Adult attachment, parenthood and intimacy

Changing families, challenging futures
6th Australian Institute of Family Studies Conference
Melbourne 25-27 November 1998

Patricia Noller
University of Queensland Family Centre

This discussion paper introduced the symposium on Attachment, parenthood and intimacy. It includes references to the other papers presented during the symposium: Attachment style and coping resources as predictors of coping in the transition to parenthood by Richard Alexander, Judy Feeney and Pat Noller; Conceptualising the Pair-Bond: Attachment, Caregiving, and Sexuality as predictors of Intimacy in Adult Romantic Relationships by Charles Farrugia and Lydia Hohaus; and The transition to parenthood : A study in progress by Lydia Hohaus, Judy Feeney and Pat Noller.

In terms of research on personal relationships, there is no doubt that adult attachment theory is the flavour of the decade. Since Hazan and Shaver's seminal work in 1987, there have been a plethora of studies relating adult attachment to just about anything. Adult attachment theory has built on the work of Bowlby and Ainsworth who focused on the attachment bond between the parent and the child.

Hazan and Shaver (1987) constructed a 3-item forced-choice measure designed to allow adult participants to categorize themselves as secure, avoidant or anxious-ambivalent, categories used by Ainsworth to describe young, children's attachment to their mothers (see table 1.) They found that these attachment styles existed in similar proportions in adults as in children in Ainsworth's studies (about 60% secure, 25% avoidant and 15% anxious-ambivalent). Initially, I suspect, some authors, at least, added in attachment style as another variable in their study to see what would happen. More recently, studies have been designed to look at more complex Issues and to add to our understanding, of the central concepts of the theory.

Issues that have been explored by researchers include the stability of attachment ratings (Baldwin & Fehr, 1995), how attachment style affects partner choice (Frazier et al., 1996), links between attachment, caregiving and relationship functioning (Carnelly, Pietromonaco & Jaffe, 1996). the impact of attachment style on couples' relationships (Gaines et al., 1997; Noller & Feeney, 1994; Feeney, Noller & Callan,1994), and violence in couple relationships (Dutton et al., 1994; Roberts, 1998).

In addition, ways of measuring adult attachment have changed, with the three-style formula being replaced to some extent by Bartholomew's theorizing that, if styles of attachment were plotted on Bowlby's dimensions of view of self and view of other, then there would be four styles (see table 2). In addition, there have been a number of attempts to develop measures that would provide continuous rather than categorical measures of attachment-related constructs. All three of the papers presented in this symposium use the Attachment Style Questionnaire (Feeney, Noller & Hanrahan, 1994), which was our attempt to develop a continuous measure so that individuals could be assessed on a number of attachment-related dimensions, rather than being placed in somewhat ill-fitting categories. It is a 40-Item likert scale measure which contains five subscales: Preoccupation, Need for approval, Discomfort with closeness, Confidence in self and others and Relationships as secondary.

The studies to be presented here are distinctive in several ways. Firstly, rather than focusing on the categories of attachment style, the focus is entirely on continuous measures. Charles Farrugia, in his study, uses the five scales of the original ASQ, whereas Hohaus et al. and Alexander et al. use the two major underlying dimensions of Anxiety over abandonment and Discomfort with closeness. Similar scales were originally constructed from the 15 statements in Hazan and Shaver's 1987) categorical measure, and have been used by us and by others in an attempt to use continuous measures and to deal with the problem of small cell sizes, particularly in the insecure groups, when categorical data are used. All three studies reported here are based on self-report questionnaires only.

Secondly, rather than just looking at attachment style differences, these studies
have been designed to test Hazan and Shaver's (1994) model of the three behavioural
systems: Attachment, Caregiving and Sexual mating (see figure 1). Farrugia looks at the extent to which various measures of the three behavioural systems predict intimacy in adult romantic relationships in a sample of students, with the prototypical pair bond of Hazan and Shaver being conceptualized in terms of intimacy. Farrugia also looks at the functions of attachment, and the extent to which they also predict intimacy (see figure 2). His study is a single wave study, with the data gathered at a single point in time.

Hohaus paper and the Alexander paper are both based on the same sample: a large group of couples having their first baby. In addition, they both report on the first two waves of this four wave longitudinal study. For the Hohaus et al. paper, the goal again was to test Hazan and Shaver's ( 1994) model, and particularly to look at whether the model applied differentially to couples having their first baby than to a childless control group.

The paper by Alexander focuses on attachment, affect regulation and coping. First the author looks at the different patterns of correlations for the two groups of couples. Then the two attachment dimensions (Anxiety and Discomfort), the coping resources of self esteem and social support, the infant-related, variables of infant difficultness and the experience of parental strain are used to predict two styles of coping: problem-focused and emotion focused. The three papers are designed to test Hazan and Shaver's (1994) model of the three behavioural systems in predicting intimacy, marital satisfaction and coping, respectively .

___________________________________________________________________________

The three papers are presented at this point.

___________________________________________________________________________

Farrugia's results show clear support for the Hazan and Shaver model, with measures of attachment, caregiving, and sexual mating, and of attachment functions all being related to intimacy, as measured by the Psychosocial Intimacy Questionnaire of Tesch (1985). Regression analyses showed that the four sets of variables together accounted for a large amount of the variance in intimacy scores, with each set contributing unique variance. It is important to note that in all of the studies, only some aspects of the three behavioural systems were related to the criterion variable, whether intimacy marital satisfaction or coping were involved.

The attachment scale , Confidence, a measure of attachment security, was the only measure of attachment that predicted intimacy, the correlations with the other dimensions being generally weaker. Measures of caregiving that predicted intimacy were Proximity vs Distance and Sensitivity vs Insensitivity, with Cooperation vs Control not being included in the analysis.

The only measure of the sexual relationship which predicted intimacy was Sexual communication satisfaction. It makes sense that the Sexual behaviour index is not predictive, given that the actual activity engaged in is probably less important than whether the members of the couple are satisfied. It is also important to realise that the measure of sexual communication used here contains a number of items that reflect the quality of the sexual relationship such as openness, willingness to be involved and to meet the partner's needs. It is less obvious why the Pleasure mean did not predict intimacy, but as I do not know the scale, I cannot comment. The limitations of this study include the use of university students only in the sample, the recruitment of only one member of the couple and the lack of a longitudinal design. I get quite concerned about the widespread use of university students in studies of romantic relationships. A large amount of what we know about love, for example, is based on the experiences of 18 and 19 year olds, and with all due respects to the young, I doubt that they are really the world experts on love or on love relationships. A further problem is the fact that these students are not even representative of their age group, being better educated and generally higher in socioeconomic status than many of their peers.

I have to admit that I did not understand the reason why only one member of the dyad was included in the study - something to do with changes in the relationship if both members of the couple keep records of their sexual activity. It was not clear from the method section that the participants were keeping diaries, but rather completing a questionnaire at a single point in time. I think this point needs some elaboration. Although I am a strong advocate of involving more than one member of a relationship in research, I also understand the difficulties that can ensue. While it is possible that including both members would have dealt with the problem of balancing the numbers of each sex in the sample, it may also have reduced the sample size overall, given that it is generally easier to get one member of a couple to be involved in research than to have to persuade both members of a couple.

With regard to the use of a longitudinal design, I'm not sure that I have ever seen a list of problems as comprehensive as the one Charles has provided for us. I think I will file it away so that I can use it when 1 want to justify not using a longitudinal design. While clearly it would be interesting to see whether the three behavioral systems assessed at one point in time can predict intimacy at a later point, it is true that longitudinal studies are expensive, and do take a long time to complete. In addition, the longitudinal usefulness of the three behavioural systems in predicting intimacy and other related variables is tested in the other two studies.

A surprising aspect of the Hohaus et al. study is the lack of change over time in the measures of the three behavioural systems, particularly given that almost half of the couples had been through a major transition which would be expected to affect all three systems. I would have predicted a group by time interaction, with the transition group changing on at least some of the systems, and the control group staying the same. The gender differences arc generally in line with findings from studies of marital relationships and marital communication, and also reflect gender stereotypes. They support the proposition that there are differences between husbands and wives in how they behave in close relationships.

When aspects of marital satisfaction were used as the dependent variables, group differences were in evidence, with those in the transition group reporting less quality time together than the control group. The interaction effects (gender by time, gender by time by group) showed that wives in the transition group were more likely than any other group to report less quality time together after the birth of the baby. There was also evidence that wives reported a decreased amount of quality time together after the birth, whereas husbands reported an increase. There are clearly differences in perceptions here. I suspect that wives are reporting on time alone together, whereas husbands are counting time together with the baby and the wife. The items tend to focus on leisure time, and given that the babies are only between four and six weeks old, the wives are probably not having much leisure time. In addition, wives are likely to be most affected by extra chores, lack of sleep etc.

The regression. analyses support the relevance of the three behaviouraI systems for predicting marital satisfaction for both husbands and wives, at both time 1 and time 2, with all regression analyses indicating significant prediction. Results were not strong, however, and for pregnant wives, the group of most interest, none of the variables provided any unique contribution. Anxiety was the most consistent predictor from the attachment set, Sensitivity was the most consistent predictor from the caregiving set and Sexual satisfaction was the most consistent predictor from the sexuality set. It is difficult to know whether different measures of the systems would have produced stronger results.

The most interesting question, of course, is whether the behavioural systems at time I predict marital satisfaction at time 2. At this stage, that question has not been answered, and we will need to wait for further episodes emerging from this study in progress. For this study, such an analysis would tell us whether the relationship of the couple during pregnancy could predict the quality of their relationship after the birth of their first child.

The Alexander et al. paper focuses on attachment theory as a theory of affect regulation, with attachment behaviours serving to regulate the distress associated with perceived threat. As noted in the paper, Individuals with a secure attachment style (in terms of Alexander's paper, low in both anxiety and discomfort) being able to acknowledge their distress and seek support from others, those avoidant in attachment style (high in discomfort and low in anxiety) tending to restrict both their acknowledgment of distress and their support seeking, and those who are anxious ambivalent in attachment style (high in anxiety but low in discomfort) tending to focus very strongly on their distress and their need for support. The main focus of the study is to look at the links between attachment, coping resources, infant-related variables, and the coping strategies used in dealing with the transition to parenthood. In this study, regression analyses were used to predict coping strategies at time 2 from attachment at time 1, for the transition couples only, although it was difficult to work out from the paper which variables in the analyses were assessed at which time. More clarity is needed here. The results were generally weak for problem-focused coping and for social support seeking, but stronger for emotion-focused coping for both husbands and wives.

The results for husbands' emotion-focused coping suggest that emotion-focused coping is predicted by high anxiety, low self-esteem and high parental strain, with self-esteem mediating the link between attachment (particularly anxiety) and coping. For wives, the pattern was similar, except that high support seeking, as well as low self-esteem and high parental strain, predicted emotion-focused coping.

Thus this study provides support for the importance of attachment, along with coping resources and infant-related variables in predicting emotion- focused coping. For both husbands and wives, attachment, and particularly anxiety, was important in predicting the extent to which the new parents used emotion-focused coping. For husbands only, low anxiety was related to the use of problem-focused coping. It is important to keep in mind that emotion- focused coping is generally considered less constructive than problem-focused coping, because the emphasis is on dealing with the affect (e.g., by crying, getting out of the situation etc.), rather than trying to solve the problem.

Overall, the three studies provide support for the importance of considering attachment issues in studying romantic and marital relationships, and particularly those going, through a stressful transition. Attachment was significantly related to intimacy in romantic relationships, and to marital satisfaction and coping in married couples experiencing the birth of their first child.

There was also clear support for Hazan and Shaver's (1994) model of the three behavioural systems as predicting the quality of the pair bond between couples in a romantic relationship, including marriage. It is important to be able to show the validity of this model through empirical research, for both couples in a romantic relationship and those going through the transition to parenthood.

Given that the Hohaus et al. paper and the Alexander et al., paper both come from two waves of a four-wave study, we can presumably look forward to other presentations that will increase our understanding of the importance of attachment in adult relationships. It will be important to see how well attachment and the three behavioural systems at each phase of the study will predict the state of these relationships six months after the birth. I'm sure we all look forward to hearing more as further data become available.

References

Baldwin, M., & Fehr, B. (1995). On the instability of attachment style ratings. Personal Relationships, 2, 247-261.

Carnelly, K., Pietromonaco, P., Jaffe, K. (1996). Attachment, caregiving and relationship functioning in couples: Effects of self and partner. Personal Relationships, 3, 257-279.

Dutton, D. G., Saunders, K., Starzomski, A. & Bartholomew, K. (1994). Intimacy, anger and insecure attachment as precursors of abuse in intimate relationships. Journal of Applied Social Psychology, 24, 1367-1386.

Feeney, J. A., Noller, P. & Callan, V. J. (1994). Attachment style, communication and satisfaction in the early years of marriage. Advances in Personal Relationships, 5, 269 308.

Frazier, P. A., Byer, A. L., Fischer, A. R., Wright, D. M. & DeBord, K. A. (1996). Adult attachment style and partner choice: Correlational and experimental findings. Personal Relationships, 3, 117-137.

Gaines, S. O., Reis, H. T., Summers, S., Rusbult, C., Cox, C. L., Wexler, M. 0., Marelich, W. D. & Kurland, G. J. (1997). Impact of attachment style on reactions to accommodative dilemmas in close relationships. Personal Relationships,4, 93-114.

Hazan, C. & Shaver, P.R. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 5, 511-524.

Hazan, C. & Shaver, P.R. ( 1994). Attachment as an organizational framework for research on close relationships. Psychological Inquiry, 5, 1-22.

Noller, P. & Feeney, J. A. (1994). Relationship satisfaction, attachment, and nonverbal accuracy in early marriage. Journal of Nonverbal Behavior, 1 199-2211.

Roberts, N. D. (1998). Communication in violent relationships. Unpublished doctoral dissertation, University of Queensland.

Table 1

Secure
I find it relatively easy to get close to others and am comfortable depending on them and having them depend on me. I don't often worry about being abandoned, or about someone getting too close to me.
Avoidant
I am somewhat uncomfortable being close to others. I find it difficult to trust them comletely and to allow myself to depend on them. I am nervous when anyone gets too close and, often, love partners want me to be more intimate than I am comfortable being.
Anxious/ambivalent
I find that others are reluctant to get as close as I would like. I worry that my partner doesn't really love me or won't want to stay with me. I want to merge completely with another person, and this desire sometimes scares people away.

Table 2

	MODEL OF SELF (Dependence)
	Positive (Low)	Negative (High)

Positive (Low)	SECURE Comfortable with intimacy and autonomy	PREOCCUPIED Preoccupied (Main) Ambivalent (Hazan) Overly dependent
MODEL OF OTHER (Avoidance)	_________________________________________
Negative (High)	DISMISSING Denial of Attachment Dismissing (Main) Counter-dependent	FEARFUL Fear of Attachment Avoidant (Hazan) Socially avoidant

Return to Conference Papers Index
Australian Institute of Family Studies, Level 20, 485 La Trobe Street, Melbourne, Vic 3000, Australia. Telephone: (03) 9214 7888; International: 61 3 9214 7888. Facsimile: (03) 9214 7839; International: 61 3 9214 7839. URL: http://www.aifs.gov.au/