To us it is remarkable that so many couples would end their relationship not long after the spouses professed their unending love for one another, often at great expense and in front of family and friends. How does this shift happen? Can we predict which couples will run into trouble and which will succeed? What factors are most important to consider in understanding how marriages deteriorate and change?

As you may know, about half of all first marriages end in divorce or permanent separation. Many of these divorces and separations happen in the first five years after the wedding. The UCLA Marriage and Adult Development Study was conducted to answer one basic question: why do so many marriages end or become dissatisfying? We choose to address this question for two reasons. First, it presents an interesting scientific puzzle. Second, and more importantly, we believe that working to answer this question will put us in a better position to enable future generations of couples to have stronger and more fulfilling relationships.

The scientific literature includes many studies of marriage, and we seek to draw from and extend this literature. In 1995 two members of the research team, Benjamin Karney and Thomas Bradbury, published an article that summarized all 115 known longitudinal studies of marriage - representing more than 45,000 couples. Four of the most important conclusions they reached concerned the methods that researchers typically used in longitudinal studies. It is valuable to consider these four conclusions because they limit what many of these studies can tell us. These conclusions also point directly to the procedures we followed in this study:

First, Karney and Bradbury observed that in many longitudinal studies the participating couples were, on average, several years along in their marriage when the study began. It is interesting to know, for example, how marriages change between year 5 and year 6, but this may not tell us much about how couples negotiate the first few years of marriage that often prove to be so difficult. And most studies of couples who have been married for 5 years or so will overlook the many other marriages that started at the same time but ended in divorce before they could be studied. We believe that this is a very important oversight, because these divorcing couples might be of great relevance to developing intervention programs.

To address this concern, we began this study with newlywed couples and we recruited them to participate in the study shortly after marriage. We also focused our attention on people who had not been married before, because there is evidence that first and subsequent marriages differ in important ways.

 

A second limitation of many prior studies is that they tend to examine change over relatively short periods of time. Again, this can be useful, but studying marriages over only one or two years may not yield enough information to understand fully how they change and evolve over significant spans of the marital career. In our first study we started with 60 couples and examined them over as many as 4 years.

In this study, we started with 172 newlywed couples and we now plan to examine as many of them as possible over 12 years.

 

A third concern, related closely to the second, is that most longitudinal studies evaluate marriages at just two points in time. This has two unfortunate consequences. First, it provides a relatively imprecise picture of how marriages fluctuate and change. Information will be lost with this approach, and it can make couples look better or worse than they might be otherwise. Second, having just two assessments of marriage forces researchers to use somewhat uninformative statistical approaches, which do not capture the process of change in marriage very well.

To address these concerns, we attempt to collect data twice a year from each couple. We believe that this represents the marriage better, and it allows us to use statistical tools that reflect the ebb and flow of relationships.

 

Reading over all 115 of these articles led Karney and Bradbury to a fourth concern about the published research: the vast majority of studies examined just one variable (or one set of conceptually similar variables, such as different indices of personality) as a predictor of how well marriages functioned over time. This approach holds some merit, as it can be useful to know just how important one variable (like personality) is in understanding the course of marriage. However, knowing how important one variable is relative to some other variable is almost always more informative. And most scholars would probably agree that the outcome of a marriage is determined by more than one variable or one set of closely related variables. This concern was addressed by looking over all of the variables in the published literature that appeared to be influential in marriage. Combined with a review of existing theories, this led Karney and Bradbury to three main domains of influence in explaining how marriages succeed and fail:

Enduring Strengths and Vulnerabilities, or the personal characteristics, experiences, and histories that a spouse would bring to any marriage,

Adaptive Processes, or the ways that couples contend with differences of opinion, differing goals, and individual or marital difficulties and transitions, and

Stressful Events, or the developmental transitions, situations, incidents, and chronic and acute circumstances that spouses and couples encounter.

To address the concern that previous studies were focusing on just a few variables thought to be important in determining the course of marriage, we collected data in all three of these domains - enduring strengths and vulnerabilities, adaptive processes, and stressful events. We tried to collect data in these domains more than once and using different kinds of methods, including questionnaires, interviews, and direct videotaped observation of couples.

So that is our study in a nutshell: it is an effort to collect data from newlywed couples, over a significant span of time in their relationship, at frequent intervals, capturing as best we could who the individual spouses were prior to marriage, how they interact with one another following marriage, and the stresses and strains they encounter along the way. With these data in hand, we believe we will get closer to resolving the mystery of why marriages fail.

We do not mean to imply that this study is flawless, or that all studies have to be like ours to be informative. (In fact, we ourselves have published papers using established couples, short spans of time, two periods of data collection, and a small set of variables.) However, we do believe that this study incorporates many, perhaps even most, of the features needed to advance our understanding of how marriages develop, change, and deteriorate.

 

In an earlier study, we recruited couples from newspaper and radio advertisements. This worked reasonably well, but we could not be sure how many people learned about the study and decided to not participate. It is also hard to tell how the people who respond to advertisements differ from those who do not respond. This is important, because in our research the goal is to make discoveries that generalize to many couples, not just those who are studied. If an unusual group of couples is studied, then the findings may apply only to them.

With these concerns in mind, we recruited couples for this study by examining the marriage licenses filed in Los Angeles County between May 1993 and January 1994. We gained special permission from the county to have access to the licenses, and several times over the course of several months we sent pairs of research assistants - and a portable computer -- to the county office to scan the microfilmed records. The marriage licenses included, for both spouses, their address, age, years of education, occupation, and number of previous marriages. We transcribed the addresses for those people who were older than 18 years of age, wives who were 35 years old or less (to allow the possibility that all couples might become parents over the course of the study), people who had at least a tenth grade education (to ensure that they could understand our questionnaires), and people who were entering their first marriage. All other licenses were ignored.

We mailed letters to the addresses on the licenses, inviting people to earn $75 for participating in a longitudinal study of newlywed marriage. A postage-paid post card was included with each letter, which couples could complete (with names and telephone numbers) and mail back to the lab if they were interested in learning more about the project.

Of the 3,606 letters that we sent out, 637 couples (17.8%) expressed interest in participating, 41 letters were returned as undeliverable (1.1%), and 2,928 letters (81.2%) went unanswered. This response rate is similar to that of other studies recruiting married couples from public records. It becomes important to know, of course, how the people who responded compare to those who did not respond. Our analyses showed that the respondents were more likely to cohabit premaritally (42.9% vs. 35.3%), had more years of education (wives: 15.4 vs. 14.5; husbands: 15.2 vs. 14.6), and the wives were older (26.6 years vs. 26.2 years); respondents also had higher status jobs. These are noteworthy differences, but they are probably small enough to indicate that the couples who responded were not atypical or unusual from the larger population. We published an article that discusses in detail this recruitment procedure.

A member of our research team then telephoned those couples who returned the postcards, and the assistant then interviewed one of the spouses. In this interview we verified the information on the marriage licenses, and we excluded from further consideration those people who already had children, couples in which at least one partner could not speak English, and couples who had plans to move from the area. Ideally we would have included people with a child and people who were non-English speakers, but we did not because (a) we were interested in knowing how couples would adjust to the arrival of their first child and (b) we did not have funds to hire non-English speakers who could interview people in the lab and code their videotapes.

The first 172 couples who met all of our eligibility criteria and kept their laboratory appointment formed the sample. All couples were married less than 6 months when they began the study.

 

The women in the study were 26.0 years of age and had 16.2 years of education, on average. Their median annual income was somewhere between $11,000 and $20,000. Sixty-one percent were Caucasian, 15% were Asian American/Pacific Islander, 5% were African American, 16% were Latina/Chicana, 2% were Middle Eastern, and 1% identified as "other". Most of the women were employed (85%), and they were most likely to work in a semi-professional job (e.g., technician, travel agent, customer service representative). Forty-seven percent listed their religion as Protestant, 26% as Catholic, 5% as Jewish, 3% as Mormon, 17% as None, and 3% as Other.

The men averaged 27.6 years of age and 15.6 years of education. Their median annual income ranged from $21,000 to $30,000. Sixty-seven percent were Caucasian, 13% were Asian American/Pacific Islander, 4% were African American, 15% were Latino/Chicano, and 1% were Middle Eastern. Most husbands were employed (91%), and they were most likely to work in some kind of semi-professional job. Forty-one percent listed their religion as Protestant, 31% as Catholic, 5% as Jewish, 2% as Mormon, 19% as None, and 2% as Other.

The vast majority of participants in most prior longitudinal studies of marriage are Caucasian. We see our inclusion of people who identify themselves as ethnic minorities - 39% of the wives and 33% of the husbands - as a very strong feature of this project.

We know that this sample is above average in education: 36% of the sample has completed college, compared to 25% of California residents and 22% of U.S. residents. We also know that our couples were married 1-2 years later than the average man and woman, for whom the national norms in 1994 were 24.4 and 26.7, respectively. And we know that our participants had similar (for men) or slightly lower (for women) personal incomes compared to others in California (where the median personal income was $22,915 in 1990). Our sample of couples is very representative of the proportion of Catholic, Jewish, and Mormon people in California but it may over-represent Protestant people slightly.

So what is the bottom line? Our sample does not seem to be unusual on any particular dimension for newlywed couples. As a result, we believe that that the people participating in this study are reasonably representative of the population of newlyweds in California and that our findings can be generalized in a meaningful way to the larger population of newlyweds in the United States.

What else do we know about our couples? As you might expect, spouses were very satisfied with their marriage at the beginning of the study, according to standard measures of marital satisfaction. They reported few symptoms of depression at this time, though we know from interviewing spouses in our first lab session that 22% of the men and 29% of the women reached a diagnosable level of depression at some time in their life. (By comparison, the lifetime prevalence for depression in the US is about 17%.) We also know from these interviews that significant numbers of participants also had diagnosable or nearly diagnosable Alcohol Abuse (21%) or Substance Abuse (17%). About half (46%) of the participants had a parent who had divorced.

If you participated in this study, you know that we collected all kinds of information about you and your marriage in a variety of ways - we administered questionnaires, we interviewed you, and we videotaped you and your partner as you talked about different aspects of your relationship.

As it turns out, the questions we asked tend to fall into three main categories. First, we wanted to find out about your personality, your experiences while you were growing up, and the kinds of dating relationships you had before you met your spouse. We think of these as Enduring Strengths and Vulnerabilities, or the personal characteristics, experiences, and histories that you and your spouse would bring to any marriage, no matter whom you married. Second, we wanted to know how you and your spouse talk to each other, especially while you were discussing problem areas in your relationship or things you might want to change about yourself. We use this information as a way of gauging couples' Adaptive Processes, or how they work together on differences of opinion, differing goals spouses might have, and the stresses and strains that life sends our way. And finally, the third area of interest to us was the various Stressful Events that you might have encountered as individual spouses and as a couple. We know that individuals and couples vary widely in the kinds of developmental transitions and chronic and acute circumstances that they encounter, and we wanted to know more about these for each participant.

Basically, we are curious to know how these three sets of factors operate - either alone or in combination with one another - to produce different kinds of marriages. Read More

Three other topics were of great interest to us...

The first, and most obvious, were spouses' perceptions of how satisfied they were with their marriage. We also asked about whether spouses were thinking about separating from their partner or whether they were contemplating divorce. These questions were asked on every questionnaire, and they help to give us a picture of how marriages change over time.

Second, we administered a special set of questions once before and twice after the arrival of a couple's first child. This helped us to know about people's expectations for becoming a parent, the effects of the child on the marriage, and who was involved in various aspects of caring for the new baby.

Third, we were curious to know about couples' experiences with premarital counseling. Because one of our key goals is to use information collected in the project to develop a program that will benefit future generations of couples, we used this opportunity to find out about the kinds of programs that couples normally received in their community and how much couples valued them.

 

Between 1993 and 1998 we studied the marriages of 172 newlywed couples. We collected a lot of information from them, and we learned a great deal as a result. In 2001 we initiated a new project to study these couples and, if they became parents, their children as well. We anticipate continuing this data collection through 2005.

We are continuing this study for three main reasons.

First, we will be examining how marriages develop over a longer span of time. The factors that make a marriage satisfying in the third year may well differ from the factors that are beneficial to a marriage in the eighth year. The new data will allow us to study topics such as this, and the information we gather will allow us to develop workshops for couples at different points in their marital career.

Second, although numerous shorter-term longitudinal studies of newlyweds have been conducted, few longer-term longitudinal studies are available in the scientific literature. We therefore know relatively little about the various challenges that crop up in marriage across time, how they are linked to prior experiences in the marriage, and how they might affect the marriage in the years that follow.

Third, many couples now have young children, and we are eager to learn how earlier development in the marriage is related to how parents raise their children, and how the children themselves grow and change.

The bottom line is this: we believe that a great deal has been invested in this project - by our laboratory staff and by the participating couples themselves - and we believe that many more questions about marriages and families remain to be answered. The information that couples have already provided to us serves as a rare and valuable resource, and as a foundation for learning much more about what makes marriages and families work.