{"title":"The Great EFT Schism: Testing Johnson's and Greenberg's Alternative Models of What Drives Couple Satisfaction","authors":"Júlia Halamová","doi":"10.1111/famp.70214","DOIUrl":null,"url":null,"abstract":"<div>\n \n <p>Emotion-focused therapy for couples (EFT-C) is organized around two theoretical accounts: Johnson's (2019) attachment-based model, which treats attachment as the primary motivational system in romantic relationships, and the Greenberg–Goldman (2008) multi-motivational model, which adds regard and attraction to attachment as three co-equal systems. Their relative explanatory power has not been directly compared. The present study tested whether the multi-motivational model explains more variance in relationship satisfaction than an attachment-only model in a non-clinical community sample. A nationally quota-representative sample of 1090 individuals nested in 545 couples completed measures of relationship satisfaction (CSI-16), attachment behaviors (BARE), interpersonal attraction (IAS), and level of regard (BLRI). Hierarchical regression and relative weights analysis compared model performance, with generalized estimating equations and cluster-bootstrap confidence intervals accounting for non-independence within couples. The attachment-only model explained 47% of variance; adding attraction and regard increased this to 70.5% (Δ<i>R</i><sup>2</sup> = 0.235, Cohen's <i>f</i><sup>2</sup> = 0.80). Contrary to expectations, regard emerged as the dominant predictor (41.3% of explained variance), followed by attraction (29.8%) and attachment (29.0%). Perceived partner regard uniquely predicted satisfaction (<i>p</i> < 0.001), whereas self-reported regard for partner did not (<i>p</i> = 0.663). Conclusions about clinical mechanisms cannot be drawn from this cross-sectional, non-clinical design. Nonetheless, the findings are inconsistent with the strong form of the attachment primacy hypothesis and consistent with the Greenberg–Goldman multi-motivational account. Comprehensive assessment of distressed couples may benefit from attending to regard and attraction alongside attachment.</p>\n </div>","PeriodicalId":51396,"journal":{"name":"Family Process","volume":"65 4","pages":""},"PeriodicalIF":2.2000,"publicationDate":"2026-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Family Process","FirstCategoryId":"102","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1111/famp.70214","RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"FAMILY STUDIES","Score":null,"Total":0}
引用次数: 0
Abstract
Emotion-focused therapy for couples (EFT-C) is organized around two theoretical accounts: Johnson's (2019) attachment-based model, which treats attachment as the primary motivational system in romantic relationships, and the Greenberg–Goldman (2008) multi-motivational model, which adds regard and attraction to attachment as three co-equal systems. Their relative explanatory power has not been directly compared. The present study tested whether the multi-motivational model explains more variance in relationship satisfaction than an attachment-only model in a non-clinical community sample. A nationally quota-representative sample of 1090 individuals nested in 545 couples completed measures of relationship satisfaction (CSI-16), attachment behaviors (BARE), interpersonal attraction (IAS), and level of regard (BLRI). Hierarchical regression and relative weights analysis compared model performance, with generalized estimating equations and cluster-bootstrap confidence intervals accounting for non-independence within couples. The attachment-only model explained 47% of variance; adding attraction and regard increased this to 70.5% (ΔR2 = 0.235, Cohen's f2 = 0.80). Contrary to expectations, regard emerged as the dominant predictor (41.3% of explained variance), followed by attraction (29.8%) and attachment (29.0%). Perceived partner regard uniquely predicted satisfaction (p < 0.001), whereas self-reported regard for partner did not (p = 0.663). Conclusions about clinical mechanisms cannot be drawn from this cross-sectional, non-clinical design. Nonetheless, the findings are inconsistent with the strong form of the attachment primacy hypothesis and consistent with the Greenberg–Goldman multi-motivational account. Comprehensive assessment of distressed couples may benefit from attending to regard and attraction alongside attachment.
期刊介绍:
Family Process is an international, multidisciplinary, peer-reviewed journal committed to publishing original articles, including theory and practice, philosophical underpinnings, qualitative and quantitative clinical research, and training in couple and family therapy, family interaction, and family relationships with networks and larger systems.