I decided in this post to have a look at a few of the chapters in the first part(s) of the book. As earlier mentioned I lost my notes and highlights to these parts of the book due to computer trouble, making it much more difficult and time consuming to blog this stuff, but I wanted to cover some of that stuff even so because if I don’t I’ll forget the details (to the extent that I have not already – I should caution that this post provides relatively ‘lazy coverage’ as I felt it to be completely out of the question to select material from the book to talk about here using the same criteria I normally employ).
An obvious but important conclusion from the chapter on The Affective Structure of Marriage, which is a chapter that among other things covers multiple conceptual models dealing with relationship change, is that: “various models of marital change are useful because no single pathway describes changes in all, or even most, marriages. Even among couples sharing a similar outcome (e.g., divorce), there is considerable variation in the course toward that outcome. This implies that attempts to develop a single explanation or description of divorce are likely to be, at best, incomplete. Concluding that multiple models are useful is merely recognizing that there are multiple developmental processes in marriage.” Relationships may change for all sorts of reasons, and there is no full model out there which explains everything. This stuff is complicated. It’s noted in the chapter that some of the attempts people have made at trying to e.g. predict which couples divorced over a given time period turned out to perform really quite well on one sample (using longitudinal data – this is not just unsophisticated cross section analyses we’re dealing with), but then it turned out later that they perform really quite horribly on validation samples using the same type of data to predict outcomes in different couples. Similar stimuli may have different effects depending on how long people have been together. Different models deal with time frame aspects in different ways.
I’ll mention a few results from the literature covered in that chapter here. One is that couples who were initially more affectionate and less antagonistic were happier 13 years later than were other couples who were still together at that point in time but had lower initial levels of affection/higher antagonism. It’s also been found that couples which are high in antagonism early on in the relationship (‘lots of drama’) are more likely to divorce early on; disillusionment after a few years of marriage seems to be a better predictor of divorce years later, with initial affection being an important moderating variable in the sense that couples who were initially higher in affection were together for a longer period of time before they eventually divorced. Shorter relationship duration at the time of marriage seems to predict divorce. Personality characteristics such as (trait) (presumably also state-, US) anxiety and neuroticism are associated with relationship dissatisfaction and divorce risk. I should probably once again emphasize that the only reason why I’m not providing effect sizes here is that the authors do not, and so I’m not able to. Some conclusions from the chapter:
“one of the most exciting nascent trends in the marital literature involves the recognition that there is not a single unitary process leading to marital distress and divorce […]. Some couples begin marriage with lower marital satisfaction than most other couples but remain married indefinitely, whereas other couples begin marriage very satisfied but end up divorcing. Moreover, the predictors of dissatisfaction and divorce are not always the same; for instance, stable characteristics such as trait anxiety appear to be more strongly related to satisfaction than they are to divorce […]. Even the processes leading to divorce are not uniform, with some couples who eventually divorce beginning marriage with high levels of hostility and divorcing quickly, others beginning marriage with moderate amounts of both positive and negative elements before becoming quite low in affection, and still others beginning marriage with exceedingly high levels of affection that are not sustained over the early years of marriage. Also, the predictors of divorce are different for divorces that occur earlier in marriage compared with those that happen later in marriage. […] Being high in conscientiousness [for example] appears to diminish the chances that one will divorce early in marriage but does not appear to prevent eventual divorce”.
It should be noted, as they also do, that much of this research is based on what they in the book call ‘observational data’, which in this context means data obtained by actually observing the individuals, usually in a lab, and then coding specific behaviours in specific ways. They didn’t just ask people if they were affectionate towards each other; they tried to estimate whether or not they were, based on behaviours they could observe. There are problems with this sort of data and they talk about that in the chapter; for example it has been argued (I think I may have talked about this before in my coverage) that the most effective kind of support may well be invisible support (“actions that take place outside the recipients’ awareness”, or supportive actions which are provided “in such a skillful way that, although the information about the transaction is available to the recipient, the transaction is not coded as enacted support”) – and this sort of support is difficult to observe in a lab; whereas on the other hand the most visible sort of support, which is the easiest type to code by observers, may be counterproductive (such actions may provide a signal to the partner that the other party considers him/her too incompetent to handle the task on his/her own, which may lead to self-doubt etc. in the recipient), perhaps making interpretations slightly more difficult than one might think they are. A related problem seems to me to be that not providing support may in some situations be the optimal approach to take by the partner (‘my partner obviously doesn’t need my help right now, and if I were to provide support in this situation this would not be helpful’), and so such behaviours may be indicative of a strong relationship – yet that’s not how such behaviours will be coded in the studies. There are some problems here.
Next, a few observations dealing with divorce and postdivorce relationships. This data is old, but better than nothing: “Most divorced adults find another romantic partner. In the United States, the probability of cohabiting after the dissolution of first marriage is 70% after 10 years […] Census estimates project that in the United States nearly 85% of divorced people remarry […]. Although the remarriage rate is lower in other Western societies, most divorced people eventually cohabit or remarry […] It is an almost universal finding that children have more difficulty adapting to parental remarriage than do the adults.” I thought I should mention in a slightly unrelated context that I recently came across a Danish article about how children are dealt with here in the divorce context; I was not surprised to learn that women get custody in 90% of the cases – the politicians are thinking about changing this (this did surprise me), which has caused some organizations to argue that it’s a bad idea to change this state of affairs (again, not surprising). I’ve blogged US data on this stuff before – go have a look at the archives/use the search function if you’re curious, I’m too lazy to provide a link. I believe the US numbers are reasonably similar. An important observation made in the chapter is that parenting roles have evolved over time, and that the institutional setup had not really evolved with them at the time this book was written: “Child support policies have been predicated on the notion of fathers having only one set of children to support. In fact, increases in multiple marital and cohabiting relationships means that nearly 75% of remarried men have multiple sets of children to support (emotionally and financially) both inside and outside their current relationship.” It’s important to observe in this context that the proportion of all marriages which were remarriages was really high in the US, and that the remarried couples made up a big proportion of the total: “About half of all U.S. marriages are remarriages for one or both partners” (data from the U.S. Census Bureau, 2000). Things may or may not be different today.
Some observations from the chapter about personal relationships in adolescence and early adulthood: “Friendships and romantic relationships are tightly interwoven in adolescence and early adulthood. Unsupervised mixed-gender peer groups during adolescence provide opportunities and supportive environments for “pairing off” between group members. By mid-adolescence, most individuals have been involved in at least one romantic relationship; by the early years of early adulthood, most are currently participating in an ongoing romantic relationship (Collins, 2003). […] Existing findings point to a shift in the qualitative characteristics of dating relationships between the ages of 15 and 17 years, and dating among early adults seems similar in key ways to dating among late adolescents. After age 17, the likelihood of being involved in a romantic relationship changes little […] Having a romantic relationship and the quality of that relationship are associated positively with romantic self-concept and, in turn, with feelings of self-worth […], and longitudinal evidence indicates that by late adolescence, self-perceived competence in romantic relationships emerges as a reliable component of general competence […]. Whether adolescent romantic relationships play a distinctive role in identity formation during adolescence is not known, although considerable speculation and some theoretical contentions imply a link […] The most widely studied patterns have to do with variations in the timing of involvement in both romantic relationships and sexual activity, typically showing that early dating and sexual activity are risk factors for current and later problem behaviors and social and emotional difficulties […] The social worlds of those involved in romantic relationships differ from those who are not because romantic partners quickly become dominant in the relationship hierarchy […]. Although romantic interconnections initially are predicated on principles of social exchange, commitment drives participants to transform this voluntary relationship into one that is more obligatory and permanent […]. Eventually, most early adults marry and reproduce, further transforming the relationship and marginalizing remaining friendships, thus effectively ending the peer group’s dominance of relationship experiences”.
And finally some data and observations from the chapter about close relationships in middle and late adulthood: “The majority of adults in the United States are married, but the proportion is smaller in old age than earlier in adulthood (ages 35 to 54 years = 71.3%, 55 to 64 years = 74.2%, and 65 or older = 56.7%), and a notable sex difference in the proportion married exists between men and women aged 65 or older (75.7% versus 42.9%, respectively). The majority of households comprise family households (68%), usually of married couples (52%), but 32% of adults live in non-family households, including the 26% who live alone [do keep in mind that many of those 26% are involved in romantic relationships as well, though the characteristics of the relationships they have are different]. Among persons aged 75 years or older, however, the proportion living alone is much higher (39.6%) because of the greater likelihood of being widowed (ages 35 to 54 years = 1.6%, 55 to 64 years = 6.7%, 65 to 74 = 19.6%, and 75 or older = 41%, U.S. Census Bureau, 2003). […] the proportion of householders with children of any age at home remains above 50% even in the 45- to 54-year-old age group (Russell, 2001). […] One of the key findings of research on the causes and consequences of relational difficulties in adulthood is that negative dimensions of interactions have stronger effects than positive ones on relationship quality and satisfaction.”