Mate Preferences

Mate Preferences: Implicit Methods Facial Attractiveness
• One issue with self-reports is that they are unconstrained by real-world relationships
– Possible to report strong preferences for everything, including things that are opposites – Traits do not all vary in a perfectly orthogonal way; some things are going to be correlated with others, either in a positive or negative way – We say that self-report studies lack ecological validity because of this issue

Mate Preferences
• One way to address this issue is to try and design experiments that force subjects to make trade-offs in their preferences
– By allowing potential mates to have only one of two traits, it can be inferred that the one that is chosen is preferred more strongly than the one that is not

Trade-offs in Mate Preferences
• One study used this method to compare preferences for three trait clusters:
– Warmth/Trustworthiness – Attractiveness/Vitality – Status/Resources

• Vignette design held one of the three constant (at either a high or low level), then asked the subjects to chose which one of the remaining two traits they would prefer a high level of in a mate

Trade-offs in Mate Preferences

Trade-offs in Mate Preferences
• Results seem to suggest that warmth/trustworthiness most important dimension in potential mate, no matter what other combination of traits present
– Depends on context to some extent – Consistent across men and women

• Still has some issues:
– Not very ecologically valid, as there is no reason to think these sets of traits in particular should be mutually exclusive – Since only three clusters of traits used, provides very low-resolution idea of how preferences might operate – Not providing relevant input cues

Percent choosing warmth/trustworthiness over status/resources

Percent choosing attractiveness/vitality over warmth/trustworthiness

(Fletcher et al, 2004)


Budget Allocation
• A different approach that still forces trade-offs, but may have other advantages is the budget allocation method:
– Essentially get a budget of ‘mating dollars’ to spend on building a mate – Allocate some number of dollars to each trait to buy higher levels of that trait – Trade-offs are implicit; money spent on one trait is money that is unavailable to spend on other traits – Allows for more comprehensive analysis of relevant traits than trade-off studies alone – Can vary budgets from very restricted to very generous – Can argue that people might really have something akin to different budgets based on their own attributes

Budget Allocation

(Li & Kenrick, 2006)

Budget Allocation
• Notice that patterns are slightly different than those of pure self-report
– Kindness still at or near the top – Spending on attractiveness suggests greater importance – See large difference in preference for resources

Mate Preference
• To avoid issues of ecological validity, need to make sure preference mechanisms can evaluate mates that are representative of real people • To avoid issues associated with self-report, need a method that does not rely on knowledge of and accuracy in reporting preferences
– Ideally need something implicit that allows preference mechanisms to operate (measure outputs) and then to try and backwards engineer *how* they operate

• Still suffers from lack of ecological validity
– No rules on what sort of mate can be created; participants free to spend money on two traits that might be negatively correlated in real-life without constraint – Can produce preferences that do not map onto real people

Policy-Capturing Methodology • Present a series of individuals who vary on specific traits to raters, have raters judge their attractiveness as mates. Then test correlation between traits and mate attractiveness ratings. • Can use multiple regression to test influence of multiple traits simultaneously (e.g., how important is kindness when physical attractiveness is held constant?)

Perceptual Policy Capturing
• Use policy-capturing logic and facial photographs of real people • Rate the photos on how appealing the person is as a relationship partner • Also rate photos on a number of traits, such as kindness, intelligence, physical attractiveness • Use trait ratings to infer which were most important in making mate judgments


Perceptual Policy-Capturing Results
• General finding: People self-report that kindness and intelligence are most important to them, but physical attractiveness is actually strongest predictor of mate attractiveness • Cues that generate mate attraction may not be those to which subjects have conscious access

Policy Capturing – Other Designs
• Shuar (hunter-gatherer group in Bolivia: Pillsworth et al., 2008):
– High school students who know each other and rated each other on traits like physical attractiveness and likely future wealth (good provider)
• Both sexes self-report physical attractiveness unimportant • When rank classmates for mate attractiveness, though, physical attractiveness is strong predictor for men and women • Good provider traits predict female choices better than male




What is physical attractiveness? • Do humans possess specialized mechanisms that are designed to find particular physical cues attractive? • An evolutionary approach predicts that any such cues should have signaled health, fertility, or other beneficial traits (e.g., resource acquisition abilities, parenting abilities) in ancestral environments • Can compare this position against null hypothesis that attractive traits are arbitrary (‘beauty in eye of beholder’)

Early Emergence of Attractiveness Judgments
• Had been generally argued that children learn norms of attractiveness over first few years of life • Langlois et al. (1987) exposed infants to pairs of photos of the faces of adult women in which one had been prerated (by adults) as attractive and the other pre-rated as unattractive • Video camera filmed the infants, and raters scored fixations to the left and right sides

10 9 8 7 6 5 4 6-8 mnths 2-3 mnths Attractive Unattractive

Early Emergence of Attractiveness Judgments

• Langlois et al. (1990) – 12 month olds played more and had more positive affect when in presence of adult stranger in attractive vs. unattractive mask • Infants played more with attractive vs. unattractive doll • So infants seem to find interaction with attractive people (attractive faces) more rewarding than unattractive people

Looking time (secs)


Agreement on Attractiveness Judgments • Many studies show that raters from distinct cultures generally show good agreement on judgments of facial attractiveness (correlations between .50 and .90) • Infant and cross-cultural findings suggest may be species-typical frames for categorization of facial stimuli in terms of attractiveness • Case for specialized attractiveness perception mechanisms would be much stronger if can show that features people find attractive do predict health, fertility, or other reproductive benefits

• Fluctuating Asymmetry (FA) – Asymmetry of the two sides of traits that are symmetrical on average in the population • FA indexes “developmental instability” – because two sides are regulated by same genes, asymmetry implies disruption of the developmental program for traits • Causes of developmental instability: pathogens (bacteria, viruses), plant toxins, mutations, nutritional deficits • Benefits of choosing symmetrical partners – Indirect benefits – Symmetry may index possession of genes well-adapted to local parasites, genes able to resist toxins, lack of mutations – Direct benefits – Symmetry may indicate better overall condition and thus greater ability to provide resources, compete against other individuals, etc. • Nonhuman species: FA predicts growth rate, survival probability, and mating success

Symmetry and Facial Attractiveness
• Different techniques converged on finding that more symmetrical faces found more attractive in both men and women (e.g., measuring left and right feature sizes and comparing) • Construct chimeras (combine left w/ left and right w/ right to create symmetrical faces
– Mealey et al. (1999): Identical twins for whom genes are same but developmental stability may vary – Rated twins’ faces for attractiveness, also rated similarity of left/left and right/right chimeras – Similarity ratings of left/left and right/right faces predicted attractiveness ratings – More similar L/L and R/R combinations suggest those faces came from individuals who were more symmetrical

Symmetry and Facial Attractiveness
• Another way to investigate importance of symmetry is to manipulate symmetry of photographs
– Measure coordinates of facial features on left and right sides of face, then move those features so that they are at the average position of the two – Produces faces that are exactly the same on the left and right sides (so totally symmetrical) – Compare attractiveness of manipulated faces with unmanipulated faces

Facial Averageness
• Possible benefits associated with averageness:
– May signal greater heterozygosity (different alleles of genes inherited from mother and father)
• Not in-bred • Greater genetic diversity in immune system genes • Absence of mutations that distort feature sizes

Independent Effects of Averageness and Symmetry
• Averaged composites tend to be very symmetrical: is this why averageness is more attractive? • Rhodes et al. (1999): Created composite average faces from 24 faces. Then warped real faces toward or away from the average, making them more or less average w/o making symmetrical • Also artificially manipulated symmetry of same images

– Stabilizing selection: average traits sometimes closer to functional optimum
• Too large/too small selected against as inefficient

• Langlois & Roggman (1990): Composite faces formed by averaging shape and texture of multiple faces are judged more attractive than the individual faces


More Attractive than Average
• Average faces are attractive and are more attractive than most individual faces • …but are not the most attractive faces • Faces can be made more attractive than average by changing features in specific ways • Composites made from attractive vs unattractive faces are themselves more attractive

Facial Feminization
• Perrett et al. (1998) created composite average male and female faces • Then “feminized” face by exaggerating differences between average female and average male face; “masculinized” face by reducing difference between face and average masculine face • Subjects consistently prefer feminized face shapes to average female face shapes • Suggests feminizing improves on attractiveness over averageness • Similar studies with masculinization of male faces, but these results more complicated

Sexual Dimorphism and Sex Hormone Markers • Masculinity and femininity may signal different qualities than symmetry (developmental stability) and averageness (heterozygosity)
– So may contribute independently to judgments of facial attractiveness – What causes masculinity/femininity?
• Sex hormones

• Cues of sex hormone concentrations may signal other qualities:
– Women: estrogen signaling fertility – Men: testosterone signaling immunocompetence/general condition