24 ways to be compatible with your relationship partners: Sex differences, context effects, and love styles ☆

Whether two people can get along is likely an important consideration in mate choice but it is relatively neglected by researchers. We provided an initial exploration of the concept of romantic/sexual compatibility by asking participants ( N = 274) how different/similar they want to be from their partners on 153 items that we reduced to 24 factors (e.g., appearance, conformity, leisure) which had good fit and metric invariance and (some) scalar invariance. To understand individual differences in scores on these factors, we considered participant's sex, relationship seriousness (i.e., casual vs. romantic), and love styles. Similarity was considered more important by women and in serious relationships. Participants preferred similarity in specific factors consistent with their love styles, like romanticism if one is high in eros. Results are discussed in relation to evolutionary and sociocultural models of mate choice.

"complementarity hypothesis" is scarce and assortative mating is common (Luo, 2017), the concept and the mechanisms of compatibility are ripe for research because existing research is often sociological in nature (Malouff et al., 2010) and compatibility may be also be part of a larger, evolved system of mate choice (e.g., compatibility may increase reproductive success; Apostolou & Christoforou, 2021;Luo, 2017;Wu et al., 2020).
Couples tend to be alike in age, level of education, political opinions, religious views, and intelligence (Gonzaga et al., 2007;Watson et al., 2004).This assortative mating could be an effect of mere propinquity (Bossard, 1932;Ellsworth, 1948;March & Jonason, 2023) and social homogamy (Luo & Klohnen, 2005).Therefore, partners might be alike because people with a similar background, living near one another are more likely to encounter one another.We predict that participants will, therefore, show a preference for similarity for the indexes of class, opinions, religion, and intellect (H1).Secondly, women are choosier and more oriented towards long-term relationships in mate selection than men (Kenrick et al., 1990;Li & Meltzer, 2015).Evolutionary (Buss & Schmitt, 2019) and social (Shoemake, 2007) psychologists agree that women are more cautious when it comes to mating, but they differ in ☆ Authors' notes: the second author is supported by the Charles University Research Centre program UNCE/HUM/025 (204056), the second and the third authors are supported by the Czech Science Foundation grant (23-05379S), the third author was partially funded by a grant from the National Science Centre of Poland (2019/35/B/HS6/00682).
their reasoning with the former pointing to asymmetries in minimum obligation to offspring characteristic of female mammals whereas the latter pointing to the social sanctions women suffer for engaging in sex (e.g., the sexual double standard; Koehn & Jonason, 2018).For these reasons we predict that women will prefer similarity more than men do (H2).Moreover, because there is a greater need for compatibility in serious relationships given their length, seriousness, and potential offspring (Buss & Schmitt, 2019), we predict that similarity will be preferred more in the long-term than in the short-term context (H3).
Furthermore, people's perceptions of compatibility could be linked to their love styles because love styles operate like personality traits affecting people's life choices and values (Michalska et al., 2023).There may be six love styles (i.e., Eros, Ludus, Storge, Pragma, Mania, and Agape) which represent different attitudes and beliefs towards love and predict narrow band attitudes and behaviors in relationships (Hendrick & Hendrick, 1987).People's ideas of compatibility might be linked to their love styles.Specifically, those characterized by the eros love style (i.e., strong, passionate) may want a partner like them in terms of romanticism because they value emotive aspects of relationships (H4), whereas people characterized by the pragma love style (i.e., similar in terms of background and lifestyle) may be looking for like-minded partners especially in terms of lifestyle, family, and food because they value functional aspects of relationships (H5).
In this exploratory study, we ask participants if they prefer someone similar or different in a wide range of characteristics for long-term and short-term relationships.We draw on a wide literature on relationship preferences to create a taxonomy of compatibility and to explore individual differences therein.We attempt to understand latent patterns in them through factor analyses.Then we examine sex differences, mating context effects, and individual differences in loves styles to these preferences.

Participants & procedure
Data was collected from 345 Italian volunteers from Facebook groups, SurveyCircle, and snowball sampling university students. 1The necessary sample size was determined based on the average effect size in personality psychology (r ≈ 0.20; Gignac & Szodorai, 2016) and guidelines (N ≈ 250) set for reducing estimation error in personality psychology (Schönbrodt & Perugini, 2013).We removed 71 participants for missing data or failed attention checks.Our final sample consisted of 274 participants (49 men, 225 women), aged 19 to 64 years old (M = 27.89,SD = 8.39) who were mostly heterosexual (84 %), single (38 %), and never married (54 %).Participants were first informed about the nature of the study, provided tick-box consent, and were then asked to complete a between-subjects (i.e., to reduce fatigue) survey.Upon completion participants were thanked, debriefed, and again asked if they consented to the use of their data after learning what the study was about (three opted out).This study was approved by the Ethical Committee for Psychological Research at the University of Padua (#4500), and the data (along with appendixes) can be found on the Open Science Framework.2

Measures
Participants were provided an ad hoc list of 153 items (OSF Appendix A) describing potentially similar or different characteristics between partners taken from the literature on compatibility and assortative mating (OSF Appendix B).Participants were asked how much like them (1 = very different; 7 = very similar) they preferred their partner to be for each item for a long-term (n = 152) or short-term (n = 122) relationship.Examples of items are "Current city of residence", "Jealousy", and "Music genre".These items were factor analyzed as reported in the Results.
To measure love styles, we used the Italian translation (Agus et al., 2018) of the 24-item Love Attitudes Scale-Short Form (Hendrick et al., 1998).This scale is divided into six subscales (four items each) representing the love styles of Eros (e.g., My partner and I have the right physical "chemistry" between us), Ludus (e.g., I enjoy playing the "game of love" with my partner and a number of other partners), Storge (e.g., Our love is the best kind because it grew out of a long friendship), Pragma (e.g., One consideration in choosing my partner was how he/she would reflect on my career), Mania (e.g., When my partner doesn't pay attention to me, I feel sick all over), and Agape (e.g., I would rather suffer myself than let my partner suffer).Participants indicated how much they agreed (1 = strongly disagree; 5 = strongly agree) with the items which were averaged into indexes of Eros (α = 0.68), Ludus (α = 0.49), Storge (α = 0.81), Pragma (α = 0.65), Mania (α = 0.58), and Agape (α = 0.70).

Results
We conducted exploratory and confirmatory factor analyses on the 153 items which were rated in short-and long-term contexts based on the preferred level of similarity.We pooled the short-and long-term contexts together and conducted an exploratory factor analysis using principal axis factoring with Promax rotation (Kaiser-Meyer-Olkin = 0.63; Bartlett's χ 2 [11628] = 22,108.06,p < .001).The analysis yielded 45 factors with eigenvalues greater than one, explaining 73.88 % of the variance (OSF Appendix C).We then conducted confirmatory factor analysis, testing the model fit of seven factors at the same time, yielding 24 factors with good fit (OSF Appendix D) and measurement invariance across relationship context (OSF Appendix E) when we only included factors with at least two items on them in the EFA.
We examined the role of context (Table 1) and participant's sex (Table 2) in accounting for individual differences in compatibility preferences independently given the small sample of men.Participants, with reference to the long-term context more than to the short-term context, found a partner compatible when similar in lifestyle, opinions, romanticism, morals, family, food, religion, conformity, leisure, residence, and enthusiasm.Differently, in the short-term context, more than to the long-term context, people perceived a partner compatible when similar in appearance and intellect.Moreover, men, more than women, found a partner compatible when similar in emotions and activity, while women found a partner compatible when similar in lifestyle, opinions, morals, conformity, appearance, and empathy more than men.As for love styles, women (M = 2.43, SD = 0.75) were more pragmatic (t[272] = − 3.31, p < .001,Hedges' g = 0.53) in their love style than men (M = 2.03, SD = 0.77) in, while men (M = 3.17, SD = 0.79) were more agapic (t[272] = 5.74, p < .001,g = 0.91) in theirs than women (M = 2.52, SD = 0.70).
In Table 3 we report the correlations between love styles and the compatibility indexes.Overall, people who were high in eros indicated a preference for similarity in romanticism, morals, conformity, appearance, empathy, humor, and enthusiasm, while people who were high in ludus preferred difference in residence.People who were characterized by storge showed a preference for similarity in emotions, appearance, and job and for difference in lifestyle, opinions, and origins.People who were high on pragma preferred similarity in lifestyle, sociality, romanticism, morals, family, food, religion, conformity, appearance, empathy, and intellect.People who were characterized by mania showed a preference for similarity in romanticism and for difference in sensation and conformity.Lastly, those characterized by agape showed a preference for difference in opinions, sociality, sensation, class, conformity, appearance, empathy, speech, and intellect.
We also explored whether these correlations differed by relationship And last, we explored within-sex differences in the two relationship contexts with the same alpha-correction in women only (insufficient men in the sample).The correlation between preference for similarity in lifestyle (z = − 3.05, p = .001)and ludus was stronger in the short-term context (r[98] = 0.19) than in the long-term context (r[127] = − 0.22).The correlation between preference for similarity in activity (z = − 2.96, p = .002)and storge was also stronger in the short-term context (r[98] = 0.33) than in the long-term context (r[127] = − 0.06).Lastly, the correlation between their preference for similarity in romanticism (z = 3.13, p = .001)and pragma was stronger in the long-term context (r [127] = 0.25) than in the short-term context (r[98] = − 0.17).

Discussion
Compatibility is a key factor in mate choice not only because it affects the quality and overall satisfaction of a relationship, but also because of its potential adaptive importance (Luo, 2017).Compatible partners are more likely to create a stronger emotional bond, communicate better, and cooperate more (Luo, 2009).In turn, their parental investment should be higher, and their offspring's health and longevity may improve (Wu et al., 2020).Nevertheless, research on relationships has mainly focused on characteristics that attract potential partners or affect couple's success (Joel et al., 2020;Malouff et al., 2010).Moreover, studies on compatibility have factors such as age, personality traits, physical appearance, and education (Lampis et al., 2018;Watson et al., 2004).In this study we investigate what features men and women want to be compatible on and explore corresponding individual differences in love styles.
First, participants desired a partner who was similar in opinions, morals, conformity, lifestyle, and residence (H1).Participants could be looking for a similar partner in opinions, morals, and conformity not Table 1 Internal consistency (Cronbach's α/Pearson's r) estimates and descriptive statistics for the 24 ways to be compatible across mating context.only because it is convenient when raising a child but also because it is psychologically rewarding (Luo, 2009).It might also be that they prefer similar partners as a mere effect of propinquity (March & Jonason, 2023).In fact, people are more likely to meet people from the same background and people from the same background tend to share similar values, opinions, and attitudes (Watson et al., 2004;Luo & Klohnen, 2005;Wu et al., 2020).Therefore, the preference for similarity in opinions, morals, conformity, lifestyle, and residence might be an effect of passive assortment, reflecting influences of social homogamy rather than active selection.Interestingly, participants did not express a preference for similarity for those elements which usually display positive assortment within couples (e.g., level of education, religious views, intelligence, etc.).Second, women preferred a similar partner more than men for opinions, morals, conformity, empathy, lifestyle (tentatively), and appearance (tentatively; H2), while men only preferred similarity more than women in activity and emotions (tentatively).This result is consistent with parental investment theory in that women must invest more in raising any offspring that might be created from a sexual episode and have been over evolutionary time, they are also more selective than men when considering a potential partner (Kenrick et al., 1990;Li & Meltzer, 2015).Moreover, in this study choosiness was higher for the major characteristics that seem especially likely to influence childrearing practices such as lifestyle, opinions, and morals.Having similar opinions, and lifestyles may encourage emotional and cooperative bonds between parents, facilitating child-rearing and reducing potential conflicts as well (Wu et al., 2020).
Third, participants preferred a similar partner more when referring to a long-term relationship, rather than a short-term relationship (H3).Specifically, in the long-term more than in the short-term context, they preferred a similar partner in lifestyle, opinions, romanticism, morals, food, conformity, residence, enthusiasm, family (tentatively), religion (tentatively), and leisure (tentatively), whereas in the short-term context they only preferred similarity more as for appearance and intellect (tentatively).This is consistent with the fact that individuals become more selective when choosing a partner for a long-term relationship because fewer resources are invested and there is less chance for a child to be created (Li & Meltzer, 2015;Thomas et al., 2020).However, this could also be an artifact because people perceive long-term relationships as more acceptable which may inflate ratings in that context.Short-term relationships may be less acceptable because of strong religious beliefs (Rowatt & Schmitt, 2003), 3 greater chances of unwanted pregnancy and sexually transmitted diseases (Tadinac & Hromatko, 2006), and increased risk of reputational damage (Apostolou, 2009), all of which could be normative forces (Koehn & Jonason, 2018) driving down men and especially women's responses to questions about them.
With reference to love styles, women were more pragmatic in their love style than and men were more agapic; these sex differences replicate prior effects (Dion & Dion, 1993;Worobey, 2001).In addition, participants were more likely to prefer similarity or difference in specific features, consistent with their love style.Participants high in eros (i.e., pursuing a passionate, intense love) preferred similarity in romanticism (tentatively; H4), while those high in ludus (i.e., game-playing love) preferred difference in residence (tentatively).Those high in pragma (i.e., practical view of love, looking for a suitable partner) preferred similarity in most characteristics, even more practical ones such as family, food, and lifestyle (tentatively; 5); alternatively, those high in mania (i.e., obsessive, dependent love) only preferred similarity in romanticism (tentatively).Surprisingly, those high in agape (i.e., all-giving, selfless love) preferred difference in many characteristics, even relevant ones such as opinions, appearance, and conformity (tentatively).Loves styles might be general relationship styles which then predict downstream effects in people's loves, such as mate and similarity preferences (Michalska et al., 2023).The relationship between preferences for similarity and love styles is relatively new and we only conducted exploratory tests, therefore, more work is warranted.

Limitations & conclusions
While ours is one of the only studies to consider compatibility between partners in a broad way, our study was characterized by several limitations.First, the study relied on an especially WEIRD-albeit Italian-sample of people who were volunteers and survey-takers on Table 3 Correlations between the 24 ways to be compatible and Love Attitudes.various online data collection portals.If men are higher in selfish traits like psychopathy and narcissism than women are and these traits reduce the probability of volunteering (i.e., an altruistic act), we might expect men to be under-represented in this kind of research.Moreover, if women are more likely to choose people-oriented professions like psychiatry (Kashicki et al., 2023), they may be over-represented because we sampled students from psychology.Second, we focused on love styles, but there are myriad other traits we could consider like attachment, authenticity, and personality traits and pathologies.Third, some of our measures (e.g., the ludus love style, the mania love style) had low rates of internal consistency, even when considering more the liberal threshold for basic research (Schmitt, 1996) which may be related to our sampling method, linguistic concerns in psychometrics, the way the love styles measure was created (i.e., items/factors created to serve a theory), and the brevity of the scale (i.e., α is correlated with the number of items of a scale).And last, our measure of compatibility (1) may be more about similarity preferences than whether people are "compatible", (2) has not been validated externally, (3) may be factor-loaded to women's definition of compatibility despite showing sex differences, (4) may be subject to experimenter bias, and (5) have inflated our Type 1-error which we attempted to address in the Results for our moderation tests and in the Discussion by highlighting the "tentative" (ps < 0.05 to 0.01) nature of such effects.Future research might adopt these 24 compatibility features but not necessarily the items themselves.
Despite these limitations, our study provides further understanding of romantic and sexual compatibility and differences by sex and context therein.Future studies should (1) replicate these results, (2) better examine the factorial structure of compatibility and the correlation between compatibility and satisfaction, and (3) examine how perceptions of compatibility are related to other individual differences like attachment, self-esteem, and relationship beliefs.For now, we have provided new insights into the broad range of factors that may bear on people's evaluations of the compatibility of their partners and individual differences in those ratings.

CRediT authorship contribution statement
AM conceptualized, conducted analyses, and wrote the paper.ZC conducted factor analysis.PKJ advised on the analyses and assisted in the writing the paper.
context and sex independently constituting 12 tests for each of the 24 traits.We reduced Type 1 error inflation within each compatibility index by treating 0.003 as the new threshold (without such a correction we had 34 cases of moderation out of 288 potential tests [11.81 %]).Participants high in ludus preferred a different partner in lifestyle (z = − 2.80, p = .003)in the long-term context (r[152] = 0.18) more than in the short-term context (r[122] = − 0.05).Lastly, participants high in storge preferred a similar partner as for activity (z = − 2.83, p = .002)in the short-term context (r[122] = 0.28) more than in the long-term (r [152] = − 0.06).Men (r[49] = 0.41) more than women (r[225] = − 0.02) preferred a similar partner in class (z = 2.81, p = .003)when characterized by storge, however, given the low number of men in the sample, these results should be interpreted with caution.

Table 2
Descriptive statistics and sex differences and the 24 ways to be compatible.