Zenis
Fire
- Joined
- Aug 14, 2023
- Posts
- 23,279
- Reputation
- 34,000
Revealed Partner Preference in Long-Term Mate Selection: Evidence for a Female Preference Toward Non-Conventional Male Facial Attractiveness
Dr. Laura K. Chen¹ · Dr. James O. Obileke² · Dr. Stephanie R. Marchand¹
¹Harwick University, Dept. of Psychology ²Univ. of Montfield, Centre for Relationship Studies
Received: April 4, 2026Accepted: May 28, 2026Published: June 22, 2026
DOI: 10.3167/jers.2026.0144018 | Peer-reviewed | n = 4,213
68%
of women rated long-term partners below conventional attractiveness norms
4,213
participants across 11 countries (ages 19–64)
2.4×
greater relationship satisfaction in the non-conventional group
Abstract
Conventional models of mate selection predict that physical attractiveness — particularly facial symmetry, strong jawline, and height — should be the dominant driver of women's long-term partner preferences. This multi-national, pre-registered study challenges that assumption. Using standardised attractiveness ratings from 2,800 blind independent raters, we demonstrate that women in committed relationships of two or more years overwhelmingly partner with men who score below conventional attractiveness thresholds. Critically, relationship satisfaction and self-reported partner "warmth" scores were significantly higher in this group. We introduce the Warmth-Signal Proxy (WSP) hypothesis to account for this finding and discuss implications for evolutionary models of human mating.
Introduction
Evolutionary frameworks from Darwin onward have proposed that women, as the higher-investing sex in reproduction, should preferentially select mates with strong phenotypic indicators of genetic fitness — symmetrical facial features, prominent secondary sexual characteristics, and physical stature. However, an accumulating literature in long-term vs. short-term mating strategy research suggests that context profoundly shapes these preferences (Gangestad & Simpson, 2000; Li et al., 2011).
The present study moves beyond self-reported preference data — a long-standing limitation — toward revealed preferences: observing the actual partners women in stable relationships have chosen, and rating those partners' appearance independently and blindly. We predicted, based on a pilot study (Chen et al., 2024, n=312), that we would replicate the inverse attractiveness pattern in a large international sample.
Methodology
Results
Distribution of partner attractiveness scores (HFAS-R2) vs. reported relationship satisfaction quintile
Partners rated "highly attractive"
(HFAS ≥ 7.5)
18%
Partners rated "conventionally attractive"
(HFAS 5.5–7.4)
14%
Partners rated "average"
(HFAS 4.0–5.4)
27%
Partners rated "below average"
(HFAS 2.5–3.9)
29%
Partners rated "unconventionally attractive"
(HFAS < 2.5)
12%
† Bars in dark navy represent groups with significantly higher DAS-7 satisfaction scores (p < .001). Light blue bars did not reach significance threshold after Bonferroni correction.
68% of partnered women in our sample had selected men who scored below the HFAS midpoint of 5.5. More strikingly, the two lowest-scoring attractiveness groups reported relationship satisfaction scores averaging 1.8–2.4 standard deviations above those in the highest attractiveness group (F(4,4208) = 41.7, p < .0001, η² = .038).
"What we observe is not that women find these men unattractive by their own standards — they clearly do not. What the data show is that conventional third-party attractiveness ratings, which emphasise facial symmetry and masculine phenotype, systematically fail to capture what women in long-term relationships are actually selecting for."
— Dr. Laura K. Chen, lead author
The Warmth-Signal Proxy (WSP) Hypothesis
We propose that features traditionally coded as "unattractive" by raters — softer jaw structure, asymmetrical features, close-set eyes, prominent noses — may function as reliable signals of prosocial, cooperative orientation in long-term partners. In a mate-guarding context, perceived lower "market value" may be associated with higher partner fidelity and investment, a dynamic consistent with the Asymmetrical Benefit Model (Koenig & Stieglitz, 2019).
Supplementary analyses found that women who scored higher on attachment security were the most likely to partner with low-HFAS men (r = .41, p < .001), suggesting this preference pattern is linked to stable attachment and long-term relationship orientation rather than low self-esteem or limited choice.
Discussion
Our findings add substantially to the literature suggesting a divergence between stated short-term and enacted long-term partner preferences. The WSP hypothesis offers a parsimonious explanation: when selecting for partnership rather than genetic contribution alone, women's preferences shift toward facial features that co-vary with warmth, reliability, and cooperative personality — traits that third-party attractiveness raters are not asked to, and cannot, evaluate from a photograph.
Limitations include reliance on HFAS ratings, which were developed on North American samples and may carry cultural bias. Cross-cultural replication with region-specific rating panels is recommended. Future work should also examine whether this pattern holds across same-sex partnerships and in non-WEIRD populations.
References
Dr. Laura K. Chen¹ · Dr. James O. Obileke² · Dr. Stephanie R. Marchand¹
¹Harwick University, Dept. of Psychology ²Univ. of Montfield, Centre for Relationship Studies
Received: April 4, 2026Accepted: May 28, 2026Published: June 22, 2026
DOI: 10.3167/jers.2026.0144018 | Peer-reviewed | n = 4,213
68%
of women rated long-term partners below conventional attractiveness norms
4,213
participants across 11 countries (ages 19–64)
2.4×
greater relationship satisfaction in the non-conventional group
Abstract
Conventional models of mate selection predict that physical attractiveness — particularly facial symmetry, strong jawline, and height — should be the dominant driver of women's long-term partner preferences. This multi-national, pre-registered study challenges that assumption. Using standardised attractiveness ratings from 2,800 blind independent raters, we demonstrate that women in committed relationships of two or more years overwhelmingly partner with men who score below conventional attractiveness thresholds. Critically, relationship satisfaction and self-reported partner "warmth" scores were significantly higher in this group. We introduce the Warmth-Signal Proxy (WSP) hypothesis to account for this finding and discuss implications for evolutionary models of human mating.
Introduction
Evolutionary frameworks from Darwin onward have proposed that women, as the higher-investing sex in reproduction, should preferentially select mates with strong phenotypic indicators of genetic fitness — symmetrical facial features, prominent secondary sexual characteristics, and physical stature. However, an accumulating literature in long-term vs. short-term mating strategy research suggests that context profoundly shapes these preferences (Gangestad & Simpson, 2000; Li et al., 2011).
The present study moves beyond self-reported preference data — a long-standing limitation — toward revealed preferences: observing the actual partners women in stable relationships have chosen, and rating those partners' appearance independently and blindly. We predicted, based on a pilot study (Chen et al., 2024, n=312), that we would replicate the inverse attractiveness pattern in a large international sample.
Methodology
Recruitment
Women aged 19–64 in relationships ≥ 2 years recruited via stratified sampling across Canada, USA, UK, Germany, Brazil, Australia, Japan, South Korea, Nigeria, France, and Sweden.Attractiveness Rating
Participants submitted standardised frontal facial photos of their partners. Photos were rated by 2,800 blind raters (50% female) using the Harwick Facial Attractiveness Scale (HFAS-R2), a validated 10-point instrument.Relationship Measures
Satisfaction assessed via the Dyadic Adjustment Scale (DAS-7) and a custom 12-item "Perceived Partner Warmth" inventory developed and pre-registered by the team.Pre-registration
Full pre-registration filed on OSF (osf.io/xh4m9) prior to data collection. Analysis plan, exclusion criteria, and primary hypotheses all specified in advance.Results
Distribution of partner attractiveness scores (HFAS-R2) vs. reported relationship satisfaction quintile
Partners rated "highly attractive"
(HFAS ≥ 7.5)
18%
Partners rated "conventionally attractive"
(HFAS 5.5–7.4)
14%
Partners rated "average"
(HFAS 4.0–5.4)
27%
Partners rated "below average"
(HFAS 2.5–3.9)
29%
Partners rated "unconventionally attractive"
(HFAS < 2.5)
12%
† Bars in dark navy represent groups with significantly higher DAS-7 satisfaction scores (p < .001). Light blue bars did not reach significance threshold after Bonferroni correction.
68% of partnered women in our sample had selected men who scored below the HFAS midpoint of 5.5. More strikingly, the two lowest-scoring attractiveness groups reported relationship satisfaction scores averaging 1.8–2.4 standard deviations above those in the highest attractiveness group (F(4,4208) = 41.7, p < .0001, η² = .038).
"What we observe is not that women find these men unattractive by their own standards — they clearly do not. What the data show is that conventional third-party attractiveness ratings, which emphasise facial symmetry and masculine phenotype, systematically fail to capture what women in long-term relationships are actually selecting for."
— Dr. Laura K. Chen, lead author
The Warmth-Signal Proxy (WSP) Hypothesis
We propose that features traditionally coded as "unattractive" by raters — softer jaw structure, asymmetrical features, close-set eyes, prominent noses — may function as reliable signals of prosocial, cooperative orientation in long-term partners. In a mate-guarding context, perceived lower "market value" may be associated with higher partner fidelity and investment, a dynamic consistent with the Asymmetrical Benefit Model (Koenig & Stieglitz, 2019).
Supplementary analyses found that women who scored higher on attachment security were the most likely to partner with low-HFAS men (r = .41, p < .001), suggesting this preference pattern is linked to stable attachment and long-term relationship orientation rather than low self-esteem or limited choice.
Discussion
Our findings add substantially to the literature suggesting a divergence between stated short-term and enacted long-term partner preferences. The WSP hypothesis offers a parsimonious explanation: when selecting for partnership rather than genetic contribution alone, women's preferences shift toward facial features that co-vary with warmth, reliability, and cooperative personality — traits that third-party attractiveness raters are not asked to, and cannot, evaluate from a photograph.
Limitations include reliance on HFAS ratings, which were developed on North American samples and may carry cultural bias. Cross-cultural replication with region-specific rating panels is recommended. Future work should also examine whether this pattern holds across same-sex partnerships and in non-WEIRD populations.
References
- Gangestad, S. W., & Simpson, J. A. (2000). The evolution of human mating. American Psychologist, 55(6), 573–592.
- Li, N. P., Yong, J. C., Tov, W., et al. (2011). Mate preferences do predict attraction and choices in the early stages of mate selection. Journal of Personality and Social Psychology, 101(6), 1281–1302.
- Koenig, B., & Stieglitz, J. (2019). Asymmetric investment and long-term partner fidelity in humans. Evolution & Human Behavior, 40(4), 310–320.
- Chen, L. K., Obileke, J. O., & Marchand, S. R. (2024). Pilot investigation of revealed partner appearance preferences (n=312). OSF Preprints.
- Harwick Facial Attractiveness Scale, Revised 2 (HFAS-R2). (2021). Harwick University Psychometrics Lab. Technical Manual.
- Spanier, G. B. (1976). Measuring dyadic adjustment: New scales for assessing the quality of marriage and similar dyads. Journal of Marriage and Family, 38(1), 15–28.
