Contribution of marital conflict to marital quality in short and long-term marriages: An actor-partner interdependence model.

AIMS: In the field of family research, previous studies have made great strides toward understanding the relationship between marital conflict and quality. However, they have only studied couples in short-term marriages. Therefore, much remains to be unraveled with regard to long-term marriages. We aimed investigate the comparative contribution of aspects of marital conflict to marital quality in short-and long-term marriages in Iranian families.
MATERIALS AND METHODS: Using random clustered sampling, 400 dyads in intact first marriages were surveyed across eight provinces of Iran. Complete surveys for both husbands and wives were returned for 162 households (couple's response rate: 40.5%). Survey measures included demographics questionnaire, Barati and Sanai's Marital Conflict Questionnaire and Blum and Mehrabian's Comprehensive Marital Satisfaction Scale. Structural equation modeling was used to test the actor-partner interdependence model of marital conflict-marital quality.
RESULTS: Generalized additive models were incorporated to define what constitutes short-and long-term marriages. Based on the models regressed, duration ≤ 10 years was defined as short-term, whereas duration ≥ 25 years was labeled long-term. In short-term marriages (n = 44), decreased sexual relations, increased daily hassles and sidedness in relations with parents were negatively associated with marital quality in both actor and actor-to-partner paths. In long-term married couples (n = 46), only increased daily hassles (P < 0.001) and disagreement over financial affairs (P = 0.005) contributed to actor paths and only sidedness in relationships with parents showed significant negative association to marital quality in actor-to-partner paths.
CONCLUSIONS: Different themes of conflict contribute to the diminished level of marital quality in early and late stages of the marriage. Conflicts over sex, relationship with extended family and daily hassles are emphasized in the early years of marriage, while in later years; dispute over money and daily hassles dominate.

INTRODUCTION

Early research in the field of marriage and family structure highlights the role of marital quality on an individual's global happiness and provides support to the notion that an unsatisfactory marital relationship adversely influences both physical and mental health of the individual.[1234] Kiecolt-Glaser and Newton, in their meta-analysis of 64 articles, concluded that diminished marital quality is not only associated with a diverse range of physical and mental health problems, but is probably the most important indicator of global happiness as well, often surpass the effect sizes reported for work and social network-related satisfaction.[5]

Although previous studies have made great strides toward understanding the interrelationship between marital conflict and quality, they have often included couples in the early stages of the marriage. Therefore, much remains to be unraveled with regard to long-term marriages. For instance, it is not known whether similar themes of marital conflict contribute to diminished marital satisfaction, or they tend to change as marriage advances over time. Answering such questions has significant implications for our perception of how long-term marriages function and what supporting/interfering forces to this dynamic are.

Gottman,[6] based on the trajectory derived from experimental data, suggests that couples are especially prone to divorce in the first 0-7 years and then 16-24 years of marriage. An enduring marriage therefore could be defined as an intact marriage that has lasted more than 24 years; the probability of divorce in such relationship is low. It should be noted however, the curvilinear path of marital quality does not ensue an identical pattern across populations and might indeed be dependent upon the sample under study. Only few authors have cast doubt on the very existence a curvilinear pattern between marital duration and marital quality;[78] yet the shape of the curve has been an area of heated debate. The majority of previous research has argued that marital quality assumes a U-shaped pattern with decrements in the first half, followed by increments in the second half of the married life.[91011] More recently, VanLaningham, et al., using the US national data of married couples under 55 years collected over five waves in 1980, 1983, 1988, 1992 and 1997 observed that although the association of marital duration with marital quality is curvilinear, its shape more likely resembles a declining spline rather than the commonly narrated U-shape.[12] When only the last three waves were included in the analysis, the results yielded a consistent declining pattern with apparent flattening of the trajectory in the middle years of married life.[12]

Given the paucity of available data on the contribution of marital conflict to marital quality in long-term marriages, especially in the context of a collectivist background, we aimed to conduct the present study.

MATERIALS AND METHODS

Methods

Participants and application

For the purpose of this study, a sample size of 160 couples in intact marriages with marital duration between 2 and 38 years was used to draw the trajectory of marital quality over marital duration. Cut-points for marital duration were extracted from critical time points and dyads were categorized as being in short- or long-term marriages accordingly. Then, using structural equation modeling, the relative contribution of four essential themes in marital conflict to marital quality was investigated and compared between short- and long-term married couples. We hypothesized that although similar declining pattern of marital quality might be in effect in both short- and long-term marriages, the relative contribution of different areas of conflict to marital quality is likely to be distinct for each period. Four major areas of conflict including money, relationship with parents and in-laws, daily hassles and sexual relations were evaluated. These four areas, although do not capture the entirety of marital conflict, are nevertheless significant areas of distress for the married individual and each has been shown to significantly affect marital quality in the published literature.[1314151617]

A cluster sampling scheme was developed to randomly select households across seven provinces in Iran. These seven provinces were chosen to reflect the economic, social and cultural diversity of the Iranian society. The survey was conducted between February and May 2011 and included 400 households. In the first step, the range of postal codes in urban areas of selected provinces was fed into a software program (MATLAB 6.1, The MathWorks Inc., Natick, MA). Except for Tehran - the capital of Iran-where two clusters were assigned, one cluster of 50 households was assigned for each province. In the second step of sampling, households within each cluster were visited and a sealed package containing survey questionnaires, a brief guideline on how to fill each form and a consent form describing the purpose and nature of the study were delivered. Neighboring households to the right of the index household were then approached and research team delivered the packages. Sampling was continued until each cluster consisted of 50 households. Households were not enrolled if no one was available at the time of the visit, the primary household resident was not married, was under 18 years of age, had been divorced, or his/her spouse was deceased. Each package included two sets of questionnaires, one for the wife and one for the husband. The accompanying guideline specifically asked the participants to complete the forms at the comfort of their homes, but independently of each other. Delivered packages were collected the next day at a time convenient for the household residents. 342 households returned the completed questionnaires for at least one spouse indicating individual response rate of 84%. Complete forms for both husbands and wives accompanied by signed consent forms were available for 162 couples, indicating couple response rate of 40.5%. Ethics committee of the Behavioral Sciences Research Center of the Baqiyatallah University of Medical Sciences approved the study protocol.

Materials

Surveys used in the present research include an information form collecting data regarding demographic characteristics of the participants, Barati and Sanai's marital conflict questionnaire (MCQ) and Blum and Mehrabian's Comprehensive Marital Satisfaction Scale (CMSS).

Information form

In this portion of the survey, participant's date of birth, age at getting married, duration of marriage, number of children, highest level of education attained, perceived level of socio-economic status and current job status were recorded. Duration of marriage was averaged if the response provided by husband and wife differed by up to 3 years. If the difference was larger than 3 years, the response deemed unreliable and was recorded as missing. Participants were asked to categorize their socio-economic status as low, middle, or high. If each spouse provided different answers, the highest given level was recorded for the couple. Participants classified themselves in one of the four categories of job status: Employed (either part-time or full-time), retired, unemployed and homemaker (in which the participant identified her principal duty as keeping the house and taking care of children).

MCQ

Based on their experience on marital discord among Iranian couples, Barati and Sanai developed a 42-item self-administered questionnaire.[18] MCQ is specifically tailored to capture areas of conflict that are overlooked in western models of marital conflict. Domains used for the purpose of the present study include decreased sexual relations (items 4, 10, 15, 28, 33), increased daily hassles (items 5, 11, 16, 22, 29, 34, 39, 41), sidedness in relations with extended family (items 1, 6, 12, 17, 19, 23, 25, 30, 35, 37, 40, 42) and disagreements over financial affairs (items 2, 8, 13, 20, 26, 32, 38). Increased frequency of daily hassles includes items that inquire about the frequency of disagreements and struggles about daily activities and events as well as the frequency of engaging in negative methods for solving them (i.e. withdrawal and heightened emotional response). Sidedness in relations with extended family members involves increased relationship with mother and father and at the same time decreased relationship with in-laws.

Each statement on the 42-item list is rated using a five-item likert-type response scale and respondent's score can range from a minimum of 42 to the maximum of 210. Greater scores indicate a higher level of marital conflict. A good level of internal reliability for the questionnaire has been reported (Cronbach's alpha = 0.74).[19] Cronbach's alpha for the domains were: 0.71 for decreased sexual relations, 0.76 for increased daily hassles, 0.78 for sidedness in relations with extended family and 0.71 for disagreements over financial affairs.[19]

CMSS

CMSS, developed by Blum and Mehrabian (1999), is a 35-item self-administered questionnaire that records self-perception of each partner regarding the intimate interaction and general satisfaction in the married life.[20] Respondents declare their level of agreement to each statement using a nine-level scale that ranges from −4 (very strong disagreement), to 0 (neutral), to +4 (very strong agreement). Scores from individual statements are then added up to calculate a final score. In the present study, a translated version of the questionnaire was used.[21] The Persian version of the CMSS was validated in a sample of 200 married individuals and an excellent level of internal consistency (Cronbach's alpha = 0.90) has been documented.[21] In addition, validity of a two-factor (marriage satisfaction, marriage dissatisfaction) structure model with χ2 = 77.94, degrees of freedom (Df) =61, P = 0.071, comparative fit index (CFI) =0.98 and root mean square error of approximation (RMSEA) =0.05 has been demonstrated.[21]

Statistical analysis

Descriptive analyses for dyads in short-and long-term marriages were generated using Statistical Package for Social Sciences version 17.0 (SPSS Inc., Chicago, Illinois). Generalized additive models (GAM) for describing trajectory of marital quality over marital duration were constructed using STATISTICA version 8.0 (StatSoft Inc., Tulsa, Oklahoma). Structural equation modeling (SEM) was performed using Amos version 18.0 (Amos Development Corporation, Crawfordville, Florida).

GAM

GAMs are an extension of generalized linear models that allow semi-parametric smooth functions to be fitted on data using splines.[22] These data-driven models produce curves that in comparison with linear approaches, better capture the true non-linear structure of the sampled data.[22] For the purpose of this study, possible contribution of marital duration to marital quality (assessed by CMSS) was investigated using generalized additive regression splines, separately for husbands and wives. Normal distribution was assumed for marital quality and the log function was considered as the link. Df for each model were automatically calculated according to the generalized cross validation rule.

Actor-partner interdependence model

The actor-partner interdependence model is a model used for analysis of dyadic data where observations in one individual are dependent upon and correlate with the other.[23] The interdependence model provides an alternative to the traditional approaches where codependence of data is usually overlooked.[23] In fact, the premise of interdependence contrasts with the assumption of parametric statistical techniques (e.g. regression and analysis of variance) where independence of data is indivisible.[24]

Using the actor-partner interdependence model, we aimed to investigate whether different domains of marital conflict in husbands and wives affect marital quality in both. Accordingly, domains of marital conflict were directed toward marital quality for both actor and partner. Equality constraints were placed for actor and partner paths. Residual variances were assumed for dependent variables and were constrained to be equal.

For each model, fit indices including χ2, P value for χ2, Df, CFI and RMSEA (90% confidence interval) were calculated. In addition, for actor and actor-to-partner paths, standardized regression coefficients were calculated and t-test for statistical significance was performed. χ2 tests whether the fitted model is significantly different from the perfect fit; hence greater values indicate more prominent deviation. For CFI and RMSEA acceptable fit was delineated using the cut-off points provided by Kline.[25] According to Kline's recommendations, CFI > 0.90 and RMSEA ≤ 0.08 indicate acceptable fit for the model.[25]

RESULTS

Marital duration and marital quality

Data for the enrolled 162 couples was fitted on GAM for husbands and wives, separately and the derived graphs are illustrated in Figure 1. For husbands, the fitted spline clearly demonstrated a curvilinear pattern (non-linear P = 0.008, calculated Df = 4.01, GAM coefficient = 0.0004). As evident in Figure 1 (top panel), for husbands, marital quality declined over first 8-10 years of marriage. The middle years (10-25 years) enjoyed a reversal pattern, while in the later years (25-38 years) a sudden decline in the marital quality was observed. A similar pattern was identified in wives (non-linear P = 0.013, calculated Df = 4.02, GAM coefficient = 0.0004). As demonstrated in Figure 1 (bottom panel), wives experience a descend in marital quality during the initial years of marriage, then the quality remains relatively stagnant with a slow tendency toward improvement until 22-25 years and finally starts to sharply decline towards the end. However, the overall gradient of change is less pronounced in wives compared with husbands. It appears that marital quality of wives is less influenced by marital duration and remains more stable over time. The cut-offs for definition of short-and long-term marriages were inferred from splines. In both sexes, the curvilinear pattern showed two distinct critical time periods when the direction of the marital quality trajectory alternates. These two critical points (first at 10 years and second at 25 years) partition the observed trajectories into three different periods of decline/rise-stability/decline. According to this devised model, short-term marriage is defined marital duration ≤10 years and long-term as marital duration ≥25 years.

Figure 1

Generalized additive models for marriage duration and marital quality in husbands (top panel), and wives (bottom panel). A complex and significant curvilinear pattern in both models is identifiable

Description of analyzed dyads

Baseline characteristics of couples in short-and long-term marriages are presented in Table 1. Short- and long-term marriage groups encompassed of 44 and 46 couples, respectively. Couples in short-term marriages were married for an average of 7.5 years and had at least one child. More than 90% perceived themselves as being middle or high on the socio-economic status scale. In short-term marriages, husbands were on average 5 years older than wives. 49% of couples had university degrees and another 43% completed high-school. Husbands were categorized as either employed (93%), or unemployed (7%). About 47% of wives were in the labor force, another 47% were homemakers and the remaining 6% regarded themselves as unemployed.

Table 1

Baseline characteristics of couples and individuals in short-term and enduring marriages

Couples in long-term marriages were in an enduring relationship for an average of 26.5 years. Except for one couple who had only one child, others had at least two children. In addition, all but one couple perceived themselves as being middle or high in the socio-economic status spectrum. Husbands were on average 4.5 years older and more than 90% at least had completed high school. Two-thirds of the husbands were still in the labor force, less than one-third were retired and the remaining 6% were categorized as unemployed. Wives on the other hand, were less likely to have a university education and only 24% were in the paid labor force [Table 1].

Marital conflict and marital quality

The results of actor-partner interdependence model for short-and long-term marriages are depicted in Tables 2 and 3. The corresponding SEM model is also demonstrated in Figure 2. Goodness of fit indices, generated for each group separately, showed excellent fit of the model. Of note, despite acceptable fit indices in both groups, the generated parameters for short-term marriages were slightly more desirable. Regarding actor paths in short-term marriages, regression coefficients for three areas of conflict, namely decreased sexual relations, increased daily hassles and sidedness in relationship with parents significantly and negatively associated with marital quality. Actor-to-partner paths also showed that the same three indicators in actors contribute to decreased satisfaction in partners. In both series of coefficients, the largest observed value was for sidedness in relationship with parents (−0.239 for actor path and − 0.309 for an actor to partner path). In long-term marriages, two domains significantly linked to diminished marital quality were increased daily hassles and disagreement over financial affairs. However, actor-to-partner path indicated that only sidedness in relations with parents contributes to partner's decreased marital quality.

Table 2

Goodness-of-fit indices for actor-partner interdependence models in short-and long-term marriages

Table 3

Standardized path coefficients for actor-partner interdependence models in short-and long-term marriages

Figure 2

Actor-partner interdependence model for dyads in short- and long-term marriages

DISCUSSION

Two limitations of the present study deserve mention. First, the cross sectional nature of our study precludes inferring causal relationships between marital duration and marital quality. Delineating trajectory of marital quality over the life course requires that couples are followed from the beginning of married life until the end. These follow-ups usually necessitate collecting information via structured interviews conducted at multiple points in time. Snapshots taken at different times over the life course could be brought together to form an integrated path capable of demonstrating the extent of change over the married life. However, collection of these data is not possible unless an extraordinary amount of resources is available. Moreover, conduct of such studies is extremely time-consuming often exceeding the time allocated for a specific research project and the research life of the scientist itself. In developing countries, such as the country we live in (Iran), these predicaments are further intensified. In the absence of such cohorts, cross-sectional data provide invaluable insights into the complex dynamics of marital quality and the forces shaping it.

The second limitation relates to the shortcomings in the enrollment of divorced/separated couples. Studies investigating marital conflict and satisfaction are to include couples who have broken marital ties since these individuals are often the ones with the least satisfaction in married life.[26] In individualist cultures, about two-thirds of all first marriages end in separation or divorce.[27] Failure to include unhappy couples would possibly skew the results towards a diluted level of marital conflict and at the same time an augmented level of marital quality. In the present study, our sampling scheme was tailored to include only couples in intact first marriages and no divorced couples were included. It should be noted however, the percentage of couples getting divorced in Iran is significantly lower than the observed numbers in western societies where individualism dominates. In 2010, marriage-to-divorce-ratio in Iran has been reported to be 6.1/1000; about 3 times higher than the reports generated from 44 US states (1.9/1000).[2829] Despite significant transformation over the past decades, the collectivist tradition of Iranian culture still strongly values marriage over divorce and as a result the marriage to divorce ratio of the country is more than 3 times higher than the United States.[2930] Hence, although lack of divorced individuals in the studied sample poses and important limitation with regard to interpretation of the results, we believe that the magnitude of the imposed bias is indeed quite finite.

Notwithstanding the mentioned limitations, our results were able to closely resemble the curvilinear pattern identified in the longitudinal study of US married couples (1988-1997 waves).[12] VanLaningham, et al. presented solid evidence that marital quality consistently declines over the life course, although flattens after about 18-20 years into the married life.[12]

Our findings demonstrate that in both husbands and wives three splines can be distinguished over the marital course. The first spline corresponding to the first 8-10 years of marriage is characterized by a rapid decline in marital quality. After reaching a nadir at this critical point, the curve takes an upturn that is more evident among husbands. The third critical point occurs at around 23-25 years where marital quality again, begins to decline and falls to levels lower than ever. Our study was not specifically designed to determine the relative contribution of forces shaping the trajectory of marital quality over time. However, using the breadth of available research in the area, a number of inferences could be drawn describing why the curvilinear pattern is the way it is. A number of studies have highlighted the effect of having children on the initial decline in marital quality and reasoned that parenting is a stressful experience demanding an enormous amount of time and energy from the dyad.[313233] Resources previously available for mutual caring and reciprocity are directed towards the newborn, which causes turbulence in the otherwise stable married bond.[1132] Although the effect of advent of children on the first wave of decline observed in short-term marriages is understandable, it is not clear if and why subtraction of children from home is associated with a further decline about 15-18 years later. Based on their experience with US couples, VanLaningham, et al. suggested that departures of children from home produces an ignorable effect on marital quality and is transient at best.[12] We add to this argument that family bonds in the collectivist cultures are stronger than those in individualist ones and hence departure of children might in fact be a detrimental event to the family structure.[34]

It is conceivable that external factors namely retirement and leaving the labor force also contribute to the second wave of decline.[113536] In our study, about 30% of husbands in enduring relationships have had entered retirement, whereas the rate for wives was only 4%. Indeed the second wave of quality decline was more precipitous in husbands compared with wives. It remains open to question however, why an event like retirement would propel marital quality downward and exerts negative effects on the married life. Previous research has suggested that retiring, especially from high-stress jobs is associated with improved marital quality.[36] It has also been proposed that the retirement of one spouse while the other is still considered working might be detrimental to marital quality.[37] When both husband and wife enter retirement concurrently, the effect is inverse and the marital satisfaction benefits.[37]

Using the actor-partner interdependence model, we have shown that various domains of conflict negatively affect the quality of married life of not only self, but also the partner. This is particularly important since it reveals that conflicting experiences are unquestionably shared and even if exert limited effects on the actor, may inadvertently be conceived hurtful by the partner. For instance, in short-term married couples sidedness in relationship with parents negatively associates with marital quality in both actors and partners. Nonetheless, the regression coefficient for actor-to-partner path is relatively larger, suggesting the partner is more disturbed by the act, than the actor him/herself.

Daily hassle was the only theme that affected both short- and long-term married couples. The path coefficient grew larger in enduring relationships, albeit at the same time actor-to-partner effect rendered non-significant. Congruent with our findings, in a sample of elderly couples aged 55-75 years, daily hassles has been implicated in decreased marital quality and that this association is mediated through intimacy level of the couple.[13] Daily arguments over in and out of the house chores tend to be repetitive, stressful and in the long run, significantly frustrating.[1338] The fact that hassles continued to negatively affect both short-and long-term marriages supports the notion that these struggles often stem from personality and mental traits that are formed long before marriage and tend to be relatively stable throughout the married life.[3839]

Decreased sexual relationship affected both actor and partner in only short-term marriages while disagreements over money posed as an important indicator of diminished marital quality only in couples married for longer periods. Finally, favoring parents over in-laws was an important determinant of partner marital quality in both short- and long-term marriages.

In line with our findings, Byers in an analysis of 87 individuals in long-term relationships found little evidence supporting a causal link between sexual and marital satisfaction.[40] On the other hand, sexual satisfaction was associated with marital quality in couples married for one to 3 years, confirming that the importance of sexual relationship might decline with advancement of marriage.[41]

It is hypothesized that in the context of collectivist cultures, extended family and most importantly parents and in-laws retain a significant role throughout dyad's married life. Favoring relationship with parents over in-laws implies that the process of integrating new members into the extended kinship has not successfully occurred. Since in-laws of one spouse are parents to the other, one-sided relations with parents could be upsetting and become a source of marital conflict. Inversely, since members of the extended kin are often deeply involved in conflicts of the married couple, an unbalanced interference in the conflict could weaken a previously established a bond between spouse and in-laws, directing it towards an unbalanced relationship with parents. Familism is an indispensible part of family structure and values within collectivist cultures. Contrary to the individualist tradition where one assumes responsibility and depends on self, the familistic perspective assumes equal priority for the members of immediate and extended family as well as the person him/herself.[34] Members of the extended family and in particular, parents and in-laws form an inherent part of individual's social network which can affect marital quality in a multi-faceted manner.[16] Parents and in-laws of the married couple could provide either support (by helping in the time of need and adding to the resource available for conflict resolution), or interfere (by fostering reprehension and siding with one spouse in the struggles) in the married life.[42]

Papp, et al. argue that money is an inherent domain of conflict in marriages and perhaps the most important of them.[14] According to their assessment, disagreements over financial affairs are more pervasive, repetitive, threatening to marital functioning, associated with negative emotions and at the end tend to remain unresolved.[14] Previous literature also lends support to the notion that money arguments are indeed an essential contributor to marital conflict and might include disagreements over who spends the money, how money is spent, or how assets/debts are shared.[154344] Assets tend to accumulate with advancing age and financial responsibilities toward children is relaxed as they grow older and become independent.[15] Therefore, it is expected from the hypothesis that short-term marriages be bothered by money conflicts more than their long-married counterparts. What caught us by surprise was the observation that disagreements over finances emerged as a major theme of conflict in long-but not short-term marriages. These findings were recorded, in spite of the fact that couples in both short- and long-term marriages reported similar ratings of their socio-economic status.

CONCLUSION

Our findings underscore the importance of marital conflict in diminishing marital quality in both short- and long-term marriages. With taking advantage of a comparative approach, we propose that although the quality of marriage indeed consistently declines in both marriages of short- and long-duration, different themes of marital conflict contribute to this trajectory at each period. In particular while conflicts over sex, relationship with extended family and daily hassles are emphasized in the early years of marriage, areas of conflict in couples in enduring marriages are relatively limited and include money and daily hassles. Future longitudinal studies need to focus on elucidating the factors and events that contribute to this transition.