Keeping Marriages Healthy, and Why Its So

Difficult
By Benjamin R. Karney

Benjamin Karney is an Associate Professor of Social Psychology and co-director of the

Relationship Institute at the University of California, Los Angeles. His research focuses on how marriages change or
remain stable over time, and in particular how relationship maintenance is constrained or enhanced by the contexts in
which it takes place. Currently this includes research on marriages in the military, funded by the Department of
Defense, and marriages in low-income populations, funded by the National Institute on Child Health and Human
Development. He received the Gerald R. Miller Award for Early Career Achievement from the International
Association for Relationship Research in 2004 and has twice been the recipient of the National Council on Family
Relations Reuben Hill Research and Theory Award for outstanding contributions to family science. His textbook,
Intimate Relationships (coauthored with Thomas Bradbury), will be published by W. W. Norton in January, 2010.

People rarely change their minds about subjects that are important to them. Those who favor gun control today are
likely to favor gun control ten years from now, and those who vote for Democratic candidates today are likely to do so
throughout their lives.
Yet intimate relationships, and marriages in particular, are the exception to this rule. After two people stand before
everyone important to them in the world and publicly declare that they love each other and intend to remain together
for the rest of their lives, everything social psychology has learned about the stability of publicly declared opinions
suggests that these will be the most stable opinions of all (Festinger, 1957). Yet of course they arent. Despite the
almost uniform happiness and optimism of newlyweds, most first marriages will end in divorce or permanent
separation (Bramlett & Mosher, 2002), and the rate of dissolution for remarriages is even higher (Cherlin, 1992). In
most cases, this represents a drastic and unwanted change in a highly valued belief, a change that is emotionally and
financially costly to both members of the couple. Even in marriages that remain intact, newlyweds initially high levels
of marital satisfaction tend to decline over time (VanLaningham, Johnson, & Amato, 2001). How can we account for
this change? How is it that marital satisfaction declines so frequently, despite our best efforts to hold on to the positive
feelings that motivate marriage in the first place? And what is it those couples that maintain their initial happiness are
doing right?
What couples that stay happy are doing right
Understanding how marital satisfaction changes requires first understanding how thoughts and opinions about a
marriage and a spouse are structured. Our representations of our partners are complex and multifaceted, consisting
of perceptions that range from specific and concrete (e.g., My spouse makes great pancakes.) to global and
evaluative (e.g., My spouse is wonderful!) (John, Hampson, & Goldberg, 1991). Although we are generally
motivated to believe the best about our partners, we are not equally motivated or able to protect all our beliefs at all
levels of abstraction (e.g., Dunning, 1995). For example, if my partner actually makes terrible pancakes, it is neither
possible nor terribly important to believe otherwise. However, if I am to stay happily married, it is desirable to find a
way to believe that my spouse is wonderful, and it is possible to do so by identifying and focusing on specific
perceptions that might support this global belief.
That is what happy couples do. When couples in the early years of marriage are asked to rate which specific aspects
of their relationships are most important to the success of their marriage, they generally point to whatever aspects of
their relationship are most positive, and the spouses who demonstrate this tendency most strongly are the ones who
are the happiest with their relationships overall (Neff & Karney, 2003). This selection process does not happen only at
the beginning of the relationship. Over time, as specific aspects of the relationship change, with some parts becoming

more positive and some becoming more negative, the couples who stay happiest overall are the ones who change
their beliefs about what is important in their relationships accordingly, deciding that whatever aspects of the marriage
have declined must not be so important after all (Neff & Karney, 2003). As a consequence of this continued process
of selective attention, global evaluations of a marriage tend to be pretty stable from day to day, as these are the
evaluations we are motivated to protect, but perceptions of specific aspects of the marriage tend to vary, more
positive on good days and less positive on bad days (McNulty & Karney, 2001).
So what happens to those less positive specific perceptions? They dont disappear. Even happy newlyweds readily
acknowledge that their partners are not perfect in every way (Neff & Karney, 2005). Staying positive about the
relationship requires that spouses find ways to integrate their perceptions of specific problems and disappointments
within an overall positive view of the marriage. One way spouses can do this is by generating explanations for a
spouses failings that limit any broader implications those failings may have. For example, if my spouse is distant and
withdrawn one evening, deciding that my spouses behavior is a symptom of a difficult day at work (rather than a sign
of a lack of interest in me) means that the behavior has no global implications for my marriage. For spouses who tend
to make these sorts of charitable explanations for their partners disappointing or irritating behaviors, global
evaluations of the marriage remain relatively stable from day to day even when perceptions of specific aspects of the
relationship are fluctuating. For spouses who make less charitable explanations, blaming each other for faults and
missteps, specific perceptions and global evaluations are more closely linked, such that the entire marriage seems
less rewarding on days when specific elements are bad and the entire marriage seems more rewarding on days
when specific elements are good (McNulty & Karney, 2001). In other words, making charitable explanations severs
the link between specific negative perceptions and global evaluation of the marriage, leaving the global evaluations
more resilient. Couples who are able to acknowledge their partners faults while maintaining positive views of their
marriage overall have more stable satisfaction over time (Karney & Bradbury, 2000) and they are less likely to divorce
in the early years of marriage (Neff & Karney, 2005).
Why is maintaining a relationship so difficult?
If this sort of integration is so beneficial, and if happy newlyweds are already doing it, why do newlyweds initially high
levels of marital satisfaction nevertheless decline so frequently? The short answer is that making allowances for a
spouses inevitable shortcomings is difficult, and especially so because marriages and other intimate relationships do
not take place in a vacuum. The way that spouses think about and respond to each other is a product of broader
forces that affect marriages and intimate relationships. As research identifies more of the processes that contribute to
stability and change in marital satisfaction, models of these processes have expanded to account for those broader
forces. One framework that attempts this is the Vulnerability-Stress-Adaptation Model of Marriage (i.e., the VSA
model; Karney & Bradbury, 1995). Consistent with the research described above, the VSA model (see Figure 1)
describes adaptive processes (e.g., solving problems, explaining each others behavior) as directly affecting how
marital satisfaction changes over time. The model further suggests that these processes themselves are facilitated or
constrained by spouses enduring vulnerabilities (e.g., cognitive styles, personality traits, childhood experiences) and
the stressful circumstances they encounter outside the relationship (e.g., work load, financial strains, health
problems).

Figure 1: The Vulnerability-Stress-Adaptation Model of Marriage (Karney & Bradbury, 1995)

Research informed by the VSA model suggests two general reasons why spouses attempts to maintain their initially
high marital satisfaction may fall short over time. First, some people are naturally better at it than others. For
example, when asked to write open-ended paragraphs about issues in their marriages, some spouses recognize that
there can be two sides to every conflict and that compromises are possible. Others write only about their own
perspective, failing to recognize that other perspectives are possible, let alone valid. When couples who have written
these paragraphs are then invited to discuss real marital issues, the ability to recognize multiple perspectives
emerges as a significant predictor of the quality of their discussions, as rated by outside observers (Karney & Gauer,
in press). Where does this ability come from? A likely source is exposure to more or less successful problem-solving
in early childhood. Indeed, wives whose parents divorced when they were children and husbands whose childhood
family environments were highly negative also have more difficulty resolving problems together, and are at risk for
declines in marital satisfaction as a result (Story, Karney, Lawrence, & Bradbury, 2004).
Second, maintaining a relationship takes energy, and in some contexts that energy is in short supply. It is not enough
that couples have the ability to address problems effectively if they lack the capacity to exercise those abilities in the
moment. Unfortunately, in the context of stress, even couples who are normally effective at maintaining their
relationships may find it difficult to do so. To evaluate this possibility, recently married couples were asked about the
kinds of explanations they made for each others negative behaviors every six months for the first four years of their
marriages (Neff & Karney, 2004). At each assessment, they were also asked to describe and rate the stressful events
they had been exposed to outside of the marriage (e.g., stress at work, financial strains, problems with friends or
extended family, health issues, etc.) during each six month interval. Controlling for changes in their marital
satisfaction over that time, the way spouses understood each others negative behaviors at each assessment was
significantly associated with the stress they had been under during that period. When stress was low, spouses on
average were able to generate more charitable explanations for each others negative behaviors, preventing those
behaviors from affecting their global feelings about the marriage. But after periods of relatively high stress, the same
spouses who had demonstrated this ability were significantly less likely to exercise it, and so were more likely to
blame their partners for negative behaviors that they had previously excused.
In addition to highlighting the main effects of enduring vulnerabilities and stressful circumstances on marriage, the
VSA model suggests that these relatively independent sources of influence on marital processes interact. That is,
among individuals with comparable levels of enduring vulnerabilities, those who encounter stressful circumstances
will have an especially hard time maintaining their relationships, and among individuals encountering similar levels of
stress, the ones most at risk for relationship problems are the ones who also have numerous enduring vulnerabilities.
Survey research that oversampled from low-income and underrepresented communities (Rauer, Karney, Garvan, &
Hou, 2008) confirms these sorts of interactions, showing that the associations between relationship satisfaction and
any particular constraint on adaptive processes (e.g., mental health problems, financial strain, substance abuse)
becomes stronger in the presence of other risk factors.
So, why is it so difficult to maintain the initial positive feelings that characterize most newlywed couples? It is difficult
because some disappointments are inevitable in any long-term committed relationship, because some spouses lack
the ability to respond to those disappointments effectively, and because even spouses who have the ability may
encounter stressful circumstances that prevent them from exercising their abilities when they are most needed.
Implications for helping couples succeed
Dominant approaches to strengthening marriages and other intimate relationships focus almost exclusively on
adaptive processes, i.e., teaching couples a set of skills for resolving problems and dealing with disappointments
when they arise (e.g., Markman, Stanley, & Blumberg, 1994). The VSA model and the research informed by it
suggest that there may be a limit to what these approaches can accomplish. Individuals coping with significant
personal vulnerabilities may not be able to change their behaviors. Even couples that know perfectly well how to
respond to each other effectively may lose their capacity for effective adaptive processes when under stress. In light
of these broader forces affecting relationships, policies that address individual well-being and current sources of
stress on family life may be as effective at promoting healthy relationships as any interventions that target
relationships directly. Research on the effects of public policies on marital outcomes supports this idea. In Norway, for
example, after the government began offering cash incentives to parents that elected to forgo state-subsidized
childcare and stay home with their children, divorce rates fell significantly even though the new policy did not target
marriages directly (Hardoy & Schne, 2008). Policies like these that simply make life easier for families and
individuals may contribute to an environment that supports marriages and other intimate relationships. In such an
environment, more spouses and partners may prove capable of maintaining their relationships on their own.

http://www.apa.org/science/about/psa/2010/02/sci-brief.aspx

Longitudinal theories of marriage

How does marital happiness or satisfaction change over time? The answer appears to
depend on how the measures of marital satisfaction have been obtained. It is widely
accepted among researchers that satisfaction follows a Ushaped trajectory: a pattern of
early decline followed by a levelling out during the parenting years and an improvement
when children leave home (Van Laningham, Johnson and Amato 2001). However, there is
also evidence to suggest that marital satisfaction declines regardless of the presence or
absence of children (Clements, Cordova, Markman and Laurenceau 1997). In addition,
other patterns of change in marital happiness have been found, with continuous declines,
continuous increases, and relatively little change across the life span having been reported.
Support for the U-shaped curve tends to come from cross-sectional research (Karney and
Bradbury 1997; Orbuch, House, Mero and Webster 1996), although contradictory results
are to be found among such studies. Linear trends have been found via longitudinal
analyses, but these studies may also be reflecting period effects (changes due simply to
growing older; de Vaus 2001) rather than actual developmental changes. The issue is
further complicated if satisfaction is reported retrospectively. Findings based on
retrospective reports of satisfaction tend in general to be viewed with caution, although
retrospective reports actually may be more accurate because the distance afforded by time
provides a more balanced perspective (Mackey and O'Brien 1995).
In spite of a large body of research, there are few studies following couples over periods of
more than ten years, so researchers have yet to achieve a clear understanding of the
trajectory of marital satisfaction in the longer term (Van Laningham et al. 2001). As with
cross-sectional studies, the outcomes of longitudinal research have also been inconsistent.
Until recent years the inconsistency of findings could have been due in part to the inability of
statistical techniques to deal adequately with the extent and rate of individual change
(Karney and Bradbury 1997). The problem is that while two individuals may be similarly
satisfied at a given point in time, their journey to that point may have been vastly different.
As well as being likely to have commenced the relationship with different levels of
satisfaction, any changes in satisfaction across a given period of time may have been rapid

for one individual and slow for the other; the same event may trigger a rise in satisfaction for
one and a decline for the other. In addition, Karney and Coombs (2000) noted that
analysing mean patterns of change can conceal variability across individuals. Although
satisfaction in their sample of wives showed an overall decline across a twenty-year period,
for a substantial minority of women satisfaction remained constant or increased across
assessment intervals.
It will take time for studies employing advanced techniques that do allow for analysis of
variability across individual trajectories to generate sufficient data to allow firm conclusions
about changes in relationship patterns over extended periods.

Karney and Bradbury's Vulnerability-Stress-Adaptation model

In a landmark paper, Karney and Bradbury (1995) advanced a theoretical framework to
explain changes in marital quality and stability across time and across couples. Based on
their review of over 100 longitudinal studies of marriages and an evaluation of the utility of
four commonly-cited theoretical perspectives - social exchange theory (as applied by
Levinger 1965), attachment theory (Bowlby 1969), crisis theory (as described by McCubbin
and Patterson 1982), and behavioural theory - Karney and Bradbury identified some
general themes that could provide the basis for understanding how and why marriages
survive or break down.
Elements of the four theories highlight aspects of relationships that can provide the
foundations for an integrated theoretical framework. For instance, social exchange theory
suggests that the perception of a relationship is coloured by the rewards and costs
associated with the relationship, the sort of relationship the individual thinks he or she
deserves and their perception of the rewards and costs of being in a relationship with
someone else. Attachment theory specifies certain characteristics of each partner that will
contribute to the functioning of the relationship. Crisis theory suggests that events such as
the transition to parenthood or the experience of unemployment will have an impact on
marital quality. Behavioural theory points to the importance of couple interaction and how
members of couples cope with relational issues, conflicts and transitions.
Based on their analyses of previous research, Karney and Bradbury identified three classes
of variables that, when combined in a single framework, optimise what we can learn about
the processes by which marital satisfaction and stability change over time. The three key

elements of their theory are outlined below. The relationship of the elements to each other is
presented in Figure 1.
Figure 1. A vulnerability-stress-adaptation model of marriage

Enduring vulnerabilities: the individual strengths and weaknesses each spouse

brings to the relationship. These stable characteristics can include their personality,
beliefs and attitudes about marriage, their family of origin and social background.

Stressful life events: incidents, transitions, or circumstances encountered by the

couple that can impinge on their relationship and create tension or stress.

Adaptive processes: the ways in which a couple addresses conflict, how they
communicate, how they support each other and the ways in which they think about
marriage, their spouse and their spouse's behaviour.

According to Karney and Bradbury's model, the ways in which couples deal with the life
events they encounter are the key contributors to the couple's perceptions of the quality of
their marriage. The couple's adaptive processes are a product of the interaction between
the individual spouse's enduring vulnerabilities and the type and severity of the life events
they encounter. Thus satisfaction and stability may be relatively high for a couple who have
few enduring vulnerabilities and poor adaptive processes if those qualities are tested
infrequently. On the other hand, repeated or chronic exposure to stressful events may test
even those marriages where the spouses are normally well equipped in terms of their
individual capacities to cope and their particular patterns of interaction. As Halford (2000)

points out, life events can have both negative and positive effects on a relationship
depending on the strength of the couple's adaptive processes.
A couple's accrued experience in dealing with difficult or stressful circumstances will alter
spouses' perceptions of the quality of their relationship and vice-versa: satisfaction with the
marriage is likely to lead to more positive interactions and behaviours, while engaging in
positive interactions and behaviour is likely to enhance marital satisfaction and perceptions
of quality. Alternatively, unrealistic expectations or dysfunctional patterns of communication
may increase the likelihood of relationship problems and declines in satisfaction over time
(Olsen and Fowers 1986; Olsen and Larsen 1989; Sanders, Halford and Behrens 1999).
Ultimately, repeated failures of adaptation will undermine the stability of the marriage,
leading to increasing frequency of thoughts of divorce; successful adaptation will strengthen
or maintain the relationship and reduce the chances of eventual dissolution of the marriage.
Karney and Bradbury's (1995) theory incorporates personality, family variables, and life
events into an integrated framework that allows for the processes underlying marital change
to be clearly revealed and examined. The following theory by John Gottman is more
narrowly focused, drawing particular attention to the ways in which marital quality and
stability can be eroded.

Gottman's cascade theory of marital dissolution

Gottman's (1993) process theory of a cascade towards marital dissatisfaction and
dissolution incorporates both behavioural and social exchange theories. Gottman looked
beyond the collection of factors blamed for the sharp increase in rates of marital failure
(easier divorce laws, women's financial independence), noting that they do not offer
explanations for why some marriages last and others are dissolved. In contrast to theories
and advice offered by therapists who have generated their material based on their work with
the couples they happen to see in their practice, Gottman's theory is based on scientific
research with hundreds of couples over many years.
A core premise of the cascade theory involves conflict, long considered to have only a
negative impact on a marriage. In the laboratory, couple interactions are studied intensively
over about 20 hours, including a session during which the couple is videotaped discussing
an issue that is creating tension between them. Physiological readings are taken (heart
rate), behavioural responses coded (facial expressions, gestures, reactions), questionnaires
completed and interviews conducted. The couples also rate their own and their spouse's

emotions during the conversation (to gauge how well they read their spouse's emotions).
Trained raters then code the recorded conversation for a range of emotions such as disgust,
contempt, belligerence, and validation. These measures are then correlated with the
questionnaire and interview data to uncover the 'hidden emotional dynamics' of the
relationship (Gottman 1994: 26).
Based on his research, Gottman (1994: 28) concludes that a 'lasting marriage results from a
couple's ability to resolve the conflicts that are inevitable in any relationship'. The key lies in
the balance between positive and negative behaviours. Couples whose positive interactions
outnumber their negative interactions are known as 'regulated': marital stability is stronger
when the ratio of positive to negative behaviours is at least 5:1. Those marriages where
negativity prevails are labelled 'non-regulated' and are more likely than regulated ones to be
unhappy marriages in which separation and/or divorce are or have been considered
(Gottman 1993; Lindahl, Malik and Bradbury 1997).
Not all negative behaviours lead directly to marital distress and dissolution, but four
behaviours in particular, known as the 'Four Horsemen of the Apocalypse', have been
identified as critical in the process by which a marriage can move towards dissatisfaction
and dissolution. Criticism tends to lead to contempt, which in turn leads to defensiveness
and finally stonewalling (withdrawal).
Gottman's research has also found evidence for a process of change over time in spouses'
perceptions of their relationship that results in a 'distance and isolation' cascade. At some
point their spouse's negativity becomes overwhelming, unexpected, and/or intense to the
degree that the spouse reaches a level of desperation such that he or she will do anything
to stop the behaviour. When this point is reached a perceptual shift occurs in which the
feelings of love, respect and safety are replaced by feelings of hurt, sadness, being
threatened, fear and anger. Once this perceptual shift has occurred it can be very difficult to
view the marriage in any other light and the likelihood of maladaptive attributions that
confirm a negative view of the reasons underlying other behaviours is increased.
In a more positive vein, related research suggests that long-married couples are better able
to manage their emotions such that, on the whole, they experience less distress and greater
marital satisfaction than do younger and middle-aged couples (reported in Carstenson,
Graff, Levenson and Gottman 1996). This may be partly because they have learned to
soften conflict with affection, and partly because, over time, some conflicts either resolve

themselves by disappearing, or Australian Institute of Family Studies Research Paper No.

28, July 2002 6 they lose their power to threaten or arouse strong emotions. The need to
resolve every issue may also diminish over time as spouses' priorities and behaviour
change in light of approaching old age.

Applying the theories to marital quality and stability

The aspect of marriage that is pivotal to both Karney and Bradbury's and Gottman's
theories is conflict management - how couples deal with their differences, how they argue
and express themselves both verbally and nonverbally.
Karney and Bradbury's model is wider ranging than Gottman's, incorporating specific
assertions about the role of individual characteristics and life events. The methods of
investigating both theories stand in stark contrast to the literature that has explored the
question of why marriages last. Both adopt quantitative approaches, making measurements
largely from direct observation (Gottman) or from questionnaire responses (Karney and
Bradbury), and although both are ostensibly concerned with investigating marital breakdown
they can both provide rich sources of data to enhance our understanding of how marriages
can be long lasting, happy and rewarding for both spouses.
The 'strengths' approach has tended to be investigated using qualitative methods, although
neither quantitative or qualitative methods is precluded for either approach. Despite the
apparent common sense in doing so, relatively few studies have explicitly set out to uncover
the 'secrets' of lasting marital success by going directly to the source - long-married couples
themselves.
https://aifs.gov.au/publications/why-marriages-last/longitudinal-theories-marriage

IS MARITAL CONFLICT IMPORTANT?

The attention given marital conflict is understandable when we consider its
implications for mental, physical, and family health. Marital conflict has been linked
to the onset of depressive symptoms, eating disorders, male alcoholism, episodic
drinking, binge drinking, and out-of-home drinking. Although married individuals are
healthier on average than the unmarried, marital conflict is associated with poorer
health and with specific illnesses such as cancer, cardiac disease, and chronic pain,
perhaps because hostile behaviors during conflict are related to alterations in

immunological, endocrine, and cardiovascular functioning. Physical aggression

occurs in about 30% of married couples in the United States, leading to significant
physical injury in about 10% of couples. Marriage is also the most common
interpersonal context for homicide, and more women are murdered by their
partners than by anyone else. Finally, marital conflict is associated with important
family outcomes, including poor parenting, poor adjustment of children, increased
likelihood of parent-child conflict, and conflict between siblings. Marital conflicts that
are frequent, intense, physical, unresolved, and child related have a particularly
negative influence on children, as do marital conflicts that spouses attribute to their
child's behavior (see Grych & Fincham, 2001).

WHAT ARE MARITAL CONFLICTS ABOUT?

Marital conflicts can be about virtually anything. Couples complain about
sources of conflict ranging from verbal and physical abusiveness to personal
characteristics and behaviors. Perceived inequity in a couple's division of labor is
associated with marital conflict and with a tendency for the male to withdraw in
response to conflict. Conflict over power is also strongly related to marital
dissatisfaction. Spouses' reports of conflict over extramarital sex, problematic
drinking, or drug use predict divorce, as do wives' reports of husbands being jealous
and spending money foolishly. Greater problem severity increases the likelihood of
divorce. Even though it is often not reported to be a problem by couples, violence
among newlyweds is a predictor of divorce, as is psychological aggression (verbal
aggression and nonverbal aggressive behaviors that are not directed at the
partner's body).

http://www.psychologicalscience.org/journals/cd/12_1/Fincham.cfm

John Mordecai Gottman (born April 26, 1942) is a professor emeritus in psychology known for his
work on marital stability and relationship analysis through scientific direct observations, many of
which were published in peer-reviewed literature. The lessons derived from this work represent a
partial basis for the relationship counseling movement that aims to improve relationship functioning
and the avoidance of those behaviors shown by Gottman and other researchers to harm human
relationships.[1] His work has also had a major impact on the development of important concepts
on social sequence analysis. Gottman is a professor emeritus of psychology at theUniversity of
Washington. With his wife, Julie Schwartz, Gottman heads a non-profit research institute (The
Relationship Research Institute) and a for-profit therapist training entity (The Gottman Institute). [2]
Gottman was recognized in 2007 as one of the 10 most influential therapists of the past quarter
century. "Gottman's research showed that it wasn't only how couples fought that mattered, but how

they made up. Marriages became stable over time if couples learned to reconcile successfully after a
fight."[3]

Predictions of divorce[edit]
Gottman developed multiple models, scales and formulas to predict marital stability and divorce in
couples, and has completed seven studies in this field.[4] These studies regarding newlywed couples
are most well known.
This work concludes that the four negative behaviors that most predict divorce are criticism of
partners personality, contempt (from a position of superiority), defensiveness, andstonewalling, or
emotional withdrawal from interaction. On the other hand, stable couples handle conflicts in gentle,
positive ways, and are supportive of each other.[5]
He developed the Gottman Method Couples Therapy based on his research findings. The therapy
aims to increase respect, affection, and closeness, break through and resolve conflict, generate
greater understandings, and to keep conflict discussions calm. [6] The Gottman Method seeks to help
couples build happy and stable marriages.

Studies[edit]
Gottmans predictions are based on perceived marital bond. In his 2000 study, Gottman conducted
oral interviews with 95 newlywed couples. Couples were asked about their relationship, mutual
history, and philosophy towards marriage. The interview measured the couple's perceptions of their
history and marriage by focusing on the positive or negative qualities of the relationship expressed in
the telling of the story. Rather than scoring the content of their answers, interviewers used the Oral
History Interview coding system, developed by Buehlman and Gottman in 1996, to measure
spouses' perceptions about the marriage and about each other. Therefore, the couples perception
was used to predict marital stability or divorce. The more positive their perceptions and attitudes
were about their marriage and each other, the more stable the marriage. [7]
His models partly rely on Paul Ekman's method of analyzing human emotion and microexpressions.
1992[edit]
The original study was conducted by Gottman and Buehlman in 1992, in which they interviewed
couples with children. An a posteriori modeling yielded a discriminant function that discriminate who
has divorced with 94% accuracy.[8] Gottman believed that since early married life is a period of

change and adjustment, and perceptions are being formed, and sought to predict marital stability
and divorce through couples perceptions during the first year of marriage. [9]
1998[edit]
In a 1998 study, Gottman developed a model to predict which newlywed couples would remain
married and which would divorce four to six years later. The model fits the data with 90% accuracy.
Another model fits with 81% percent accuracy for which marriages survived after seven to nine
years.[10]
2000[edit]
Gottmans follow-up study with newlywed couples, published in 2000, used the Oral History Interview
to predict marital stability and divorce. Gottman's model fit with 87.4% accuracy for classifying
couples who divorce (or not) within the couples first five years of marriage. He used couples
perceptions about their marriages and each other to model marital stability or divorce. [9]

Critiques[edit]
Gottman has been criticized for describing this work as accurately predicting divorce, when generally
this work involves simply fitting statistical models to a data set, not making predictions about events
in the future.
A paper by Richard E. Heyman, "The hazard of predicting divorce without cross
validation"[11] analyzes 15 divorce prediction models and questions their validity.
1. When analyzing a given dataset, it is possible to overfit the model to the data, which will work
extremely nice for this dataset, but will not work when tested on fresh data.
2. Ninety percent prediction may actually mean much less when considering false positives and
the low base rates of divorce.
"Overfitting can cause extreme overinflation of predictive powers, especially when oversampled
extreme groups and small samples are used, as was the case with Gottman et al. (1998; n = 60
couples for the prediction analyses) and nearly all of the other divorce prediction studies ... published
studies that find extraordinary initial predictive results may aid us in improving models of risk by
identifying important risk factors. Nonetheless, dissemination of 'predictive power' results in the
popular media must await supportive data on sensitivity, specificity, and predictive value when the
predictive equation is applied to independent samples. By recognizing both the value and limitations
of predictive studies, professionals and the public alike will be served best." [11]

The author shows his points by creating a divorce prediction model with a data set, and showing its
low validity when the above considerations are tested.
Prof. Gottman never published a reply to this critique.
Journalist Laurie Abraham also disputed the prediction power of Gottman's method. Abraham writes,
"What Gottman did wasn't really a prediction of the future but a formula built after the couples'
outcomes were already known. This isn't to say that developing such formulas isn't a valuable
indeed, a critical first step in being able to make a prediction. The next step, however one
absolutely required by the scientific method is to apply your equation to a fresh sample to see
whether it actually works. That is especially necessary with small data slices (such as 57 couples),
because patterns that appear important are more likely to be mere flukes. But Gottman never did
that."[12] The Gottman Relationship Institute claims that six of seven of Gottman's studies have been
properly predictive, by a non-standard definition of prediction in which all that is required is that
predictive variables, but not their specific relationship to the outcome, were selected in advance.
[13]

However, Gottman's 2002 paper makes no claims to accuracy in terms of binary classification,

and is instead a regression analysis of a two factor model where skin conductance levels and oral
history narratives encodings are the only two statistically significant variables. Facial expressions
using Ekman's encoding scheme were not statistically significant. [14]

Independent studies testing Gottman marriage courses[edit]

Building Strong Families Program[edit]

BSF 15-Month Impact Report

Independent research on the impact of Gottman's marriage strengthening programs for the general
public has further questioned Gottman couple education programs.
The largest independent evaluation of a marriage education curriculum developed by Gottman,
known as "Loving Couples, Loving Children,"[15] was conducted by Mathematica Policy Research[16] at

nine sites in five states through the federally funded, multi-yearBuilding Strong Families
Program study contracted by the U.S. Department of Health and Human Services, Administration for
Children and Families. The study group included low-income, unwed couples.
An impact report released by the Office of Planning Research and Evaluation [17] showed that the
intervention had no positive impact and, in one case, "had negative effects on couples
relationships."[18]

Supporting Healthy Marriage Project[edit]

An ongoing study by Manpower Development Research Corporation (MDRC), [19] known as
the Supporting Healthy Marriage Project(SHM), is evaluating Gottman's "Loving Couples, Loving
Children" program among low-income, married couples. The multi-year, random assignment study is
funded by the U.S. Department of Health and Human Services, Administration for Children and
Families. In an early impact study on the effectiveness of "skills-based relationship
educationprograms designed to help low-income married couples strengthen their relationships and,
in turn, to support more stable and more nurturing home environments and more positive outcomes
for parents and their children," MDRC reported [20] "Overall, the program has shown some small
positive effects, without clear indications (yet no clear negative proof) for improving the odds to stay
together after 12 months."
The program is still ongoing.

Matthews, Wickrama and Conger[edit]

A study published by Matthews, Wickrama and Conger in 1996 based on couples perceptions
showed that spousal hostility, net of warmth, predicted with 80% accuracy which couples would
divorce or not divorce within a year.[21]

Relations and effects[edit]

In multiple analyses, Gottman has shown a plethora of relations and effects in marriage and divorce,
some in peer-reviewed publications, while many others appear in Gottman's own books. Among
those are

The physical elements in marital conflict (i.e., physical effects are central to the inability to
think, etc., in conflict situations) for which he advises a 20-minute cooling period or physical
relaxation.[22]

The effects of "bids for connection." That is the smallest bids people do to connect and how
the other reacts. For example, happy couples do have many more "bids for connection" when
together, and much more "turn towards" response, and much, much less "turn away" - the most
negative reaction. The book dedicated to this element is "The Relationship Cure."

The concept of "trust," which Gottman defines as the tendency to cooperate to form "winwin" situations, and the ability to get unstuck from the loss-loss state loop (like mutual defecting
in the Prisoner's dilemma). A central feature of unhappy relationships, notes Gottman, is that
couples are stuck in loss-loss loops.[23]

The neutral affect provides a way out of negative interactions as most interactions do not
transition directly from negative to positive. The degree of neutral affect is often overlooked as a
predictor of relationship success due to the very fact that the neutral affect is simply neutral. [23]

The dynamic to cause divorce in the short term is different from that causing divorce later.
Early divorce is characterized by the "four horsemen" of bad fighting, whereas later divorce is
characterized by lower positive affect in earlier stages of the relationship.

Anger is not at all bad for relationships. Happy couples are as frequently angry as unhappy
couples. It seems that how people react to anger and how destructive they get is the crucial
factor rather than the frequency of anger or fights. Gottman even says that anger is functional in
marriage.

69% of happy couples still have *the very same* unresolved conflicts after 10 years, yet
remain happy because they do not get gridlocked in the conflict and manage to get around it. [24]

Contempt and marriage[edit]

Gottman's theory states that there are four major emotional reactions that are destructive and thus
are the four predictors to a divorce: criticism, defensiveness, stonewalling, andcontempt. Among
these four, Gottman considers contempt the most important of them all. [25]

Seven Principles[edit]
In his book The Seven Principles for Making Marriage Work, (his most popular book) Gottman
discusses behaviors that he has observed in marriages that are successful and those that are
detrimental to marriage, based on his research conducted at his lab in Seattle, Washington. He has
outlined seven principles that will reinforce the positive aspects of a relationship and help marriages
endure during the rough moments

https://en.wikipedia.org/wiki/John_Gottman