Courtship

Abstract

Courtship is a form of sexual behavior: the MeSH-indexed process through which people search for, evaluate, and choose romantic partners. Cognitive psychology treats it not as instinct but as a decision problem solved under severe limits on time, information, and attention — a case of bounded rationality rather than exhaustive optimization. Mate search runs as a sequential, satisficing process in which an aspiration threshold, not a ranked survey of everyone, decides when to stop looking, while judgments of romantic interest are read rapidly from thin slices of nonverbal behavior. Most revealing of all, the ideal-partner preferences people confidently state predict surprisingly little about who actually attracts them face to face. Three demonstrations trace satisficing mate search, thin-slice accuracy, and the gap between stated and revealed preference.

Keywords: courtship, mate choice, decision-making, bounded rationality, attention

Ask someone what they want in a partner and they will answer without hesitation: a list of traits, ranked, each with a confident weight. Watch the same person actually fall for someone and the list rarely predicts the outcome. That gap — between the specification we can state and the choice we actually make — is the central puzzle of courtship for cognitive psychology, and it is not a sign that people are confused about their own desires. It is a sign that mate choice is computed by processes the chooser has little introspective access to: a search that stops at good enough rather than best, a perception that reads interest from fleeting nonverbal cues, and a valuation that behaves quite differently in the abstract than in the presence of a real person.

That is why courtship is a revealing case for the science of decision-making and not only for evolutionary biology. The chooser faces a genuinely hard computational problem — an unknown number of candidates encountered one at a time, no way to go back, incomplete information about each, and a strong deadline imposed by attention, memory, and life history (Todd et al., 2007). The mind does not solve it by ranking the field; it solves it with fast, frugal heuristics tuned to the structure of the search, and the signatures of those heuristics — satisficing thresholds, rapid thin-slice reads, preference reversals between word and encounter — are exactly what the courtship literature keeps finding (Lenton & Francesconi, 2010; Eastwick et al., 2014).

What Courtship Is

Courtship, in the sense indexed by MeSH and used across the behavioral literature, is the set of behaviors through which an individual attracts, assesses, and secures a mate — a specific form of sexual behavior, the parent descriptor under which the classification files it. Its biological function is obvious; its cognitive interest is that it forces a decider to solve an unusually hard problem with ordinary, resource-limited machinery. The chooser must sample candidates, infer hidden qualities from sparse and sometimes deceptive signals, compare against some internal standard, and commit — all while time, attention, and memory impose hard limits on how much can be weighed at once.

The behavioral picture that motivates a cognitive reading is a persistent mismatch. People hold articulate, confident beliefs about the kind of partner they want, and those beliefs show lawful, cross-culturally patterned structure (Buss, 1989; Walter et al., 2020). Yet when the same people meet real candidates, those stated priorities predict who attracts them far more weakly than anyone expects (Eastwick & Finkel, 2008). The failure is not that the beliefs are insincere. It is that the processes generating a stated preference and the processes generating an actual choice are not the same system, and need not agree (Todd et al., 2007).

A second feature makes courtship a natural laboratory for decision research: the search has the formal structure of an optimal-stopping problem. Candidates are met one at a time, commitment forecloses further search, and there is no master list to rank. A chooser cannot optimize over a field they cannot see all at once, so the question is not which heuristic the mind should use but which one it does — and whether that heuristic is well matched to the real statistics of the mating environment (Lenton & Francesconi, 2010).

How Courtship Intersects Cognition

Courtship touches cognitive psychology at three points, each a different part of the same decision. First, search: how a chooser samples a sequence of candidates and decides when to stop, which is a satisficing problem solved with an aspiration threshold rather than full optimization. Second, perception: how a chooser reads another person's romantic interest and mate value from brief, mostly nonverbal cues — a thin-slicing problem in social perception. Third, valuation: how the preferences a chooser can state relate to the ones that actually move their choice, which turns out to be far looser than intuition allows. Table 1 lays out the three, and the three interactive demonstrations below take them in turn.

Table 1

Three Cognitive Problems the Courting Mind Must Solve

Note. The three problems are facets of one decision, not stages in a fixed sequence; a chooser searches, perceives, and values continuously and in parallel. MeSH files Courtship (D003380) under Sexual Behavior (F01.145.802.279) and records no narrower descriptors beneath it, so the organizing structure here is cognitive rather than taxonomic.

Mate search: satisficing under bounded rationality

The first cognitive problem is when to stop looking. Because candidates are encountered sequentially and commitment forecloses further search, mate choice has the formal shape of an optimal-stopping problem, and no resource-limited mind can solve it by optimization: a field that is never seen all at once cannot be ranked. What people do instead is satisfice — sample a few candidates to estimate the range available, set an aspiration threshold, and commit to the first prospect who clears it (Todd et al., 2007). This is a fast-and-frugal heuristic in the tradition of bounded rationality: it ignores most of the information an optimizer would demand and still performs well, because it is tuned to the real structure of the search environment rather than to an idealized one.

The decisive finding is that this search process is dissociable from stated preferences. When Todd and colleagues compared what people said they wanted with the choices they made in a speed-dating-style sequence, the two were driven by different information: mate choices tracked the chooser's own mate value and a simple aspiration rule, not the elaborate trait weightings they reported (Todd et al., 2007). The abundance of options matters here too. As the number of candidates grows, people do not search more thoroughly; they shift toward faster, more heuristic, cue-based screening, trading depth of evaluation for speed — a rational adaptation to attention and time limits, but one that changes what gets selected (Lenton & Francesconi, 2010). Demo 1 makes the satisficing rule concrete: candidates arrive one at a time, an aspiration threshold is set from an initial learning sample, and the first candidate to clear it is chosen — showing how threshold height trades the risk of stopping too early against the risk of searching too long.

Thin-slicing: reading interest from brief nonverbal cues

The second problem is perception: deciding whether another person is interested, and estimating their qualities, from sparse and fleeting cues. Courtship runs heavily on nonverbal signaling — Moore's observational work catalogued a large repertoire of women's nonverbal solicitation displays and showed that their frequency, not mere physical presence, predicted which approaches followed (Moore, 1985). The cognitive point is that these signals are read fast. Ambady and Rosenthal's meta-analysis established that thin slices — exposures of under five minutes, sometimes seconds — support judgments of interpersonal states and outcomes at accuracy well above chance, and that lengthening the exposure adds surprisingly little (Ambady & Rosenthal, 1992). Applied to courtship, observers shown brief clips of strangers interacting can judge who is romantically interested in whom at better-than-chance rates, reading the same nonverbal channel the participants themselves are using (Place et al., 2009).

Thin-slice accuracy is real but bounded, and that boundedness is the cognitively interesting part. The judgment is a signal-detection problem: interest cues are noisy, genuine interest and ordinary friendliness overlap, and both false alarms (seeing interest that is not there) and misses (failing to see interest that is) carry real costs. Accuracy above chance does not mean accuracy near ceiling, and the same brief exposure that enables a quick read also caps how good that read can be. Demo 2 treats interest detection as exactly this signal-detection task, with a movable decision criterion along which hits and false alarms trade off against each other as the thin slice gives only partial information.

Valuation: why stated preferences miss the real person

The third problem is the one that most unsettles intuition. People can state detailed ideal-partner preferences, and those preferences are structured, sincere, and cross-culturally patterned (Buss, 1989; Fletcher et al., 1999). Yet a comprehensive meta-analysis found that the match between a person's stated ideals and the traits of a partner they actually evaluate predicts their romantic evaluation at close to zero once the two have met in person — stated ideals forecast attraction to a described profile, but not to a real individual (Eastwick et al., 2014). The original demonstration came from speed dating: before the event, men and women reported the familiar sex-differentiated priorities, but what actually predicted their in-event desire bore no relation to those stated weights (Eastwick & Finkel, 2008).

This is not evidence that preferences are meaningless, but that stated and revealed preferences are computed differently. Preferences specified in the abstract operate on attributes — a checklist applied to a hypothetical partner; choices made in person operate on an integrated, holistic impression of a particular someone, in which the same attributes behave differently or drop out entirely (Todd et al., 2007). Machine-learning work sharpens the point: trained on everything participants reported about themselves and their ideals before meeting, models predicted general tendencies to desire and be desired, but the specific compatibility between two particular people — the thing a preference is supposed to capture — was essentially unpredictable from either person's stated preferences (Joel et al., 2017). Simulations that let stated preferences drive mate choice can reproduce real couple formation only when preferences are integrated across the whole person rather than weighted trait by trait (Conroy-Beam & Buss, 2016). Demo 3 shows the dissociation directly: a stated-ideal checklist and an in-person holistic impression are scored for the same candidates, and the two rankings come apart as the weighting shifts from attributes to the integrated whole.

Worked Example

Model mate search as an optimal-stopping problem to see why satisficing, not optimizing, is the sensible rule. Suppose a chooser will meet N = 20 candidates, one at a time, each with a quality drawn independently and uniformly from 0 to 100, and that commitment is final: accept a candidate and the search ends; reject one and they are gone for good. The chooser cannot wait for the best of all 20 because they never see all 20 at once.

A good heuristic is the aspiration-threshold rule: observe and reject the first m candidates to learn the range, then accept the first later candidate who exceeds the best quality seen in that learning sample. The classic analysis sets the learning fraction at about 1/e ≈ 37%, so the chooser samples m ≈ 0.37 × 20 ≈ 7 candidates, fixes the threshold at the best of those seven, and then takes the first of the remaining 13 who beats it. With quality uniform on 0–100, the expected best-of-seven sample threshold is 7/8 × 100 = 87.5, so

threshold θ ≈ 87.5, and the chooser commits to the first later candidate scoring above it.

The rule's logic is a trade-off between two errors. Set the learning sample too short and the threshold is low, so the chooser commits early to a mediocre prospect — stopping too soon. Set it too long and the threshold is so high that no remaining candidate clears it, so the chooser exhausts the field and must settle for whoever is last — searching too long. The 37% rule balances these, and here it yields roughly a 38% chance of landing the single best candidate and a high chance of landing one of the top few — far better than the 1/20 = 5% a random pick would give, and achieved without ever ranking the full field. Crucially, the threshold is learned from the chooser's own early sample, not read off a list of stated ideals: this is why real mate choice tracks an aspiration level and recent experience rather than the trait weightings people report (Todd et al., 2007; Lenton & Francesconi, 2010). Demo 1 runs exactly this rule, varying N and the learning fraction to show the two errors trade off.

Discussion

Reading courtship as a decision problem does two useful things. It dissolves the apparent paradox that people seem not to know what they want: the beliefs are sincere and the behavior is lawful, but they are outputs of different systems, so a weak link between stated preference and actual choice is expected rather than puzzling (Todd et al., 2007; Eastwick et al., 2014). And it locates courtship within the general science of bounded rationality rather than treating it as a sealed-off instinct: the same satisficing, cue-based heuristics and the same divergence between what a chooser can articulate and what actually drives choice appear across consumer choice, hiring, and risk — thrown into sharp relief here by the sequential, high-stakes, one-at-a-time structure of the mating environment (Lenton & Francesconi, 2010). The practical implication follows: because the binding constraints are attention, search cost, and the difficulty of forecasting compatibility, environments that change those constraints — online dating's vast, profile-filtered option pools above all — should change courtship cognition, and they do, pushing choosers toward faster, more heuristic, attribute-based screening of the very kind that predicts in-person attraction least well (Finkel et al., 2012).

Two cautions keep the account honest. First, the modern research consensus is deflationary about prediction: large-scale machine-learning work finds that neither partner's self-reported traits nor their stated preferences predict the quality of the relationship they will have, which is driven far more by how each person perceives the relationship once in it than by any pre-match attribute (Joel et al., 2020). Courtship cognition selects a partner; it does not thereby compute the relationship's future. Second, the decision framing describes the machinery of choosing within a mating market whose structure — who is available, what is valued, what signals are honest — is set by culture and ecology, and the cross-cultural regularities in stated preferences are real and must be explained even as their weak behavioral predictiveness is acknowledged (Buss, 1989; Walter et al., 2020). Naming the heuristic that chose is not the same as naming the world that shaped it.

Current Directions

The most active and most humbling front is predictive modeling. Pooling 43 longitudinal studies of couples, machine-learning analyses found that relationship quality is modestly predictable from how people describe the relationship itself, but essentially not predictable from the individual traits or stated preferences that courtship is supposed to screen on — a result that reframes mate choice as selecting an entry point rather than forecasting an outcome (Joel et al., 2020). A second front revisits the stated–revealed gap with better methods, asking whether ideals predict anything once the measure is functional rather than nominal, and whether people's patterns of preference (how they trade attributes off) carry signal their levels do not (Eastwick et al., 2014; Conroy-Beam & Buss, 2016). A third examines how the digital mating market reshapes the search itself: when the option pool is effectively unbounded and pre-filtered by searchable attributes, the satisficing threshold, the role of thin-slice perception, and the point at which people stop looking all shift, with consequences the first generation of online-dating research only began to map (Finkel et al., 2012; Lenton & Francesconi, 2010). The open question uniting them is whether courtship is best improved by better matching before the encounter or by protecting the holistic, in-person perception that stated preferences so poorly approximate.

Common Misconceptions

People know what they want in a partner, so their stated preferences should predict their choices.

Stated ideals are sincere and structured, but they predict in-person attraction at close to zero. What people can articulate (a weighted checklist of attributes) and what drives an actual choice (an integrated impression of a particular person) are computed by different processes (Eastwick et al., 2014; Eastwick & Finkel, 2008).

Choosing a mate means optimizing — finding the best available partner.

No resource-limited mind can optimize over candidates met one at a time with no going back. People satisfice instead: they set an aspiration threshold from early experience and commit to the first prospect who clears it, a fast-and-frugal rule well matched to the search environment (Todd et al., 2007).

More options mean better choices.

A larger option pool shifts people toward faster, more heuristic, cue-based screening rather than more thorough evaluation. Abundance changes the cognition of choosing — often toward exactly the attribute-based filtering that predicts in-person attraction least well (Lenton & Francesconi, 2010; Finkel et al., 2012).

Romantic interest cannot be judged quickly or from the outside.

Brief, thin-slice exposures to nonverbal behavior support judgments of courtship signaling and romantic interest well above chance, and even outside observers can read who is interested in whom — though accuracy is bounded, not perfect (Ambady & Rosenthal, 1992; Place et al., 2009).

Glossary

Aspiration threshold.

The internal standard a satisficing searcher sets from early experience and then applies, committing to the first candidate who exceeds it rather than ranking the whole field (Todd et al., 2007).

Attention.

The selective allocation of limited processing to some cues over others; one of the hard limits that forces courtship to be solved by heuristics rather than exhaustive evaluation.

Bounded rationality.

The view that real deciders use heuristics matched to their environment and their limited time, memory, and information, rather than optimizing — the framework within which mate search is a satisficing problem.

Courtship.

The MeSH-indexed set of behaviors through which an individual attracts, assesses, and secures a mate; treated here as a sequential decision problem in search, perception, and valuation.

Decision-making.

The process of selecting among options by weighing their values; in courtship the weighing is sequential, satisficing, and only loosely related to the preferences the chooser can state.

Ideal-partner preferences.

The traits a person reports wanting in a mate; structured and sincere, but weak predictors of who actually attracts them in person (Fletcher et al., 1999; Eastwick et al., 2014).

Matching hypothesis.

The proposal that people pair with others of similar overall desirability; its experimental origins lie in early dating studies of physical attractiveness (Walster et al., 1966).

Mate choice.

The act of selecting a romantic or sexual partner from available candidates; driven by an aspiration-threshold search and a holistic in-person impression rather than by a trait-by-trait checklist (Todd et al., 2007).

Optimal stopping.

The class of problems in which options are evaluated sequentially and a decision to stop is irreversible; the formal shape of mate search, solvable only by heuristic thresholds, not full optimization.

Revealed preference.

The preference inferred from what a person actually chooses rather than from what they say; in courtship it diverges sharply from stated ideals once a real partner is encountered (Eastwick & Finkel, 2008).

Satisficing.

Choosing the first option that meets an acceptability threshold rather than searching for the best possible one; the core heuristic of mate search under bounded rationality (Todd et al., 2007).

Social perception.

The reading of other people's states and dispositions from their behavior; in courtship it supplies the thin-slice judgments of interest and mate value on which the decision partly rests.

Thin-slicing.

Forming accurate-above-chance judgments from very brief samples of behavior; applied to courtship, the rapid read of romantic interest from a few seconds of nonverbal cues (Ambady & Rosenthal, 1992).

Visceral signaling.

The nonverbal courtship displays — glances, posture, proximity, gesture — through which interest is communicated and read, catalogued in observational studies of solicitation behavior (Moore, 1985).

Key Researchers

David M. Buss

(University of Texas at Austin, Department of Psychology). Established the cross-cultural structure of human mate preferences in a landmark 37-cultures study, framing courtship as evolved mate choice and providing the stated-preference data whose weak behavioral predictiveness later work had to explain. ORCID · Faculty page · Wikipedia

Paul W. Eastwick

(University of California, Davis, Department of Psychology). Showed through speed-dating research that the sex-differentiated ideal-partner preferences people state fail to predict their in-person romantic interest, and co-authored the meta-analysis placing the ideal–actual correspondence near zero. ORCID · Google Scholar · Faculty page

Eli J. Finkel

(Northwestern University, Department of Psychology and Kellogg School of Management). Led the critical analysis of online dating from the perspective of psychological science, examining how algorithmic matching and vast option pools reshape the cognition of partner search and choice. Google Scholar · Faculty page

Elaine Hatfield

(University of Hawaii at Manoa, Department of Psychology). As Elaine Walster, led the 1966 computer-dance study that founded the experimental psychology of attraction and dating, establishing physical attractiveness as the dominant in-person predictor and seeding the matching hypothesis. Official site · Wikipedia

Samantha Joel

(Western University, Department of Psychology; Relationship Decisions Lab). Applied machine learning to show that initial romantic desire is hard to predict from stated preferences, and that relationship quality is driven far more by perceived partner responsiveness than by either partner's pre-match traits. ORCID · Google Scholar · Faculty page

Peter M. Todd

(Indiana University Bloomington, Cognitive Science Program and Psychological and Brain Sciences). Reframed mate search as a fast-and-frugal, satisficing, sequential decision process, demonstrating that the heuristics driving actual mate choices are distinct from the preferences people state. Google Scholar · Faculty page

Frequently Asked Questions

How does decision-making work in courtship and mate choice?

Courtship is a sequential decision problem: candidates are encountered one at a time, commitment forecloses further search, and there is no master list to rank. Because no resource-limited mind can optimize over a field it never sees all at once, people satisfice instead — they sample a few prospects to gauge what is available, set an aspiration threshold, and commit to the first candidate who clears it. This fast-and-frugal rule is well matched to the real structure of the mating environment and performs far better than chance without ever evaluating the full field.

Why do stated partner preferences fail to predict who we actually date?

Because stated and revealed preferences are computed by different processes. A stated ideal is a weighted checklist applied to a hypothetical partner, whereas an in-person choice rests on an integrated, holistic impression of a particular someone. A large meta-analysis found that the correspondence between stated ideals and actual in-person attraction is close to zero: the preferences are sincere and structured, but they simply are not the system that drives choosing once a real person has been met.

Can people accurately judge romantic interest from brief nonverbal cues?

Yes, above chance but not perfectly. Research on thin-slicing shows that exposures of a few seconds to a few minutes support judgments of interpersonal states, including courtship signaling, at accuracy well above chance, and that longer exposure adds surprisingly little. Even outside observers watching brief clips can read who is romantically interested in whom. Accuracy is genuinely bounded, though, because interest cues are noisy and ordinary friendliness overlaps with genuine interest, so both false alarms and misses are common.

How does mate search relate to bounded rationality and satisficing?

Mate search is a textbook case of bounded rationality. The chooser faces limited time, memory, and information and a sequential, irreversible search, so optimization is impossible and heuristics take over. The core heuristic is satisficing: set an acceptability threshold from early experience and take the first option that beats it. Research comparing what people say they want with the choices they make finds that actual mate choices track this aspiration rule and the chooser's own mate value, not the elaborate trait weightings they report.

Does having more dating options lead to better choices?

Not straightforwardly. As the number of candidates grows, people do not search more carefully; they shift toward faster, more heuristic, cue-based screening, trading depth of evaluation for speed. This is a rational response to attention and time limits, but it pushes choosers toward exactly the attribute-based filtering that predicts in-person attraction least well. Large online option pools, filtered by searchable traits, intensify this effect and can change what gets selected without improving the match.

How does perception and attention shape attraction during courtship?

Attraction depends heavily on where attention is directed and what perception extracts from brief encounters. Courtship runs on nonverbal signaling — glances, posture, proximity — that perceivers read rapidly as thin slices, and the limited attention available in a real encounter is part of why choice relies on holistic impressions rather than exhaustive trait analysis. Because perception integrates cues into a single impression of a person, the attributes a chooser could list in the abstract often behave quite differently, or drop out entirely, when a real individual is perceived.

Does courtship affect future romantic decision-making?

Yes, the influence runs both ways. Each encounter updates the aspiration threshold that governs the next: a chooser learns the range of what is available and recalibrates what counts as good enough, so past experience shapes future stopping decisions. Over time this feedback means early samples disproportionately set the standard a person applies later. Decision-making leads to the courtship behavior, and the outcome of that behavior feeds back to reshape the thresholds and expectations driving subsequent choices.

What does cognitive science say about predicting relationship success from courtship?

It is strikingly deflationary. Pooling dozens of longitudinal couple studies, machine-learning analyses find that relationship quality is modestly predictable from how people describe the relationship once they are in it, but essentially not predictable from either partner's individual traits or stated preferences — the very things courtship is supposed to screen on. The lesson is that courtship cognition selects an entry point into a relationship; it does not thereby compute the relationship's future, which depends far more on how the partners later perceive and treat each other.

References

Ambady, N., & Rosenthal, R. (1992). Thin slices of expressive behavior as predictors of interpersonal consequences: A meta-analysis. Psychological Bulletin, 111(2), 256–274. https://doi.org/10.1037/0033-2909.111.2.256

Buss, D. M. (1989). Sex differences in human mate preferences: Evolutionary hypotheses tested in 37 cultures. Behavioral and Brain Sciences, 12(1), 1–49. https://doi.org/10.1017/S0140525X00023992

Conroy-Beam, D., & Buss, D. M. (2016). Do mate preferences influence actual mating decisions? Evidence from computer simulations and three studies of mated couples. Journal of Personality and Social Psychology, 111(1), 53–66. https://doi.org/10.1037/pspi0000054

Eastwick, P. W., & Finkel, E. J. (2008). Sex differences in mate preferences revisited: Do people know what they initially desire in a romantic partner? Journal of Personality and Social Psychology, 94(2), 245–264. https://doi.org/10.1037/0022-3514.94.2.245

Eastwick, P. W., Luchies, L. B., Finkel, E. J., & Hunt, L. L. (2014). The predictive validity of ideal partner preferences: A review and meta-analysis. Psychological Bulletin, 140(3), 623–665. https://doi.org/10.1037/a0032432

Finkel, E. J., Eastwick, P. W., Karney, B. R., Reis, H. T., & Sprecher, S. (2012). Online dating: A critical analysis from the perspective of psychological science. Psychological Science in the Public Interest, 13(1), 3–66. https://doi.org/10.1177/1529100612436522

Fletcher, G. J. O., Simpson, J. A., Thomas, G., & Giles, L. (1999). Ideals in intimate relationships. Journal of Personality and Social Psychology, 76(1), 72–89. https://doi.org/10.1037/0022-3514.76.1.72

Joel, S., Eastwick, P. W., & Finkel, E. J. (2017). Is romantic desire predictable? Machine learning applied to initial romantic attraction. Psychological Science, 28(10), 1478–1489. https://doi.org/10.1177/0956797617714580

Joel, S., Eastwick, P. W., Allison, C. J., et al. (2020). Machine learning uncovers the most robust self-report predictors of relationship quality across 43 longitudinal couples studies. Proceedings of the National Academy of Sciences, 117(32), 19061–19071. https://doi.org/10.1073/pnas.1917036117

Lenton, A. P., & Francesconi, M. (2010). How humans cognitively manage an abundance of mate options. Psychological Science, 21(4), 528–533. https://doi.org/10.1177/0956797610364958

Moore, M. M. (1985). Nonverbal courtship patterns in women: Context and consequences. Ethology and Sociobiology, 6(4), 237–247. https://doi.org/10.1016/0162-3095(85)90016-0

Place, S. S., Todd, P. M., Penke, L., & Asendorpf, J. B. (2009). The ability to judge the romantic interest of others. Psychological Science, 20(1), 22–26. https://doi.org/10.1111/j.1467-9280.2008.02248.x

Todd, P. M., Penke, L., Fasolo, B., & Lenton, A. P. (2007). Different cognitive processes underlie human mate choices and mate preferences. Proceedings of the National Academy of Sciences, 104(38), 15011–15016. https://doi.org/10.1073/pnas.0705290104

Walster, E., Aronson, V., Abrahams, D., & Rottman, L. (1966). Importance of physical attractiveness in dating behavior. Journal of Personality and Social Psychology, 4(5), 508–516. https://doi.org/10.1037/h0021188

Walter, K. V., Conroy-Beam, D., Buss, D. M., et al. (2020). Sex differences in mate preferences across 45 countries: A large-scale replication. Psychological Science, 31(4), 408–423. https://doi.org/10.1177/0956797620904154