A three-year study of newlywed couples found that while people generally accurately perceive their partners’ attachment patterns, they tend to overestimate their partners’ attachment anxiety and avoidance. They also often projected their own attachment orientation onto their partners. These perceptions persisted over time and were determined by the progress of the relationship. The paper was published in. Journal of social and personal relationships.
Attachment orientation is a relatively stable pattern of thinking, feeling, and behavior in intimate relationships. They are usually explained along two dimensions: attachment anxiety and attachment avoidance. Attachment anxiety refers to the tendency to worry about being rejected, abandoned, or not loved enough. People with high attachment anxiety frequently seek reassurance and may be especially sensitive to signs that their partner is becoming distant.
Attachment avoidance refers to discomfort with emotional intimacy, dependence, and dependence on others. People high in attachment avoidance may prefer to be emotionally distant and find it difficult to express vulnerability or ask for support.
These two attachment traits are thought to be independent. This means that a person can score high or low on one trait, regardless of their level of the other trait. When both anxiety and avoidance are low, this is called secure attachment. High anxiety, high avoidance, or both are commonly referred to as an insecure attachment orientation. Attachment orientation can affect romantic relationships, family relationships, friendships, and many other aspects of life.
Study author Jeewon Oh and colleagues investigated how accurate and biased people are in perceiving their partner’s attachment orientation. They also investigated how this accuracy and bias changes over time and depending on the quality of the relationship. To do this, they compared the ratings made by the participants to the participants’ partners’ own self-ratings and to the participants’ own characteristics. This method of analysis is called the truth-and-bias approach.
They analyzed data from the Early Marriage Growth Project. This was a three-wave study of newlywed mixed-sex couples, collecting data over a period of three to four years. Couples were primarily recruited from marriage license records in western Massachusetts (Amherst, Hadley, Northampton, Belchertown, South Hadley, Springfield, Pittsfield, etc.).
Data were obtained from 226 couples (i.e., 452 participants). On average, husbands were 29 years old and wives 27 to 28 years old at the start of the study. On average, the two dated for five years and lived together for just over two years. 89% of husbands and 91% of wives were employed in paid employment.
Study participants completed assessments of attachment (Experiences of Close Relationships Questionnaire) and marital quality (Dyadic Adjustment Scale). Participants answered the same attachment assessment questions about themselves and about their partners.
Results showed that participants tended to overestimate their partner’s attachment anxiety and avoidance relative to how their partner viewed themselves. However, participants’ ratings of their partners’ attachment anxiety and avoidance were still positively correlated with how their partners rated themselves, meaning that these ratings were relatively accurate in tracking partners’ own self-perceptions. For example, when people rated themselves as higher in insecure attachment, their partners also rated themselves as higher in insecure attachment. The same was true for avoidance.
Further analysis revealed that participants’ attachment ratings for their partners were strongly correlated with their own attachment ratings. This shows that the “assumed similarity bias” is influencing the judgment of the other person. For example, when people rated themselves as having high attachment anxiety, they assumed that their partner also had high attachment anxiety (i.e., they believed their partner was more similar to them than they actually were).
Interestingly, participants who reported higher levels of satisfaction with their marriage tended to overestimate their partner’s anxiety and avoidance, although to a lesser extent. However, due to avoidance, this effect was only present during the first two years of marriage. Those who reported better relationship quality also tended to exhibit higher similarity bias toward anxiety. Although people who reported higher quality of their relationships were much less likely to rate their partners as more anxious over time, they still rated their partners as less anxious overall than participants who reported lower relationship quality.
Importantly, the researchers also found that people in low-quality relationships consistently “perceived more accurately” their long-term partners’ true levels of attachment anxiety than those in happier marriages.
“While people generally had accurate perceptions of their partner’s attachment, they often overestimated their partner’s attachment anxiety or avoidance and projected their own attachment orientation. These perceptions persisted over time and depended on how the relationship was progressing,” the study authors concluded.
This study contributes to scientific knowledge regarding social cognition in romantic relationships. However, the study authors point out that newlyweds may be highly motivated to be accurate about their partners, as people who reflect on important relationship milestones tend to be accurate. Results may vary for couples with different marital statuses and lengths of relationship.
Furthermore, because this study relies on self-report and does not take into account behaviors observed by a third party, it is difficult to determine whether partners are truly overestimating their anxiety or whether spouses are simply underreporting their anxiety due to a lack of self-awareness.
The paper, “Truth and bias in partners’ perceptions of long-term attachment orientation: The moderating role of relationship quality,” was authored by Jeewon Oh, Mariah F. Purol, William J. Chopik, Fiona Ge, Sally I. Powers, and Paula R. Pietromonaco.
