People tend to misjudge how emotionally secure their spouses really are, often perceiving more attachment insecurity than their partners report themselves. A new three-year study of newlywed couples suggests that, although partners accurately recognize broad attachment patterns, they systematically overestimate both attachment anxiety and attachment avoidance.
The research, published in the Journal of Social and Personal Relationships, followed 226 mixed-sex couples during the early years of marriage. It found that partners frequently project their own attachment orientation onto each other, shaping how secure or insecure they perceive their spouse to be.
What Attachment Styles Really Measure
Attachment orientations describe how people think, feel, and behave in close relationships over time. Psychologists typically assess attachment along two primary dimensions: attachment anxiety and attachment avoidance, which influence how individuals seek closeness and respond to perceived threats of rejection.
Attachment anxiety is characterized by concerns about abandonment, rejection, or not being sufficiently loved, often leading to heightened reassurance-seeking and increased sensitivity to emotional distance. Attachment avoidance reflects discomfort with emotional intimacy or dependence, along with a preference for self-reliance and emotional distance.
These dimensions are independent, meaning an individual can score high on one, both, or neither. Low levels of both anxiety and avoidance indicate secure attachment, whereas elevated levels on either or both dimensions reflect attachment insecurity. These attachment patterns can influence romantic relationships, friendships, and family dynamics.
How the Study Was Conducted
Lead author Jeewon Oh and colleagues analyzed data from the Growth in Early Marriage Project, which followed newly married couples across three assessment waves over a period of three to four years. Most participants were recruited through marriage license records from communities in Western Massachusetts, including Amherst, Northampton, and Springfield.
The study included 452 individuals. Husbands averaged approximately 29 years of age, while wives averaged between 27 and 28 years. On average, couples had been together for five years and had lived together for just over two years at the beginning of the study. Nearly nine out of ten participants were employed in paid work.
Participants completed the Experiences in Close Relationships questionnaire, first assessing their own attachment anxiety and avoidance and then completing the same questionnaire based on their perceptions of their spouse. Marital quality was assessed using the Dyadic Adjustment Scale, which measures relationship satisfaction and overall relationship functioning.
To evaluate perceptual accuracy, the researchers applied the Truth and Bias framework. This analytical approach compares one partner’s perception of the other with that partner’s self-report while also examining the extent to which individuals project their own attachment characteristics onto their spouse.
Partners See More Insecurity Than Exists
Across the sample, spouses generally perceived their partners as more anxious and more avoidant than those partners described themselves. Despite this tendency, partner ratings remained meaningfully associated with self-ratings, indicating that spouses demonstrated a reasonable degree of accuracy when identifying overall attachment patterns.
For example, individuals who described themselves as having higher attachment anxiety were also rated as more anxious by their spouses. A similar pattern emerged for attachment avoidance, suggesting that partners can generally detect relative differences even while overestimating the overall level of insecurity.
At the same time, the researchers found strong evidence for assumed similarity bias. Individuals with higher attachment anxiety or avoidance were more likely to attribute similar levels of anxiety or avoidance to their spouse, effectively projecting their own attachment orientation onto the relationship.
This projection can blur the distinction between a partner’s actual behavior and how that behavior is perceived. Individuals who are especially concerned about abandonment or emotional closeness may be more likely to perceive similar insecurities in their spouse, even when the spouse reports feeling relatively secure.
Why Relationship Quality Matters
Relationship satisfaction emerged as an important moderator of these perceptions. Participants who reported higher-quality marriages tended to overestimate their partners’ attachment anxiety and avoidance less than those in lower-quality relationships, particularly during the first two years of marriage.
Interestingly, individuals in more satisfying marriages demonstrated a stronger assumed similarity bias for attachment anxiety. They perceived their partners as more similar to themselves with respect to anxious attachment, although they still rated their spouses as less anxious overall than participants in less satisfying marriages.
Over time, there was a modest tendency for participants who felt more satisfied with their relationship to perceive slightly greater attachment anxiety in their partners, although these perceptions remained lower than those reported by participants in less satisfying marriages. This finding suggests that the idealization often present early in marriage may gradually give way to more nuanced perceptions as relationships develop.
Notably, participants in lower-quality relationships were actually more accurate in identifying their partners’ true levels of attachment insecurity throughout the study period. This finding is consistent with previous research showing that distressed partners often monitor one another more closely, sometimes sacrificing positive perceptions for greater accuracy.
Limits of Self-Report and Future Directions
The authors caution that all measures used in the study were based on self-report, raising important questions about what constitutes “true” attachment insecurity. Some individuals may underreport attachment anxiety or avoidance because of limited self-awareness or social desirability bias.
Without observational data from independent coders or physiological measures, it is difficult to determine whether partners genuinely overestimated attachment insecurity or whether individuals understated their own insecurity on the questionnaires. Future research could integrate self-report measures with behavioral observations and reports from friends or family members.
The sample also consisted exclusively of newly married couples, who may be particularly motivated to understand one another accurately while navigating major life transitions. Patterns of perception and bias may differ among couples who have been married longer, unmarried cohabiting partners, or couples experiencing more unstable relationships.
Despite these limitations, the study provides additional insight into how psychologists understand social perception within romantic relationships. It suggests that individuals can simultaneously be both accurate and biased, recognizing broad truths about their partners while filtering those perceptions through their own attachment histories.
The findings may also have practical implications for couples therapy and relationship education. Encouraging partners to recognize how their own attachment orientation shapes their interpretations of each other’s behavior may help reduce misunderstandings and promote more accurate and compassionate perceptions.
