Romantic love is a complex emotion with significant evolutionary implications, motivating pair-bonding and increasing survival chances for offspring. While romantic love is relatively well-defined, the process of "falling in love" is less understood, with individual differences in how easily and often individuals experience this phenomenon. Jones (2011, 2015) introduced the concept of "emophilia," a relatively stable personality-like construct representing the ease and frequency of falling in love, measured using the Emotional Promiscuity Scale (EPS). Emophilia is considered the emotional equivalent of unrestricted sociosexuality, linked to romantic relationship formation but not necessarily a one-to-one relationship. Prior research on emophilia, however, primarily involved North American samples, highlighting the need for replication and cross-cultural validation. Scandinavian cultures may exhibit different perceptions and expressions of love compared to North American cultures, potentially influencing the associations between emophilia and romantic outcomes. This study aimed to investigate the psychometric properties of the EPS within a Scandinavian sample, examine its discriminant validity against established personality traits (Big Five and Dark Triad), and explore its predictive validity concerning the number of romantic relationships and instances of infidelity.

The literature review highlights existing research on romantic love, emphasizing its evolutionary significance in pair-bonding. It differentiates romantic love from the process of "falling in love," noting individual differences in this experience. Jones's work on emophilia is presented, describing it as a stable construct related to the ease and frequency of falling in love, measured by the EPS. The EPS, despite its two-factor structure (ease and frequency), often uses a composite score due to the high correlation between factors. The close relationship between emophilia and romantic relationship formation is discussed, comparing it to the concept of sociosexuality. The need for further research, including cross-cultural studies, is emphasized due to the limited previous research and the replication crisis in psychology. Existing studies show some inconsistent associations between emophilia and Big Five and Dark Triad traits, motivating a more comprehensive investigation in a Scandinavian context.

Data were collected from 2607 participants (74.6% women, 24.7% men, 0.7% other) recruited through online articles in Norwegian (VG+) and Swedish (Aftonbladet+) newspapers. Participants completed an online survey including measures of emophilia (EPS), the Big Five personality traits (Mini-IPIP), the Dark Triad traits (Dirty Dozen), demographics (age and gender), and romantic outcomes (number of romantic relationships and instances of infidelity). A confirmatory factor analysis (CFA) was conducted to assess the psychometric properties of the EPS, examining its factor structure and internal reliability. Correlation analyses explored the discriminant validity of the EPS against personality traits. Negative binomial regression analyses investigated the predictive validity of emophilia concerning the number of romantic relationships and infidelity, controlling for age, gender, and personality traits. Measurement invariance was assessed using multigroup confirmatory factor analysis (MGCFA) across gender and age groups.

Descriptive analyses of the EPS revealed satisfactory internal reliability and generally normally distributed responses, except for one item. CFA initially indicated a two-factor structure (easily and often falling in love), but a bifactor model provided a better fit, suggesting a general emophilia factor with additional specific aspects. The EPS demonstrated satisfactory discriminant validity against personality traits, with mostly small correlation coefficients (except for a medium correlation with narcissism). Negative binomial regression analyses revealed positive associations between emophilia and both the number of romantic relationships and instances of infidelity, although the effect sizes were small. Adjusting for age, gender, and personality traits did not substantially alter these associations, with the exception of unfaithfulness which showed slightly reduced effect in fully adjusted model. No significant interaction effect was found between gender and the relationship between emophilia and romantic outcomes.

The findings largely support the hypotheses. The EPS demonstrated good psychometric properties in the Scandinavian sample, similar to previous findings in North American samples. Discriminant validity was demonstrated against personality traits, suggesting emophilia is not merely a facet of existing personality traits. However, small effect sizes suggest a complex interplay with other factors. Predictive validity was also supported, with positive associations between emophilia and relationship formation/infidelity. The small effect sizes highlight the multifactorial nature of these outcomes. Potential confounders, like the directionality of the relationships between emophilia and the number of romantic relationships/unfaithfulness, are discussed. This study strengthens the generalizability of previous emophilia research, but also emphasizes the need for additional research to investigate potential cultural differences fully.

This study confirms the psychometric soundness of the EPS and demonstrates the discriminant and predictive validity of emophilia, showing positive associations with romantic relationships and unfaithfulness. However, the small effect sizes highlight the complexity of these relationships. Future research should address limitations, particularly by clarifying definitions of romantic relationship and unfaithfulness, utilizing longitudinal designs, investigating additional relevant constructs (e.g., attachment style, cognitive factors), and expanding across diverse cultural contexts.

The convenience sampling method might have introduced selection bias, limiting generalizability. Reliance on self-report data increases susceptibility to recall bias, social desirability bias, and common method variance. The lack of metric invariance across age groups is a limitation. The cross-sectional design prevents conclusions about causality.