Agreement in Married Couples: The Effects of Adult Attachment

In support interactions between two people the level of agreement between the amount of support provided and the amount of support received can have important implications about the quality and the effectiveness of the interaction. The purpose of this study was to investigate the extent to which attachment style predicts agreement between married couples regarding the occurrence of specific support behaviors during a specific interaction. Married couples (N = 190) from the Pittsburgh community were videotaped as they discussed a goal that one member of the couple hoped to accomplish. Then, both couple members reported the extent to which specific behaviors occurred during the interaction. Results indicated that individuals with an insecure attachment style were more likely to disagree with their partner than individuals with a secure attachment style. When individuals with insecure attachment styles did agree with their partners, they tended to agree about low levels of support and high levels of negativity. Implications of results and directions for future research are discussed. Agreement in Married Couples 3 Agreement in Married Couples: The Effects of Adult Attachment Communication is thought of as a major factor in achieving happy and healthy relationships. Studies have been done examining different types of communication as well as the amount of communication in which partners engage (Gaelick, Bodenhausen, & Wyer; 1985, Noller, 1980). Communication can be separated into two different typesverbal and non-verbal. Verbal communication carries the basic content of the message while non-verbal communication often indicates how the message should be interpreted (Noller, 1980). Both types of communication are important in the accurate sending and receiving of messages. A measurement of the effectiveness of communication is the degree to which couple members agree on the meaning of messages being sent to and received from one another. This agreement may depend on the encoding and decoding of messages aided by both verbal and nonverbal forms of communication. Encoding is defined as the ability to send out unambiguous messages with clear intent, while decoding is the ability to accurately recognize the intent in a message (Noller, 1980). Marriage relationships, due to their intensity and intimacy, are particularly prone to misunderstandings in communication (Bach & Wyden, 1969). The inability to communicate effectively is often thought to be a key feature of unhappy relationships (Gaelick, Bodenhausen, & Wyer; 1985) and it is believed that common miscommunications can influence partner’s feelings towards each other and their overall satisfaction with the relationship (Gaelick, Bodenhausen, & Wyer; 1985). The purpose of this investigation is to identify important factors that may predict difficulties in encoding and decoding messages accurately, thus leading to ineffective communication and misunderstandings. Specifically, we identify and examine predictors of the extent to which partners agree about whether specific behaviors occurred during a concrete Agreement in Married Couples 4 interaction. We propose that attachment style is likely to be an important predictor of agreement between couple members. Furthermore, investigating predictors of agreement between couple members regarding a specific interaction may reveal important information about the conditions under which effective communication is attained and how to promote more effective communication between partners. First, I review prior research that has investigated agreement between relationship partners. Then, I describe how attachment theory may help us understand the extent to which couple members agree about what happened during a specific interaction. Prior Research Regarding Agreement Previous studies in communication have been done looking at encoding and decoding of messages sent and received by romantic partners. In Patricia Noller’s study of non-verbal communication (1980), all couples completed a Marital Adjustment Test (Locke & Wallace, 1959) and were assigned to different groups based on their scores. Couples that had both partners score 120 or higher on the test were put in the high marital adjustment group, couples that had at least one partner score 95 or below on the test were put in the low marital adjustment group, and all other couples were put in the moderate marital adjustment group. This study showed that couples with high marital adjustment scores were able to communicate more effectively (as measured by a modified version of Kahn’s Marital Communication Scale) than those with low marital adjustment scores. The couples in the low marital adjustment group had greater difficulties sending positive messages and tended to decode messages in a more negative direction. Thus, even if one partner has the intention to send a positive message, there are factors such as low vs. high marital adjustment, that may affect how the message is sent and how it is received, leading to potential disagreement between partners. Agreement in Married Couples 5 Other studies have assessed agreement specifically with regard to social support interactions. However, it is difficult to compare results across studies because researchers have used different indices of agreement. Jacobson and Moore (1981) and Antonucci & Israel (1986) calculated agreement as a percentage of matched yes/yes answers; however, these studies did not take into consideration agreement about no support being given. A more recent study conducted by Coriell and Cohen (1995) did consider agreement about the occurrence and the nonoccurrence of support (using percentage agreement scores) by matching both yes/yes answers and no/no answers. The Coriell and Cohen (1995) study investigated agreement within dyads with regard to the occurrence of supportive behaviors. In their study, participants were college students who had an upcoming exam. These students reported behaviors that they expected from a specific support person, such as their roommate, before the exam. Then, after the exam was over, both the student and the support person reported the supportive behaviors that were received and provided respectively. The agreement between the student pairs about the occurrence of, or the meaning behind, certain behavioral transactions was called concordance. Results of this investigation indicated agreement levels of 28% for helpful behaviors. One of the variables that Coriell and Cohen predicted would be associated with concordance is intimacy, and their results showed that intimacy predicted 16% of the variance in concordance. This is consistent with the idea that the more intimate a relationship is, the more accurate are partner’s expectancies for one another and more attention is paid to ongoing behaviors. The agreement level of 28% found in Coriell and Cohen’s study is low compared with previous research. For example, in Antonucci and Israel’s (1986) study, participants were asked to name people who were close and important to them and who provided them with certain Agreement in Married Couples 6 support functions such as reassurance, respect, sick care, talk with when upset, talk with about health, and confiding. Then, agreement was measured in terms of specific veridicality (the extent to which there is agreement between an individual and his or her network member in their report of whether support is provided or received), and overall veridicality. Specific veridicality was assessed using each of the six support functions while overall veridicality was an aggregated measure that assessed the degree to which individuals and their network member agreed that any support (provided or received) had been exchanged. They also assessed the degree to which the type of relationship (as an index of relationship closeness) predicted veridicality, and it was shown that both specific and overall veridicality were highest among spouses(56%-83%), less high among other family members (46%-59%) and lowest among friends. (30%-47%). The purpose of the Jacobson and Moore (1981) study was to examine the degree to which couples agree regarding the relationship behaviors that had occurred during a 24-hour period. The study had spouses each complete a behavioral checklist everyday for 21 consecutive days. The checklist included 409 behaviors that they report as having occurred or not during the past 24 hours. The average agreement between couple members was 48% but they found that nondistressed couples tended to agree a greater percentage of the time than distressed couples, and the items in the behavioral checklist that were more concrete (“We attended a sporting event” or “We watched TV”) elicited more agreement between couple members than more inferencebased items (“Spouse was tolerant when I made a mistake” or “Spouse confided in me”). One downfall of all these previous studies regarding social support agreement is that they were unable to capture graded frequencies of support. Since agreement was measure as either yes or no, it was not possible to say whether dyads in these studies agreed about the frequency or Agreement in Married Couples 7 amount of support. One study that did measure graded frequencies was the Abbey, Andrews, and Halman (1995) study where participants (fertile and infertile married couples) were asked to rate on a five-point Likert scale the amount of support provided and received during a four week period. They found that wives’ and husbands’ perceptions of the amount of emotional support they received from their spouse were only moderately correlated with what their spouse reported providing. Although the Abbey et al. (1995) study captured frequencies of support, it still ignored the fact that two dyads could have the same agreement score but one dyad could be agreeing about low levels of support while the other could be agreeing about high levels of support. One of the only studies we could find that addressed both the amount of support and the magnitude of agreement between dyads is a study conducted by Gant (1998). This study used a standardized difference score to measure agreement (but ignored amount of support) and a categorical measure of agreement that provided information about the amount of support. The Gant (1998) study investigated predictors of agreement with college students as the recipients and non-romantic friends as the providers. Both recipient and provider were given an adaptation of the UCLA-SSI (Dunkel-Schetter, Feinstein, & Call, 1986) scale to assess social support. The results revealed a correlation of .48 between the recipients’ and providers’ total social support scores. Factors such as intimacy and self –disclosure were found to be predictors of agreement, also recipients who expressed a greater desire for support showed more agreement with their providers. Most of the previous studies done on agreement have shown that factors like marital satisfaction and intimacy are significant predictors of agreement between dyads. More intimate relationships and married couples with higher marital satisfaction/ lower distress display greater agreement between partners. Based on these findings, factors that may affect marital satisfaction Agreement in Married Couples 8 or the intimacy of a relationship would also be predicted to have an effect on agreement between dyads. Attachment as a Predictor of Agreement The current investigation tests the idea that the attachment characteristics of relationship partners predict the extent to which they agree about the occurrence of specific interaction behaviors. Attachment theory states that people are predisposed to form strong emotional bonds with particular individuals (attachment figures) and seek comfort from those individuals in times of need. Based on prior experiences in significant relationships, people are presumed to form internal working models (or mental representations) about the availability and responsiveness of close others. These working models are thought to underlie attachment styles, which include secure, anxious/ambivalent, and avoidant attachment styles. Hazan and Shaver (1987) defined secure individuals as those who feel comfortable getting close to and depending on others, anxious/ambivalent individuals as having a strong desire to get close to others but at the same time fearing rejection, and avoidant individuals as being uncomfortable getting close or depending on others. These different attachment styles have been shown to affect the way adults interact in their romantic relationships and may also affect how they perceive their romantic partners’ behaviors. Attachment style is expected to predict the extent of agreement between relationship partners (in their perception of specific interaction behaviors) because it is thought to act as a filter in the communication process (Noller, 2005). According to Noller, attachment security and insecurity affects the way people decode and encode messages. Specifically, people with secure attachment styles are more likely to agree with their partners in their decoding of both positive and negative messages, whereas people with insecure attachment styles are more likely to distort Agreement in Married Couples 9 the messages received and be relatively inaccurate in their interpretations. For example, anxiously attached individuals have the tendency to view themselves more negatively and have trouble trusting their partners as they are chronically concerned about being rejected or abandoned by others. These traits can then lead anxious individuals to distort or misinterpret actions by their partners. Avoidant individuals are comfortable without close emotional relationships; they have a strong desire for independence and self-sufficiency. They tend to suppress their emotions and distance themselves from sources of rejection (their partners). A study by Fraley (2007) showed that highly avoidant people (especially dismissive-avoidant) use defensive strategies that prevent encoding of attachment-related information that could make them emotionally vulnerable. If this is the case, then avoidant people may miss much of the attachment-related information provided in any given interaction. Thus, having either type of insecure attachment style could potentially distort an individual’s perception of their partner’s thoughts, feelings, and behaviors during an interaction, as well as make it difficult for one’s partner to interpret one’s own thoughts, feelings, and behaviors leading to less agreement between couple members. The Current Investigation The current investigation examined whether attachment style predicts the extent to which couple members agree that specific support behaviors occurred during a specific interaction. Unlike prior studies that have considered predictors of agreement, dyads in the current study were married couples who engaged in a specific discussion about one couple member’s most important goal to accomplish over the next 6 months. Then, immediately after the discussion, both couple members were asked to report the extent to which certain behaviors were enacted during the goal discussion. Agreement in Married Couples 10 We expected that using married couples as participants and having them report about a specific interaction right after its occurrence would lead to more agreement among partners than in the some of the previous studies, such as the Coriell and Cohen (1995) study. This is because there is much less of a time gap between when the support behaviors occurred and when the participants provided their report of what happened and all of our participants (married couples) are in relationships of greater intimacy. Nonetheless, we expected (as described below) that there would be individual differences in agreement. As established by Gant (1998), agreement was assessed in terms of both a difference score (representing the difference between couple member’s reports) and a categorical measure that considers the content of couple member reports as well as the magnitude of agreement.