Abstract
BACKGROUND: This study examined the domain-specific patterns associated with supraliminal attachment priming on the cognitive accessibility of attachment-related mental representations. METHODS: Seventy participants were randomly assigned to one of three experimental conditions: attachment-insecurity priming, attachment-security priming, or a non-attachment control (non-primed reference) condition. Participants underwent supraliminal priming via a guided imagery task specific to their condition, followed by a lexical decision task measuring reaction times for five word categories: proximity-related, distance-related, positive, negative, and neutral words. RESULTS: Relative differences between priming conditions emerged exclusively for attachment-related word categories. Participants in the attachment-insecurity priming condition showed faster reaction times to both proximity- and distance-related words relative to the non-primed reference condition. In contrast, participants in the attachment-security priming condition showed faster reaction times to proximity-related words than the non-primed reference condition only at low levels of attachment anxiety; no such differences were observed at higher levels of anxiety. Reaction times to distance-related words did not differ between the security priming and non-primed reference conditions. Attachment avoidance did not moderate any effects. Bayesian analyses provided affirmative evidence for the absence of priming effects in positive, negative, and neutral word categories. Given the observed effect sizes, moderation trend-level patterns should be interpreted as exploratory. CONCLUSION: These findings advance understanding of attachment system dynamics by showing that differences between insecurity and security priming in attachment-related processing depend on attachment anxiety.