Testing the Canalization Hypothesis of Attachment Theory: Examining Within-Subject Variation in Attachment Security

Keely A. Dugan, R. Chris Fraley, Omri Gillath, Pascal R. Deboeck

Research output: Contribution to journalArticlepeer-review


According to the canalization hypothesis of attachment theory (Bowlby, 1973), people’s trajectories of attachment security should become increasingly stable and buffered against external pressures as their relationships progress. The present study aimed to examine this hypothesis within the context of romantic relationships. We analyzed longitudinal data collected from 1,741 adults who completed between three and 24 survey assessments (average number of waves analyzed = 6.79, SD = 5.31; median test-retest interval = 35 days). We modeled participants’ within-person fluctuations in partner-specific security as a function of their romantic relationship length. Additionally, we examined whether attachment-related events (e.g., conflict with one’s partner) predict greater within-person fluctuations in security among people involved in newer versus more established romantic relationships. Our results suggest that people in newer romantic relationships demonstrated greater fluctuations in partner-specific attachment anxiety—both generally and in reaction to attachment-related events—compared to those in well-established romantic relationships. However, neither of these trends was observed for partner-specific attachment avoidance. These results provide partial support for the canalization hypothesis but also suggest that canalization processes may be more nuanced than previously assumed.

Original languageEnglish (US)
Pages (from-to)511-541
Number of pages31
JournalJournal of personality and social psychology
Issue number3
StatePublished - Dec 7 2023


  • attachment
  • close relationships
  • longitudinal
  • romantic relationships
  • within-person variation

ASJC Scopus subject areas

  • Social Psychology
  • Sociology and Political Science


Dive into the research topics of 'Testing the Canalization Hypothesis of Attachment Theory: Examining Within-Subject Variation in Attachment Security'. Together they form a unique fingerprint.

Cite this