of the difficulties for a female journalist is to divide her efforts in the workplace and at home. According to Mallat and Melki (2016) in their qualitative interview, female journalists tend to see marriage and motherhood as obstacles for their career. There are many obstacles experienced by female journalists who are married, triggering conflicts between roles, both at work and in household life. According to a pilot study conducted by researchers, 63.6% of female journalists felt burdened by their work and 54.5% of female journalists felt that being a journalist adds to the stress of life. Other data also state that 59.1% of female journalists find it difficult to manage their time working as a journalist and being a wife at home. This difficulties may increase the risk of divorce (Becker et al. 1977;

Work-household Spill-over Effects
Cherlin (in Cooke and Gash 2010) argued that wives with independent income are less motivated to solve problems in a marriage. This perspective resulted in what became known as the independence hypothesis, in which the work of a wife is thought to increase the risk of divorce (Becker et al. 1977). This greatly affects the marital stability in the household life of married couples. Wives with independent finance will feel less obliged to solve problems in their household due to the beliefs that they will still live well even without their husbands. Cooke and Gash (2010) also found the effects of wives' employment on marital stability vary across the countries. In West Germany with its high-quality part-time employment, couples where the wife works part time are significantly more stable. In the more liberal British and US labor markets, neither wives' part-nor full-time employment significantly alters divorce risk. In the United States, however, mothers working part time have significantly lower divorce risk. West German and British husbands' unemployment proves more detrimental to marital stability than wives' employment. These results highlight the importance of the socioeconomic context in structuring the optimal employment participation of both partners.

Marital Commitment
Marriage is certainly associated with commitment. According to Adams and Jones (1997), marital commitment is one of the important variables in understanding marital stability. Marital commitment is a variable that has a big role in explaining marital stability (Gunter 2004). Marital commitment refers to a decision to maintain the current marriage relationship (Surra et al. 1999). According to Johnson et al. (1999), marital commitment cannot be seen as a single unit, but rather consists of three different types of commitment experiences: Personal commitment, structural commitment, and moral commitment. He argues that each type of commitment performs according to its own function, thus commitment cannot be reviewed as a whole.
According to Swensen and Trahaug (1985), a high level of commitment to a person is associated with a smaller number of marriage problems. This creates more conducive conditions for marriage survivability. Therefore, commitment is said to be the main reason for the predictor of marital stability (Lauer and Lauer 1986). There is also research on marital commitment and marital stability in individuals with high religiosity.

Religiosity
Another factor that might affect marital stability is religiosity. Indonesia is well known as the biggest Muslim population in the world. For Muslims, religiosity is usually a guidance in problem solving when facing difficult situations and stressful situations. Regarding marital life, when individuals are able to manage themselves in stressful marital situations, their marriages tend to become more stable than individuals who do not have the ability to manage themselves in stressful situations (Ellison 1991). Moreover, Latifa and Amelia (2018) have found that commitment to Islamic values significantly predicts marital stability.
A study by Brown et al. (2008) found that there was a relationship between religiosity and marital stability in white Americans and black Americans. By using the theory of religiosity, it increases the ability to solve problems in difficult situations and stressful situations (Ellison 1991), such as when individuals are able to manage themselves in stressful situations, their marriages tend to be more stable than individuals with no ability to manage themselves in stressful situations. Chinits and Brown (2001) mention that a high level of religiosity will reduce the level of conflict in marriage; in that regard, marital stability will also increase. Research conducted by Lambert and Dollahite (2006), shows that individual religiosity strengthens the desire to maintain a permanent relationship (in other words, affects marital commitment), which in turn helps couples in dealing with occurring conflicts. Sullivan (2001) found that with a high level of religiosity, the level of commitment will increase and marital instability will decrease (increased marital stability). Latifa and Amelia (2018) also found that commitment to religiosity (assessed via religious participation and religious belief), especially for Muslims, significantly predicts marital stability (R 2 = 0.487). Involvement in religious activity can encourage individuals to survive and struggle successfully in difficult situations throughout their marriage life, resulting in a low desire to divorce.
Based on the explanation above, the researcher is interested in examining the influence of religiosity and marital commitment on marital stability in Indonesian female journalists. We questioned whether or not religiosity and marital commitment predict marital stability in Indonesian female journalists. The hypothesis assumes that higher religiosity and higher marital commitment may predict behaviors shown by an individual in the couple interactions and, in consequence, influence the longevity of marital relationship (defined as 'marital stability').

Participants and Sampling Procedures
Participants of this research are 200 female journalists from Indonesia, who were recruited using non-random sampling procedures. Participants were identified through news agencies and journalist communities and registered as formal journalists from news agencies or television companies or online news agencies. The description of demography is listed below: Based on Table 1, the participants' average age is 25 years old, with a mean duration of marriage of around 7 years.

Measurement and Data Analysis
Researchers used the marital stability scale according to Buehlman, Gottman, and Katz interview guidance (this guidance then developed in a scale using Indonesian language by Latifa 2015), the religiosity scale adapted from Huber and Huber (2012), and the marital commitment adapted from Johnson et al. (1999). The validity construct of the variables: Religiosity, marital commitment, and marital stability were analyzed using Confirmatory Factor Analysis (CFA). We use MPlus 8.4 (Muthen&Muthen Los Angeles, CA, USA) software in analyzing the CFA and SPSS v 25 (IBM: New York, NY, USA) for test of hypothesis. a.
Marital stability construct The measurement of the marital stability scale developed by Latifa (2015) was used based on Buehlman et al. (1992) interview guidance, namely the frequency of fondness/affection, we-ness, expansiveness, negativity, disappointment-disillusionment, chaos, and volatility. The scale is a Likert scale with a range of four points, namely: 1 = Never, 2 = Rarely, 3 = Often, and 4 = Always. Latifa (2015) found that the marital stability construct is a one-factor model.
In the test of the construct validity of marital stability, using CFA analysis, the model fit criteria are the chi-square value > 0.05, p-value < 0.01, RMSEA < 0.05, CFI and TLI > 0.90.

b.
Religiosity construct In this study, the religiosity scale was measured with the Centrality of Religiosity Scale (CRS-15) (Huber and Huber 2012), which is based on a multidimensional and interreligious model of religiosity (Huber 2003). The CRS-15 has 15 items and consists of 5 dimensions, namely intellect, ideology, public practice, private practice, and religious experience, which are adapted to the Indonesian language. The translation procedure uses first-round back translation.
In the test of the construct validity of religiosity, we use a multiple-factor model of analysis. Using CFA analysis, the model fit criteria are the chi-square value > 0.05, p-value < 0.01, RMSEA < 0.05, CFI and TLI > 0.90.

c.
Marital commitment construct Marital commitment measured by the Marital Commitment Inventory questionnaire by Johnson et al. (1999). This Marital Commitment Inventory is to measure the personal commitment, moral commitment, and structural commitment that an individual has in his current marriage. The filling of this scale uses a Likert scale with a range of four points, namely: 1 = Strongly Disagree, 2 = Disagree, 3 = Agree, 4 = Strongly Agree. The scale translation procedure uses first-round back translation.
In the test of the construct validity of marital commitment, we use a multiple-factor model of analysis. Using CFA analysis, the model fit criteria are the chi-square value > 0.05, p-value < 0.01, RMSEA < 0.05, CFI and TLI > 0.90.

Marital Stability Scale Validation Result
In the test of the construct validity of marital stability, using CFA analysis, the model did not fit with the chi-square value = 922.642, df = 377, p-value < 0.01, RMSEA = 0.085, so the model was modified until a fit model was obtained (RMSEA < 0.05). As seen in Table 2, after 4-times syntax modification, the chi-square value = 495.292, df = 337, p-value < 0.01, RMSEA = 0.048, CFI and TLI > 0.90, meaning that the model fit. The following is the factor loading coefficient for marital stability items: From Table 3, all 29 items from marital stability scale have a positive estimate (EST/SE: >1.96) and p-value < 0.05 and a positive loading factor coefficient, which means it is significant, so all items in the construct of marital stability meet the fit model criteria (RMSEA < 0.05 after modification) and can be included in the next analysis.  In the test of the construct validity of religiosity, we used a multiple-factor model of analysis. We tested the construct validation using CFA analysis. From the results of the CFA validity test using three models (first order, second order, and five correlated factors), it can be seen in Table 4 that the value of each index fit where the three models did not fit because the p-value < 0.01 and the RSMEA value > 0.05. However, from Table 4, it can be seen that the five-correlated-factor model has a better index than the other two models, namely the lowest RMSEA index (0.070), CFI (0.976), and TLI (0.968). The next step is to make modifications to the five-correlated-factor model, where measurement errors on several items are freed to correlate with each other so that after five-times syntax modification, the following results (Table 5) are obtained. From Table 5, we can see that after obtaining the RSMEA index value of 0.049 (<0.05), the CFI index of 0.989 (>0.090), and the TLI index of 0.984 (>0.090), it can be stated that the five-correlated-factor model is acceptable, meaning that the five dimensions of religiosity are independent but correlated with each other.
From the test results in Table 6 above, it shows that each factor has a strong and positive correlation with other factors. After finding a fit model, we then checked the estimation of each item to determine whether the item was valid or not. This test was done by looking at the z value for each coefficient, as shown in Table 7 below.
Based on Table 7, it can be seen that item number 12 was significant, but the item had a negative estimate (EST/SE: <1.96). Meanwhile the other 14 items had a positive estimate (EST/SE: >1.96), p-value < 0.05, and a positive loading factor coefficient, which means it was significant. The possibility of a negative score of item number 12 is probably because the statement of the item is less representative of the 'Ideology' dimension, especially for Muslims. In the test of the construct validity of religiosity, we used a multiple-factor model of analysis. We tested the construct validation using CFA. From the results of the CFA validity test using three models (first order, second order, and three correlated factors), it can be seen from Table 8 that the value of each index fit where the three models did not fit because the p-value < 0.01 and the RSMEA value > 0.05. However, the 3-correlated-factor model had a better index than the other two models, namely the lowest RMSEA index (0.101), CFI (0.828), and TLI (0.819). The next step was to make modifications to the three-correlated-factor model, where measurement errors on several items are freed to correlate with each other so that after 133-times syntax modification, the following results were obtained (Table 9): Table 9. Modifications of factor model of Marital Commitment. After obtaining the RSMEA index value of 0.050 (<0.05), the CFI index of 0.964 (>0.090), and the TLI index of 0.955 (>0.090), it can be stated that the three-correlated-factor model is acceptable, meaning that the three dimensions of marital commitment are independent but correlated with each other.

INDEX FIT 3-CORRELATED FACTORS Notes
From the test results in Table 10 above, it shows that each factor had a strong and positive correlation with other factors. After finding a fit model, we then checked the estimation of each item to determine whether the item was valid or not. This test was done by looking at the z value for each coefficient, as shown in Table 11 below.  Table 11, it can be seen that there is one item that was not significant, namely item 26 with EST/SE < 1.96 and p-value > 0.05. This item was then eliminated and then the 41 other items had a positive estimate (EST/SE: >1.96) and p-value < 0.05 and a positive loading factor coefficient, which means they were significant. After eliminating item number 26, all items in the construct of marital commitment met the model criteria (RMSEA < 0.05 after modification) and could be included in the next analysis (test of hypothesis).

Test of Hypothesis
The hypothesis test was carried out to determine the effect of each independent variable on the dependent variable with multiple regression analysis techniques. First the researcher looked at the R Square magnitude to find out what percentage (%) the proportion of the variance of the dependent variable was explained by the independent variable, which can be seen in Table 12 as follows. Based on Table 12, it can be seen that the R-Square is 0.224 or 22.4%. This means that the influence of variables of religiosity and marital commitment on marital stability explained 22.4%, while the remaining is influenced by other variables that cannot be explained this study.
In the second step, the researcher analyzed the effect of all independent variables religiosity and marital commitment) on the dependent variable (marital stability). The results of the F test can be seen in Table 13. Based on the results of the F test in Table 13, it can be seen that the p value (sig.) is 0.000 or p = 0.0000 with a p value < 0.05. Thus, there is a significant effect of the independent variables on marital stability. Table 14 explains that all independent variables had a significant effect on marital stability (sig. < 0.05). The direction of the coefficient regression of variable religiosity and marital commitment is positive, indicating that the higher the religiosity and marital commitment, the higher the marital stability. Based on Table 15, it can be seen that the results of the F test (hypothesis test) was 50.248 with sig. 0.000 (sig. < 0.05), so the null hypothesis, which states that there is no significant effect between the variables religiosity and marital commitment to marital stability, is rejected. This means that there is a significant correlation between the religiosity variable and marital commitment on marital stability. From Table 16, the regression coefficient value can be seen in the standardized coefficient column (betha). For marital commitment, the regression coefficient value is 0.431 and the significance is 0.000 (sig. < 0.05). This means that there is a significant effect of marital commitment to marital stability. The standardized coefficient (betha) is positive, meaning that the higher the marital commitment of female journalists, the higher their marital stability. The same thing applies to religiosity, whereby the regression coefficient is 0.147 with a significance of 0.022 (sig. > 0.05); this means that the null hypothesis, which states that there is no significant effect between the religiosity variable on marital stability, is rejected. The value of standardized coefficient (betha), which is positive, means that the higher the religiosity of female journalists, the higher the marital stability.

Discussion
Based on the results of the data analysis described in the results section, there is a significant effect of religiosity and marital commitment on marital stability. The results of the F test, which tested all independent variables on the dependent variable, showed a value of p = 0.000 (significant), and a total variance explained (R 2 value) of 0.224 or 22.4%, meaning that 22.4% of the high or low marital stability is influenced by the independent variable in this study. Table 16 explains that all independent variables have a significant effect on marital stability (sig. < 0.05). The religiosity variable and marital commitment have positive regression coefficients, indicating that the higher the religiosity and marital commitment, the higher the marital stability.
This study can be complementary to the studies by Brown et al. (2008), Chinits and Brown (2001), and Sullivan (2001), about the relationship between religiosity and marital stability. The theory of religiosity, defined by Huber and Huber (2012), describes religiosity as the thoughts and beliefs a person has in viewing the world, known as a personal construct system. The five dimensions, namely intellectual, ideology, public practice, private practice, and religious experience, may increase marital stability. Marital stability is a set of behaviors shown by an individual as the couple interacts and in consequence influence the longevity of marital relationship. When this set of behaviors is religious behavior, individuals (female journalists) will perceive the marriage as more stable.
As Muslims constituted the majority of the participants in this research (95%), Islamic religiosity values might contribute to marital stability. For Muslims, religiosity is usually a guidance in problem solving when facing difficult situations and stressful situations. When individuals are able to manage themselves in stressful situations, their marriages tend to become more stable than individuals who do not have the ability to manage themselves in stressful situations (Ellison 1991). Moreover, Latifa and Amelia (2018) found that commitment to Islamic values significantly predicts marital stability. With commitment on religious teachings, the individual also has the problem-solving ability in a difficult and stressful situation: Be patient and find a way to pray and avoid negative emotions. When individuals are able to manage themselves in a situation like this, it would be more likely to result in a more stable marriage than for individuals who do not have this capability. In addition, participation in religious activities as part of a religious commitment is found as a contributor to the marital stability. Religious activities as a social resource can help improve the welfare of marriage and improve the resilience during marriage life. Socializing with people who have the noble values and norms may lead to modelling behavior that can reflect on commending behaviors to family members at home. Involvement in religious activities may foster the teaching of religious beliefs, so that individuals and couples have positive spirit to face any problems or pressures encountered throughout actual married life, and this principle has to be maintained while believing that there must be solutions to all problems. Whenever the individual participates in religious activities along with her/his partner, it tends to be more stable than someone who just attends alone while her partner never participates in religious activities. Behavioral differences between the individual and their partner can increase the risk of disunity in the marital relationship, for if the individuals are present as a couple, the solidarity may increase and can be a refrain from the distractions of the household, resulting in individuals and couples having developed views, and applying family values together in agreement.
Research conducted by Lambert and Dollahite (2006) shows that individual religiosity gives a desire to maintain a permanent relationship (in other words, affects marital commitment), which then helps couples in dealing with occurring conflicts. Chinits and Brown (2001) also mentioned, a high level of religiosity will reduce the level of conflict in marriage, so that marital stability will also increase. Sullivan (2001) found that with high level of religiosity, the level of commitment will increase and marital instability will decrease (increased marital stability).
The second variable that significantly predict marital stability in this study is marital commitment. Marital commitment using Johnson et al. (1999) theory is defined as a form of the couple's conception of the future of their relationship and their motivation to continue the relationship. A high level of commitment is associated with a number of marital problems that the person can overcome. These conditions are more conducive to the marriage longevity. Therefore, commitment is said to be the main reason for the creation of marital stability (Lauer and Lauer 1986).

Conclusions and Outlook
The research findings give some implications for female journalists, stating that religiosity and marital commitment are the two things that may help marital stability, whereas marital stability can contribute to psychological well-being and their career development. As Suleyiman and Zewdu (2018) stated, a stable marriage is when the partner enjoys a healthy relationship; one where the partner is a source of emotional support, friendship, sexual satisfaction, and financial support for one another.
For future research, path analysis between religiosity and marital commitment to marital stability should be taken into account. Adding another variable that might give a bigger contribution to predict marital stability in female journalists is also a recommended procedure for future research. Another possibility is explicating the same variables for other professions that vulnerably influence marital stability, such as being a nurse, celebrity, or policewoman. Institutional Review Board Statement: All subjects gave their informed consent for inclusion before they participated in the study.

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study. There are no anticipated risks associated with the study. Information of this study is anonymous, meaning that no names or identifying data will be collected. Survey data collected will be maintained and kept for research or educational purpose only.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.