The height of choosiness: Mutual mate choice for stature results in suboptimal pair formation for both sexes

Abstract

Mutual mate choice is prevalent in humans, where both males and females have a say in their choice of partner. How the choices made by one sex constrain the choice of the other remains poorly understood, however, because human studies have mostly limited themselves to measuring preferences. We used a sample of 5782 speed-daters making 128 104 choices to link preferences for partner height to actual choice and the formation of a match (the mutual expression of interest to meet again). We show that sexual conflict at the level of preferences is translated into choice: women were most likely to choose a speed-dater 25 cm taller than themselves, whereas men were most likely to choose women only 7 cm shorter than themselves. As a consequence, matches were most likely at an intermediate height difference (19 cm) that differed significantly from the preferred height difference of both sexes. Thus, our study reveals how mutual mate choice can result in suboptimal pair formation for both sexes, highlighting the importance of assessing the mate choice process in its entirety.

Full text

The height of choosiness: mutual mate choice for stature results in suboptimal
pair formation for both sexes
Gert Stulp
a
,
b
,
*
, Abraham P. Buunk
a
,
c
, Robert Kurzban
d
,
e
, Simon Verhulst
b
a
Department of Psychology, University of Groningen, Groningen, The Netherlands
b
Department of Behavioural Biology, University of Groningen, Groningen, The Netherlands
c
Royal Netherlands Academy of Arts and Sciences, Amsterdam, The Netherlands
d
Department of Psychology, University of Pennsylvania, Philadelphia, U.S.A.
e
Department of Economics, University of Alaska, Anchorage, U.S.A.
article info
Article history:
Received 19 December 2012
Initial acceptance 22 January 2013
Final acceptance 11 March 2013
Available online xxx
MS. number: 12-00965
Keywords:
human
mutual mate choice
pair formation
sexual selection
speed dating
stature
Mutual mate choice is prevalent in humans, where both males and females have a say in their choice of
partner. How the choices made by one sex constrain the choice of the other remains poorly understood,
however, because human studies have mostly limited themselves to measuring preferences. We used a
sample of 5782 speed-daters making 128 104 choices to link preferences for partner height to actual
choice and the formation of a match (the mutual expression of interest to meet again). We show that
sexual conflict at the level of preferences is translated into choice: women were most likely to choose a
speed-dater 25 cm taller than themselves, whereas men were most likely to choose women only 7 cm
shorter than themselves. As a consequence, matches were most likely at an intermediate height dif-
ference (19 cm) that differed significantly from the preferred height difference of both sexes. Thus, our
study reveals how mutual mate choice can result in suboptimal pair formation for both sexes, high-
lighting the importance of assessing the mate choice process in its entirety.
Ó2013 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.
Finding a suitable mate to form a reproductive unit is complex,
owing to the many factors that prevent an individual from
obtaining his or her preferred partner. First, mates with the desired
properties might not be available and, even if they are, individuals
might have insufficient time to assess all available possible mates
(Reynolds & Gross 1990;Widemo & Sæther 1999;Fawcett &
Johstone 2003;Cotton et al. 2006). Second, some desired charac-
teristics might trade off against each other; for instance, attrac-
tiveness might trade off against willingness or ability to provide
parental investment (Magrath & Komdeur 2003); obtaining a mate
with the desired level of both characteristics might, as a conse-
quence, be impossible.
Relatedly, other individuals’pursuit of their own interests can
impair mating with preferred individuals. In many species,
including humans, mating is a two-sided affair: individuals who
prefer a given partner must themselves be chosen as a mate by that
individual (Johnstone et al. 1996;Baldauf et al. 2009). In addition,
third parties, especially same-sex rivals, can interfere with
obtaining one’s desired mates (Wong & Candolin 2005). Further-
more, even successful pair formation (i.e. pair bonding) always
entails the risk that, at some point in the future, the partner may
move to a more attractive alternative (Rusbult & Buunk 1993). For
these and other reasons, any given individual’s mate preferences
are unlikely to be completely satisfied.
In part because of the difficulty of tracking choice and pairing,
the study of mate choice has focused to a large extent on measuring
preferences (Courtiol et al. 2010b). How preferences translate to
actual choices and subsequent pairing remains unclear. One win-
dow onto the relationships between preferences, choice and pair-
ing is so-called speed-dating events. During a speed-dating event,
participants meet approximately 10e30 individuals in a series of
3e7 min ‘dates’after which they discretely indicate whether they
are interested in further contact (‘Yes’/‘No’). When a ‘Yes’is recip-
rocated, they make a ‘Match’, and contact details are subsequently
provided to enable participants to arrange a more traditional date if
desired (Kurzban & Weeden 2005,2007;Finkel & Eastwick 2008;
Lenton & Francesconi 2011). Although such ‘matches’do not inev-
itably lead to the formation of an actual relationship, people who
were matched with at least one person during speed dating had a
*Correspondence: G. Stulp, Department of Psychology, Grote Kruisstraat 2/1,
9712 TS, Groningen, The Netherlands.
E-mail address: g.stulp@rug.nl (G. Stulp).
Contents lists available at SciVerse ScienceDirect
Animal Behaviour
journal homepage: www.elsevier.com/locate/anbehav
0003-3472/$38.00 Ó2013 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.anbehav.2013.03.038
Animal Behaviour xxx (2013) 1e10
Please cite this article in press as: Stulp, G., et al., The height of choosiness: mutual mate choice for stature results in suboptimal pair formation
for both sexes, Animal Behaviour (2013), http://dx.doi.org/10.1016/j.anbehav.2013.03.038

10.9% chance of engaging in sexual intercourse with a ‘match’
within 6 weeks of the event, while the chance of a more serious
relationship after 1 year was 7.2% (Asendorpf et al. 2011). Thus,
speed dating is an ecologically relevant setting to study pair
formation.
Data from speed-dating events have some advantages over self-
report questionnaire- or vignette-based studies, having greater
ecological validity and allowing a look at the effects of mutual mate
choice. More importantly for the present purpose, speed dating
allows researchers to determine how mate preferences, self-
reported indications of what individuals want in a mate, translate
into the choices that individuals actually make, and how these
choices translate into subsequent potential pairing. We thus treated
the speed-dating venue as a ‘model system’that enabled us to
interrogate human mate choice processes in a manner directly
comparable to those of other species (Lenton et al. 2009). To this
end, we operationalized definitions related to preference, choice
and pairing as used in the mate choice literature (Fowler-Finn &
Rodríguez 2012a,b) for use within a speed-dating context,
focusing on partner height as a preference variable (see Table 1).
Previous studies that have addressed the interplay between
preferences, choice and pairing in speed dating have shown that
stated preferences are generally poor predictors of choice, in that
many ‘nonpreferred’individuals are also chosen (Kurzban &
Weeden 2007;Todd et al. 2007;Eastwick & Finkel 2008;
Eastwick et al. 2011). Preferences also fail to predict which potential
mates are pursued after a speed-dating event (Eastwick & Finkel
2008). Furthermore, choices made during speed-dating events
were only weakly reciprocated between partners (Luo & Zhang
2009;Back et al. 2011).
The present analysis has several advantages over previous work.
First, we examined preferences, choice and pairing simultaneously.
Second, we focused on one trait, height, which is a particularly
useful trait to study because: (1) it is an easily verified objective
measure (in contrast to, e.g. kindness or reported income); (2) both
sexes show height preferences (Courtiol et al. 2010a;Stulp et al.
2013b); (3) partner heights correlate positively (Spuhler 1982;
Stulp et al. 2011) and men are taller than their partner more often
than expected by chance alone (Gillis & Avis 1980;Stulp et al.
2013a), indicating that pairing with respect to height is
nonrandom; and (4) both male and female heights are related to
the number of children produced (Stulp et al. 2012a,b,c), indicating
that pair formation with respect to height can affect reproductive
success and thereby has evolutionary relevance. Another advantage
of our study is that a clearly defined partner preference was
available (i.e. preferred partner height), allowing a direct compar-
ison with the response to heights. This compares favourably to
previous studies, where preferences have most commonly been
measured using a subjective scale (e.g. rate on a scale how
important physical attractiveness is in an ideal romantic partner;
see Kurzban & Weeden 2005 for a notable exception). Finally,
because we could combine the specific preferences and choices of
both sexes simultaneously, we were able to assess potential con-
flicts over partner height, and so examine how mutual mate choice
affects final pairing.
Previous work indicates that preference functions for height in
both sexes do not align, creating a sexual conflict over partner
height (Baldauf et al. 2009;Courtiol et al. 2010a; see Table 1). The
present work first reproduced this finding, and, subsequently, we
tested (1) whether stated preferences for partner height translated
into actual choice during speed dating and (2) whether height was
related to responsiveness (while others might use terms such as
‘selectivity’or ‘choosiness,’we use this term to connect with the
animal literature) and desirability. Based on the preferences and
choices of speed-daters, we determined both the strength of pref-
erence and tolerance with respect to height (Table 1), and exam-
ined how these depended on a person’s sex and own height. Finally,
we tested whether (3) the conflict between the sexes over stated
height preferences affected choice and pair formation.
METHODS
Speed Dating
We used data collected by HurryDate, a firm organizing speed-
date events across North America. The procedure and data have
been described elsewhere (Kurzban & Weeden 2005,2007). In
short, men and women are invited in groups of usually upto 50 and
with an approximately equal sex ratio. Events are stratified by age
(25e35 and 35e45 are typical). During an event, all men interact
with all women for 3 min per date after which both parties
discretely register their interest in the other person by indicating
either ‘Yes ’or ‘No’on a designated scorecard. These are then stored
by HurryDate and checked for ‘matches’: cases in which both male
and female indicated ‘Yes ’to one another. Subsequently, partici-
pants are informed who their matches are, can view these in-
dividuals’online profiles, and send emails to their matches. Our
sample consisted of single men and women paying a fee to attend
the event, indicating that these individuals were genuinely
searching for a mate (and contrasts with many other studies in
which speed-daters received a reward for participating in the form
of, for instance, money or course credits e.g. Eastwick & Finkel
2008;Luo & Zhang 2009;Eastwick et al. 2011). HurryDate col-
lects survey data from their participants including their own height
and a preferred height range (i.e. a minimal and maximal preferred
height).
During a HurryDate event, women usually remain seated while
the men change positions. Given this pattern, women’s height may
Table 1
Definitions of preference measures, choice and pairing drawn from the literature and the operational definitions used in a speed-dating context
Variable General (short) definition Operational definition
Preference ranking The ranking of mates based on the trait value
with respect to likelihood of mating
The stated minimal and maximal preferred height
Strength
*
The degree to which deviations from the ideally
preferred trait value are disfavoured
The decrease in the probability of responding ‘Yes’to a
speed-dater whose height deviates from the chooser’s
acceptable height range preference
Responsiveness
*,y
The probability that an individual will respond
positively to any mate, independently of trait value
The probability of responding ‘Yes’to any speed-dater
encountered during an event, independently of their height
Tolerance
*
The range of trait values considered acceptable
by a choosing individual
The standard deviation of the mean of those heights to
which a ‘Yes’response was given
Choice Positive response to sampled mates Whether a given speed-dater gave a ‘Yes’response
Pair formation The formation of a pair to reproduce Whether a ‘Yes’response was reciprocated, and a ‘Match’formed
*
Based on Fowler-Finn & Rodríguez (2012a,b).
y
In the speed-dating literature often referred to as ‘selectivity’or ‘choosiness’.
G. Stulp et al. / Animal Behaviour xxx (2013) 1e102
Please cite this article in press as: Stulp, G., et al., The height of choosiness: mutual mate choice for stature results in suboptimal pair formation
for both sexes, Animal Behaviour (2013), http://dx.doi.org/10.1016/j.anbehav.2013.03.038

be more difficult for men to assess than vice versa. However, before
the speed-dating event starts, the speed-daters spend several mi-
nutes interacting while standing, allowing assessments of height.
Moreover, height is also readily assessed from cues while sitting, as
standing height correlates strongly with both sitting height
(r¼0.94) and arm length (r¼0.94; Torres et al. 2003). In addition,
the face can also be used as a cue to height (Re & Perrett 2012). Thus
we consider it safe to assume that men had sufficient opportunity
to assess the height of their female dates.
Although human populations can differ substantially in terms of
height, previous studies using a wide range of methodologies and
various populations (Pawlowski 2003;Fink et al. 2007;Courtiol
et al. 2010a;Re & Perrett 2012;Stulp et al. 2013b) have all yiel-
ded the same consistent mate preferences with respect to height.
Yet, studies from non-Western samples suggest that these height
preferences are not universal (Sear & Marlowe 2009;Sorokowski &
Butovskaya 2012). U.S. citizens are particularly diverse in their
ethnic background, even among those considered Caucasian. The
preferences for height observed in this study should therefore not
be considered as ‘universal’preferences for stature.
When using such an ethnically diverse sample one has to bear in
mind that assortment for ethnic background may lead to assort-
ment for height, as ethnic background and height are correlated.
Yet, because of this correlation, it is very hard to determine the
causal arrow of this relationship. Indeed, a preference for certain
heights will also lead to assortment for ethnic background. An
interesting example of the entanglement between height and
ethnicity comes from patterns of interethnic marriage: marriages
between black men and white women and between Asian women
and white men are more frequent than marriages between black
women and white men and between Asian men and white women,
and these asymmetries are best understood in terms of height
(Belot & Fidrmuc 2010). Nevertheless, in an attempt partially to
account for homogamy with respect to ethnicity, we added a var-
iable to all our logistic mixed models (see below) that coded
whether or not speed-daters were of similar ethnic background.
Including this variable in our analyses did not change our results
qualitatively (i.e. a significant term for height (or height difference)
never became nonsignificant; these results are not reported here).
Ethical approval was obtained from the Institutional Review
Board at the University of Pennsylvania. Informed consent was not
deemed necessary by this Board, as the participants engaged in
‘public behaviour’. The data obtained could not be traced back to
the individual.
Sample
We included all events in which full information was available
for all choices made by all participants in that event (i.e. full in-
formation on who said ‘Ye s’to whom). We excluded all events in
which (1) one of the individuals said ‘Yes ’to an unknown individ-
ual; (2) when a ‘Match’was reported even though both individuals
had not said ‘Yes ’to one another; (3) when ‘Yes ’was said to an
individual of the same sex (HurryDate sessions are specifically
designed for heterosexuals); and (d) when the total number of
participants in the event was lower than 15. This gave us a total of
174 speed-dating events with full information on who said ‘Ye s’to
whom in which 5782 individuals (N¼3024 females) made 128 104
choices, resulting in 9072 matches.
Analyses
All analyses were performed separately for the two sexes. We
examined the individual preferences for partner height and how
these related to an individual’s height using Pearson correlations,
and ttests were calculated to examine sex differences (using
Cohen’sdas our measure of effect size). We examined whether
height (or differences in height) affected the chance of either giving
or receiving a ‘Yes ’response using mixed models with binomial
error distribution, in which individuals of both sexes and ‘event’
were included as random effects (i.e. three random effects in total).
The Wald Ztest was used for determining Pvalues for the param-
eter estimates in the mixed models. When examining quadratic
terms of height (or height difference), we always included the
linear term in the statistical model. When the quadratic term was
nonsignificant, we present the statistical details from the model
with only the linear term of height. Height and preferences for
height were reported in inches, so we used this unit of measure-
ment in all analyses, but for the graphs we converted these data to
centimetres. All analyses were performed using the lme4 package
in R, version 2.13.1 (R Development Core Team 2008). All percent-
ages mentioned in the Results are predictions from mixed models
based on the fixed effects, which were calculated based on the
formula in Diggle et al. (2002). Confidence intervals of optima were
based on 1000 reanalyses of the data using the functions simulate
and refitinR.
RESULTS
Overall Sample
An average of 36.97 (SD ¼10.82) individuals participated per
event in the 174 speed-dating events, in which an average of
18.17 5.18 were women and 18.80 6.06 were men. Average
height for men was 179.06 (SD ¼6.87) cm (154 men did not report
height), and 165.20 (SD ¼6.72) cm for women (172 women did not
report height). See Appendix Tables A1 and A2 for more descriptive
statistics.
Stated Preferences for Partner Height
Preference ranking with respect to height was studied using the
minimum and maximum preferred height. Men were more likely
(761 of 2601; 29.26%) than women (167 of 2847; 5.87%) to report a
very low minimally preferred height (4 feet z122 cm;
c
2
1
¼526:28, P<0.0001). In contrast, women were more likely
(844 of 2847; 29.65%) than men (623 out of 2601; 23.95%) to report
a very high maximally preferred height (7 feet z213 cm;
c
2
1
¼22:39, P<0.0001). We considered the very low minimally
(4 feet) and very high maximally (7 feet) preferred heights to
indicate that there was no limit to the height of an acceptable
partner, and therefore excluded these individuals from the
following analysis. The preferred height range (maximally
preferred minus minimally preferred height) was larger in men
than in women (men: mean SD: 24.43 8.43 cm; N¼17 70 ;
women: 18.72 7.08 cm; N¼1996; t
3470.13
¼22.33, P<0.0001,
d¼0.74). Height correlated positively with minimally and maxi-
mally preferred height in both sexes (Fig. 1; men: minimum:
r
1820
¼0.35, P<0.0001; maximum: r
1955
¼0.52, P<0.0001;
women: minimum: r
2651
¼0.40, P<0.0001; maximum:
r
1981
¼0.42, P<0.0001).
Women preferred larger within-pair height differences than
men. Men’s minimally preferred height difference was 0.021
(SD ¼6.65) cm (indicating that on average men prefer to be a
minimum of 0.021 cm taller than a woman), whereas women
indicated a significantly larger minimum height difference of 8.30
(SD ¼6.95) cm (t
4314.21
¼40.96, P<0.0001, d¼1.21). A one-sample
ttest against zero revealed that women (t
2652
¼61.45, P<0.0001,
d¼1.19), but not men (t
1956
¼0.14, P¼0.890, d¼0.003) had on
average a minimal preferred height such that the male was taller
G. Stulp et al. / Animal Behaviour xxx (2013) 1e10 3
Please cite this article in press as: Stulp, G., et al., The height of choosiness: mutual mate choice for stature results in suboptimal pair formation
for both sexes, Animal Behaviour (2013), http://dx.doi.org/10.1016/j.anbehav.2013.03.038

than the female in the couple. With respect to the maximum
preferred height difference, we again found a significant contrast
between the sexes: on average, men preferred a slightly smaller
maximum within-pair height difference than women (men:
mean SD: 24.67 7.44 cm; women: 27.94 6.54 cm;
t
3637.69
¼14.28, P<0.0001, d¼0.47).
Knowing the distribution of both individual preference rankings
and actual heights enabled us to identify the potential direction and
intensity of intersexual selection acting on height (Fawcett &
Johstone 2003). To this end, we first calculated how many
opposite-sex individuals would accept a partner of a given height in
the sense that his/her height was between the reported minimum
and maximum preferred height of opposite-sex participants. We
then calculated the total number of individuals that were of
acceptable height for these opposite-sex individuals. In these cal-
culations, we also included individuals with very low minimal or
high maximal preferred heights. The ratio of these values gives the
number of same-sex people that an individual of a given height
would face as competition per opposite-sex person. For instance, a
man of 177.8 cm (70 inches) would fall within the preferred height
range of 2458 women. These 2458 women on average would accept
2101 other men. Thus, a man of 177.8 cm would compete with, on
average, 2101/2458 ¼0.85 men (see Appendix Tables A3 and A4 for
these calculations for all heights). Short men faced the greatest
number of competitors (Fig. 2), whereas men of average height had
the fewest competitors. Very tall men had more competitors than
men of average height, but fewer than short men. Relatively short
and relatively tall women faced more competition than women of
average height, but variation in competition across women was
much lower than across men (Fig. 2). This reflects our finding that
the male-preferred height range was, on average, larger than the
female-preferred height range.
Strength of Preference in Relation to Height
To establish the strength of the preferences with respect to
height we analysed the relationship between preferred height
range and choice. Strength was assessed on two levels. First, we
examined the likelihood that an individual said ‘Yes ’to a speed-
dater who fell within the reported preferred height range of that
individual. For men, the estimated likelihood of saying ‘Yes ’to a
preferred individual with respect to height was 47.9%, whereas for
a nonpreferred individual this was reduced to 42.8% (logistic
regression: Z¼7.62, P<0.0001). For women, these same values
were 32.2% for a preferred individual versus 25.4% for a non-
preferred individual (Z¼13.64, P<0.0001): a significantly greater
decrease than seen in men (interaction term: Z¼3.10, P¼0.002).
Second, for those speed-daters who fell outside the preferred
height range of a choosing individual, we assessed the extent to
which the magnitude of the deviation from the preferred height
range influenced the chance of saying ‘Yes ’. For men, we found
that the likelihood of saying ‘Yes ’to an individual who fell 1 inch
(2.54 cm) outside the preferred height range was predicted to be
40.0%, whereas this likelihood decreased by 5.7% when the indi-
vidual fell 5 inches (12.7 cm) outside the preferred range (Z¼3.01,
P¼0.003; Fig. 3). For women, we found that the likelihood of
saying ‘Yes’to an individual who fell 1 inch (2.54 cm) outside the
preferred height range was predicted to be 24.8%, while 5 inches
decreased it by 8.0% (Z¼7.87, P<0.0001). A significant interaction
was found between sex and the deviation from the preferred
height range (Z¼2.41, P¼0.016), indicating that preference
strength was stronger in women than men. Thus, both analyses
showed that women had a stronger preference than men, and that
the link between preference and choice was stronger in women
than in men.
Examining the strength of preference separately towards
heights above and below the preferred height range, we found that
women disfavoured heights that were shorter than preferred more
than those taller than preferred (interaction term: Z¼3.02,
P¼0.003; Fig. 3). The reverse was true for men: men tended to
disfavour women who were taller than preferred more than
women shorter than preferred (interaction term: Z¼1.6 6,
P¼0.097). This pattern was significantly different between the
sexes (interaction term: Z¼3.33, P¼0.001).
140
0
2
4
6
8
10
12
14
16
Men
Women
150 160 170
Hei
g
ht (cm)
180 190 200 210
Average number of competitors
Figure 2. The number of competitors in the speed-date population for men and
women in relation to their height. High values indicate that the number of individuals
of a given height is high relative to the number of opposite-sex individuals for whom
that height falls within the acceptable height range. See text and Tables A3 and A4 for
further information.
150 160 170 180 190 200
200
190
180
170
160
150
Hei
g
ht (cm)
Preferred height (cm)
Women
Men
Figure 1. Minimum and maximum preferred height (means SE) in relation to sub-
ject height for men (filled triangles) and women (open triangles). The lines reflect the
midpoint between the minimally and maximally preferred height. For men, bins below
65 inches and above 75 inches, and for women bins below 60 inches and above
70 inches, were collapsed.
G. Stulp et al. / Animal Behaviour xxx (2013) 1e104
Please cite this article in press as: Stulp, G., et al., The height of choosiness: mutual mate choice for stature results in suboptimal pair formation
for both sexes, Animal Behaviour (2013), http://dx.doi.org/10.1016/j.anbehav.2013.03.038

Tolerance in Relation to Height
The standard deviation of heights to which a ‘Yes’response was
given (including only those individuals that responded with ‘Yes ’
more than once) was on average 2.45 (0.73) for men and
2.35 (1.00) for women, a small but significant difference
(t
5203
¼5.24, P<0.0001, d¼0.15). This difference in tolerance was
not a consequence of different standard deviations of height for men
and women (Levene’s test for equality of variances: F¼0.087,
P¼0.768). This reinforces the results given above on strength of
preferences,with women displaying a significanttendency to choose
a narrower range of mates during speed dating than men. This
measure of tolerance also correlated weakly but significantly with
the reported preferred height range in men (Spearman correlation:
r
S
¼0.070, N¼1691, P¼0.004) and women (Spearman correlation:
r
S
¼0.064, N¼1690, P¼0.009), indicating that individuals who
reported a narrower preferred height range also showed less vari-
ability with respect to which heights were given a ‘Yes’response.
Responsiveness, Desirability and Pair Formation in Relation to
Height
Overall, we found that, for both men and women, those who
reported a wider preferred height range were also more responsive
in general (men: Z¼4.73, P<0.0001; women: Z¼8.63,
P<0.0001). Furthermore, men were more responsive than
women: on average, they said ‘Yes ’to 47.4% of women, whereas for
women this value was substantially lower at 30.2% (Z¼20.85,
P<0.0001).
Taller men were less responsive, but more likely to receive a ‘Yes’
response from women (which we refer to as ‘desirability’). Both
relationships were curvilinear, with maximum responsiveness at
7.2 cm above average height (95% confidence interval, CI ¼3.4e
18.9), and maximum desirability at 21.3 cm (95% CI ¼12.9e64.0)
above average height (Table 2,Fig. 4a, b). The desirability effect was
stronger than the responsiveness effect and hence taller men were
more likely to form a pair, that is, were more likely to end up with a
match (Table 2,Fig. 4c).
Female height was not significantly related to either respon-
siveness or pair formation (Table 2,Fig. 4a, c). However, very short
and very tall women were slightly less desirable, as indicated by a
significant quadratic effect of female height on the chance of
receiving a ‘Yes’response (Table 2,Fig. 4b). The most frequently
chosen female height was predicted to lie 0.38 cm (95% CI ¼5.1e
6.4) above average height although, from the graph, it is clear that
particularly very short and very tall women were least frequently
chosen, whereas women in the intermediate height range had
similar likelihoods of receiving a ‘Yes ’response. Thus, very short
and very tall women were chosen less often than those of more
intermediate heights.
Mutual Mate Choice and Pair Formation
As partner preferences are dependent on one’s own height
(Fig. 1), relative height difference may be more informative with
respect to the chance of giving a ‘Yes ’response than assessment of
potential partner height alone because it integrates the height of
both participants. We found curvilinear effects for both men and
women on the chance of giving a ‘Yes ’response with respect to
partner height differences (male minus female height; Table 3,
Fig. 5). For men, a ‘Yes ’response was most likely when the woman
was 7.1 cm (95% CI ¼1. 0e12.2) shorter than themselves, signifi-
cantly lower than the average height difference of 13.9 cm between
men and women in our sample. Women, in contrast, were most
Table 2
The effect of male and female height (in inches; mean-centered) on the likelihood of giving a ‘Yes ’response, receiving a ‘Yes ’response, and having a match during speed dating
Likelihood of giving a ‘Yes’response Likelihood of receiving a ‘Yes’response Likelihood of match
Male Female Male Female Male Female
Intercept 0.220.052
(<0.0001)
1.230.043
(<0.0001)
1.180.046
(<0.0001)
0.120.053 (0.026) 2.270.034
(<0.0001)
2.270.034
(<0.0001)
Height 0.0500.012
(<0.0001)
0.000490.010
(0.962)
0.120.010
(<0.0001)
0.00210.010 (0.834) 0.0470.0084
(<0.0001)
0.00410.0083
(0.619)
Height
2
0.00880.0027
(0.001)
*
0.00710.0024
(0.003)
0.00700.0030 (0.020)
**
Random effects
y
Choosing
individual ID
2.231.49 1.661.29 1.761.33 2.311.52 0.790.89 0.810.90
Chosen
individual ID
1.631.28 1.581.26 1.481.21 1.611.27 0.850.92 0.830.91
Event ID 0.130.36 0.0860.29 0.0810.28 0.130.36 0.0560.24 0.0540.23
Table entries show binomial logistic mixed-model parameter estimates SE and the associated Pvalue (in parentheses).
*
The squared term of height was nonsignificant (P>0.159). We present the estimates from the model without this term.
y
Parameter estimate for variance components SD.
–15 –10 –5 0 5 10 15
Deviation from
p
referred hei
g
ht ran
g
e (cm)
0
10
20
30
40
50
Likelihood of giving ‘Yes’ response (%)
Shorter than preferred Taller than preferred
By men
By women
Figure 3. The strength of the height preference: the likelihood of giving a ‘Ye s’
response with increased deviation from the preferred height range for men and
women (mean SE). Bins below 7 inches and above 7 inches were collapsed. The
likelihood of giving a ‘Yes ’response when the height fell within the height range is
plotted for comparison.
G. Stulp et al. / Animal Behaviour xxx (2013) 1e10 5
Please cite this article in press as: Stulp, G., et al., The height of choosiness: mutual mate choice for stature results in suboptimal pair formation
for both sexes, Animal Behaviour (2013), http://dx.doi.org/10.1016/j.anbehav.2013.03.038

likely to give a ‘Yes ’response when the man was 25.1 cm taller (95%
CI ¼22.1e28.8). This height difference was significantly larger than
the average height difference between men and women and also
substantially larger than the male optimum.
We then calculated the height difference with the highest
likelihood of a match (i.e. mutual score of ‘Yes’). We found a
maximum likelihood of a match at 19.2 cm (95% CI ¼16.2e22.8)
which falls in between the most chosen value for both men (7.1 cm)
and women (25.1 cm), and which was also significantly greater
than the average height difference (Table 3,Fig. 5). When we
multiply the curve of the men giving a ‘Yes’response (with respect
to the height differences) with that of the female curve, we obtain a
curve indicating the chance of a match when the chance of a ‘Yes’
being reciprocated is independent (i.e. the likelihood of having a
‘Yes ’response reciprocated is equal to that nonreciprocated). This
estimated curve was very similar to the observed height distribu-
tion of the matches (Fig. 5), suggesting that men and women were
not more likely to give a ‘Ye s’response to an individual who gave
them a ‘Yes ’response in turn; thus there is no evidence to suggest
that a given couple feels a ‘click’with one another.
Resulting Mating Patterns with Respect to Height
We correlated the height of an individual with the average
height of all matches of that individual (Fig. 6). Both male
(r
2204
¼0.128, P<0.0001) and female (r
2379
¼0.105, P<0.0001)
height correlated positively with this average height, providing
some indication of assortative mating: taller individuals tended to
be matched with taller individuals in both sexes, but men of
average height were more likely to be matched with shorter fe-
males (Fig. 6). Thus, assortative mating for height was tempered by
female choice for men much taller themselves, with the result that
men of average height, rather than shorter men, were more likely to
be matched with shorter women.
DISCUSSION
Studies of mate choice are generally restricted to the assessment
of preferences, thereby neglecting the subsequent processes that
lead to pair formation. Here, departing from previous work, we
addressed simultaneously how preferences for partner height
translated into actual choice, and how choice then translated into
pairing. We found that nonpreferred potential partners with
respect to height still had a high (albeit reduced) chance of being
chosen (42.8% and 25.4% for men and women, respectively). This is
consistent with previous studies, which show that reported partner
preferences are not strong predictors of choice during speed dating
(Kurzban & Weeden 2007;Todd et al. 2007;Eastwick & Finkel
2008;Eastwick et al. 2011).
In addition to the reasons put forward in the Introduction to
account for why preferences are not always expected to predict
choice, humans may not be able express their own preferences
accurately, or they may feel compelled to give socially desirable
answers (Todd et al. 2007;Eastwick et al. 2011). Additionally, and
perhaps most crucially, the setting in which preferences are
established may not conform to the situation in which preferences
are actually expressed. It is worth bearing in mind, therefore, that
other psychological processes besides those relating strictly to
mating decisions may explain some of the deviation of choice from
preference.
155 160 165 170 175 180 185 190
155 160 165 170 175 180 185 190
155 160 165 170 175 180 185 190
By men
For men
For women
For men
For women
Hei
g
ht (cm)
By women
10
20
30
40
50
60 (a)
(b)
(c)
10
20
30
40
50
60
Likelihood of giving ‘Yes’ response (%)Likelihood of receiving ‘Yes’ response (%)Likelihood of match (%)
20
18
16
14
12
10
8
Figure 4. The effect of male and female height on (a) the likelihood of giving a ‘Ye s’
response, (b) the likelihood of receiving a ‘Yes ’response, and (c) the likelihood of a
match (all mean SE). For men, bins below 65 inches and above 75 inches, and for
women bins below 60 inches and above 70 inches, were collapsed.
G. Stulp et al. / Animal Behaviour xxx (2013) 1e106
Please cite this article in press as: Stulp, G., et al., The height of choosiness: mutual mate choice for stature results in suboptimal pair formation
for both sexes, Animal Behaviour (2013), http://dx.doi.org/10.1016/j.anbehav.2013.03.038

Despite the imperfect mapping of preferences onto choice, we
found that women’s preferences were more strongly related to
subsequent choice than those of men. Women reported a narrower
preferred height range than men, and they were also less likely to
choose men that fell outside this range (i.e. women had a higher
strength of preference). Similarly, there was less variation in the
heights chosen by women compared to men (i.e. women also had a
lower tolerance). Finally, women were less responsive overall than
men, mirroring findings from previous research (e.g. Kurzban &
Weeden 2005;Todd et al. 2007;Eastwick & Finkel 2008;Lenton
& Francesconi 2011), but there was no influence of a woman’s
own height on her responsiveness (in line with Kurzban & Weeden
2005). Female height did, however, influence their desirability:
women of average height were most desired during speed dating.
Furthermore, based on the preferences for height expressed by men
and the actual height distribution of women, it was clear that
women of average height also had the fewest rivals to compete
with compared to shorter and taller women. These effects were
generally small, however, and did not translate into actual success,
as female height was unrelated to the chance of a match.
A contrasting pattern of results was obtained for men, where an
individual’s own height had a significant influence on his respon-
siveness: specifically, taller men were less responsive than shorter
men. The lower responsiveness displayed by taller men can be
partly explained by their increased desirability, as taller men were
most often given a ‘Yes’response by women and had to compete
with fewer rivals than shorter men. Thus, the increased popularity
of, and reduced competition for, taller men compared to shorter
men may explain their decreased responsiveness during speed
dating. Despite being less responsive, taller man were most likely to
end up with a match. Taken together, these results demonstrate
that height is considered more important by women as a mate
choice characteristic, and that men’s‘mating success’is therefore
more dependent on height than the mating success of women. Thus
female mate choice is a likely contributor to the evolution of human
sexual size dimorphism.
Stated preferences for height differences also revealed a conflict
between the sexes. In general, women preferred their partner to be
much taller, whereas men preferred their partner to be only slightly
shorter. These stated preferences were also reflected in choice: men
were most likely to choose only small partner height differences,
including those height differences in which the woman was taller
than the choosing men. Women, in contrast, were most likely to
choose much larger partner height differences, and least likely to
choose small partner height differences, particularly those that
would result in the man being shorter. Further evidence that
women disfavour men shorter than themselves is also shown by
the differences in their strength of preference: women strongly
disfavoured men who were shorter, but not those who were taller,
than their preferred height range (Fig. 3). These converging lines of
Table 3
The effect of the difference in height (male minus female height; in inches) on the likelihood of giving a ‘Yes’response by men and women, and the likelihood of a match
Likelihood of giving a ‘Yes’response Likelihood of match
Men Women
Intercept 0.110.065 (0.081) 1.670.061 (<0.0001) 2.470.051 (<0.0001)
Height difference 0.0210.011 (0.052) 0.140.011 (<0.0001) 0.0850.011 (<0.0001)
Height difference
2
0.00380.0007 (<0.0001) 0.00720.0008 (<0.0001) 0.00570.0008 (<0.0001)
Random effects
*
Choosing individual ID 2.291.51 1.711.31 0.830.91
Chosen individual ID 1.631.27 1.501.22 0.810.90
Event ID 0.130.35 0.0890.30 0.0600.25
Table entries show binomial logistic mixed model parameter estimates (SE) and the associated Pvalue (in parentheses).
*
Parameter estimate for variance components (SD).
–20 –10 0 10 20 30 40 50
Hei
g
ht difference (cm)
0
10
20
30
40
50
60
Men giving ‘Yes’ response
Women giving ‘Yes’ response
Both giving ‘Yes’ response
Likelihood (%)
Figure 5. Height differences (male minus female height), the likelihood of giving a
‘Yes ’response by men and women and the likelihood of a match (all mean SE). Bins
below 6 inches and above 17 inches were collapsed. The broken line represents the
multiplication of the curves representing the likelihood of giving a ‘Yes ’response for
both men and women.
150
178
179
180
181
162
163
164
165
166
167 Men
Women
160
Hei
g
ht (cm)
Average height of matches (cm)
170 180 190
(a)
(b)
Figure 6. The average height SE of the individuals with whom an individual of a
given height is paired for (a) men and (b) women. For every individual we calculated
the average height of all individuals with which they had a match (see text). The
horizontal line represents the average height of the opposite sex.
G. Stulp et al. / Animal Behaviour xxx (2013) 1e10 7
Please cite this article in press as: Stulp, G., et al., The height of choosiness: mutual mate choice for stature results in suboptimal pair formation
for both sexes, Animal Behaviour (2013), http://dx.doi.org/10.1016/j.anbehav.2013.03.038

evidence strongly suggest that the male-taller norm observed in
actual couples (i.e. males are more often taller than their partner
when compared to random mating; Gillis & Avis 1980) is driven by
women rather than by men.
The process of mate choice and pair formation during speed
dating resulted in assortative mating for height, but the magnitude
of assortment in the resulting matches is lower than that observed
among actual couples (Spuhler 1982;Stulp et al. 2011,2013a). A
possible explanation for a lower level of assortment is that, because
women’s choices were so strongly directed towards men much taller
than themselves, it was men of average height, rather than shorter
men, who were most likely to be paired with shorter women (Fig. 6).
Although shorter men were much less likely to find a match during
speed-dating events (Fig. 4c), they may nevertheless succeed in
finding a partner outside of this more restricted context because the
availability of average height and taller men in a population is,
obviously, finite. Once more preferred men are removed from the
mating pool, some women may be forced to compromise with
respect to partner height, and pair up with shorter men. Thus,
because shorter men are potentially still successful at finding a
partner, and because such men are more likely to be paired with
shorter women than with tall women (because of preferences in
both sexes), the observed magnitude of assortment in actual couples
will be higher than that seen in a speed-dating context.
Our most notable finding, however, concerns the manner in
which the conflict over partner height difference extended to actual
pair formation. While men preferentially chose partners with a
height difference that fell significantly below the average height
difference between men and women, women chose partners with
height differences that were significantly above this average dif-
ference (Fig. 5). This conflict in choice inevitably resulted in pairs in
which the height difference between partners was suboptimal for
both sexes, even though all parties were expressing a free choice
and rivals did not prevent this choice. Thus, our study shows how
mutual mate choice for preferred partners can lead to suboptimal
pair formation, highlighting the value of following the mate choice
process beyond the establishment of preferences through to pair
formation. Our study also illustrates the value of speed dating as a
model system, as it shows how human mate choice processes can be
studied in a manner directly comparable to those of other species.
Acknowledgments
We thank Louise Barrett for her many helpful comments on the
manuscript. We thank HurryDate for generously providing the data,
Jason Weeden for help with the data, and Alexandre Courtiol for
statistical advice. We also thank two anonymous referees for their
comments which improved the manuscript.
References
Asendorpf, J., Penke, L. & Back, M. 2011. From dating to mating and relating:
predictors of initial and long-term outcomes of speed-dating in a community
sample. European Journal of Personality,30,16e30.
Back, M. D., Penke, L., Schmukle, S. C., Sachse, K., Borkenau, P. & Asendorpf, J. B.
2011. Why mate choices are not as reciprocal as we assume: the role of per-
sonality, flirting and physical attractiveness. European Journal of Personality,25,
120e132.
Baldauf, S. A., Kullmann, H., Schroth, S. H., Thunken, T. & Bakker, T. C. 2009. You
can’t always get what you want: size assortative mating by mutual mate choice
as a resolution of sexual conflict. BMC Evolutionary Biology,9, 129.
Belot, M. & Fidrmuc, J. 2010. Anthropometry of love: height and gender asym-
metries in interethnic marriages. Economics & Human Biology,8,361e372.
Cotton, S., Small, J. & Pomiankowski, A. 2006. Sexual selection and condition-
dependent mate preferences. Current Biology,16, R755eR765.
Courtiol, A., Raymond, M., Godelle, B. & Ferdy, J.-B. 2010a. Mate choice and hu-
man stature: homogamy as a unified framework for understanding mating
preferences. Evolution,64,2189e2203.
Courtiol, A., Picq, S., Godelle, B., Raymond, M. & Ferdy, J.-B. 2010b. Frompreferred
to actual mate characteristics: the case of human body shape. PLoS One,5,e13010.
Diggle, P., Heagarty, P., Liang, K.-Y. & Zeger, S. 2002. Analysis of Longitudinal Data.
2nd edn. Oxford: Oxford University Press.
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, 245e264.
Eastwick, P. W., Eagly, A. H., Finkel, E. J. & Johnson, S. E. 2011. Implicit and explicit
preferences for physical attractiveness in a romantic partner: a double disso-
ciation in predictive validity. Journal of Personality and Social Psychology,101,
993e1011.
Fawcett, T. W. & Johstone, R. A. 20 03. Mate choice in the face of costly competition.
Behavioral Ecology,14,771e779.
Fink, B., Neave, N., Brewer, G. & Pawlowski, B. 2007. Variable preferences for
sexual dimorphism in stature (SDS): further evidence for an adjustment in
relation to own height. Personality and Individual Differences,43, 2249e2257.
Finkel, E. J. & Eastwick, P. W. 2008. Speed-dating. Current Directions in Psycho-
logical Science,17,193e197.
Fowler-Finn, K. D. & Rodríguez, R. L. 2012a. Experience-mediated plasticity in mate
preferences:mating assurance in a variable environment. Evolution,66,459e468.
Fowler-Finn, K. D. & Rodríguez, R. L. 2012b. The evolution of experience-mediated
plasticity in mate preferences. Journal of Evolutionary Biology,25, 1855e1863.
Gillis, J. S. & Avis, W. E. 1980. The male-taller norm in mate selection. Personality
and Social Psychology Bulletin,6, 396e401.
Johnstone, R. A., Reynolds, J. D. & Deutsch, J. C. 1996. Mutual mate choice and sex
differences in choosiness. Evolution,50, 1382e1391.
Kurzban, R. & Weeden, J. 2005. HurryDate: mate preferences in action. Evolution
and Human Behavior,26,227e244.
Kurzban, R. & Weeden, J. 2007. Do advertised preferences predict the behavior of
speed daters? Personal Relationships,14, 623e632.
Lenton, A. P. & Francesconi, M. 2011. Too much of a good thing? Variety is
confusing in mate choice. Biology Letters,7, 528e531.
Lenton, A. P., Fasolo, B. & Todd, P. M. 2009. The relationship between number of
potential mates and mating skew in humans. Animal Behaviour,77,55e60.
Luo, S. & Zhang, G. 2009. What leads to romantic attraction: similarity, reciprocity,
security, or beauty? Evidence from a speed-dating study. Journal of Personality,
77, 933e964.
Magrath, M. J. L. & Komdeur, J. 2003. Is male care compromised by additional
mating opportunity? Trends in Ecology & Evolution,18, 424e430.
Pawlowski, B. 2003. Variable preferences for sexual dimorphism in height as a
strategy for increasing the pool of potential partners in humans. Proceedings of
the Royal Society B,270, 709e712.
R Development Core Team. 2008. R: A Language and Environment for Statistical
Computing. Vienna: R Foundation for Statistical Computing. http://www.r-
project.org.
Re, D. E. & Perrett, D. I. 2012. Concordant preferences for actual height and facial
cues to height. Personality and Individual Differences,53,901e906.
Reynolds, J. D. & Gross, M. R. 1990. Costs and benefits of female mate choice: is
there a lek paradox? The American Naturalist,136, 230e243.
Rusbult, C. E. & Buunk, B. P. 1993. Commitment processes in close relationships:
an interdependence analysis. Journal of Social and Personal Relationships,10,
175 e204.
Sear, R. & Marlowe, F. W. 2009. How universal are human mate choices? Size does
not matter when Hadza foragers are choosing a mate. Biology Letters,5, 606e609.
Sorokowski, P. & Butovskaya, M. L. 2012. Height preferences in humans may not be
universal: evidence from the Datoga people of Tanzania. Body Image,9, 510e516 .
Spuhler, J. N. 1982. Assortative mating with respect to physical characteristics.
Social Biology,29,53e66.
Stulp, G., Verhulst, S., Pollet, T. V., Nettle, D. & Buunk, A. P. 2011. Parental height
differences predict the need for an emergency Caesarean section. PLoS One,6,
e20497.
Stulp, G., Pollet, T. V., Verhulst, S. & Buunk, A. P. 2012a. A curvilinear effect of
height on reproductive success in human males. Behavioral Ecology and Socio-
biology,66,375e384.
Stulp, G., Verhulst, S., Pollet, T. V. & Buunk, A. P. 2012b. The effect of female height
on reproductive success is negative in Western populations, but more variable
in non-Western populations. American Journal of Human Biology,24, 486e494.
Stulp, G., Kuijper, B., Buunk, A. P., Pollet, T. V. & Verhulst, S. 2012c. Intralocus
sexual conflict over human height. Biology Letters,8,976e978.
Stulp, G., Buunk, A. P., Pollet, T. V., Nettle, D. & Verhulst, S. 2013a. Are human
mating preferences with respect to height reflected in actual pairings? PLoS
One,8, e54186.
Stulp, G., Buunk, A. P. & Pollet, T. V. 2013b. Women want taller men more than
men want shorter women. Personality and Individual Dif ferences,54, 877e883.
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, U.S.A.,104 ,15011e15016 .
Torres, L. A. G. M. M., Martinez, F. E. & Manço, J. C. 2003. Correlation between
standing height, sitting height, and arm span as an index of pulmonary function
in 6e10-year-old children. Pediatric Pulmonology,36, 202e208.
Widemo, F. & Sæther, S. A. 1999. Beauty is in the eye of the beholder: causes and
consequences of variation in mating preferences. Trends in Ecology & Evolution,
14,26e31.
Wong, B. B. M. & Candolin, U. 2005. How is female mate choice affected by male
competition? Biological Reviews,80, 559e571.
G. Stulp et al. / Animal Behaviour xxx (2013) 1e108
Please cite this article in press as: Stulp, G., et al., The height of choosiness: mutual mate choice for stature results in suboptimal pair formation
for both sexes, Animal Behaviour (2013), http://dx.doi.org/10.1016/j.anbehav.2013.03.038

Appendix
Table A1
Descriptive statistics of the events (N¼174 )
MeanSD Minimum Maximum
No. of persons in event 36.7910.82 15 65
No. of men in event 18.806.06 6 35
No. of women in event 18.175.18 7 30
Sex ratio
*
0.510.050 0.35 0.63
The correlation between the number of men and women participating in the event
was r¼0.852, P<0.0001.
*
No. of men in event divided by no. of persons in event.
Table A2
Descriptive statistics of the speed-daters
NMeanSD Minimum Maximum
Female 3024
Age 3019 31.835.16 20 53
Height
*
2852 165.206.72 144.78 187.96
Ethnicity 3024
Caucasian 2305
Asian 135
African 107
Hispanic 125
Other 83
Unknown 269
Male 2758
Age 2755 34.416.06 21 68
Height 2604 179.066.87 142.24 213.36
Ethnicity 2758
Caucasian 2109
Asian 138
African 86
Hispanic 98
Other 102
Unknown 225
*
Height was originally reported in inches.
Table A3
Calculating the average number of competitors per male height (see Fig. 2)
Male
height
*
Frequency No. of women would
accept height
y
Average no. of men
accepted by these
women
z
Considered too
short by no. of
women
Considered too tall
by no. of women
Average no. of
competitors
x
142.24 1 (0.04) 171 (6.01) 2547 (97.81) 2676 (93.99) 0 (0) 14.89
154.94 1 (0.04) 194 (6.81) 2547 (97.81) 2653 (93.19) 0 (0) 13.12
157.48 4 (0.15) 207 (7.27) 2542 (97.61) 2640 (92.73) 0 (0) 12.27
160.02 6 (0.23) 235 (8.25) 2533 (97.29) 2612 (91.75) 0 (0) 10.78
162.56 11 (0.42) 298 (10.47) 2508 (96.30) 2549 (89.53) 0 (0) 8.41
165.1 42 (1.61) 455 (15.98) 2490 (95.62) 2392 (84.02) 0 (0) 5.47
167.64 107 (4.11) 685 (24.06) 2463 (94.57) 2162 (75.94) 0 (0) 3.59
170.18 175 (6.72) 1014 (35.62) 2412 (92.64) 1833 (64.38) 0 (0) 2.38
172.72 258 (9.91) 1486 (52.2) 2334 (89.61) 1360 (47.77) 1 (0.04) 1.57
175.26 312 (11.98) 1891 (66.42) 2244 (86.18) 954 (33.51) 2 (0.07) 1.19
177.8 397 (15.25) 2458 (86.34) 2102 (80.72) 385 (13.52) 4 (0.14) 0.85
180.34 351 (13.48) 2689 (94.45) 2029 (77.92) 150 (5.27) 8 (0.28) 0.75
182.88 389 (14.94) 2803 (98.45) 1980 (76.02) 18 (0.63) 26 (0.91) 0.71
185.42 209 (8.03) 2676 (93.99) 1982 (76.11) 4 (0.14) 167 (5.87) 0.74
187.96 174 (6.68) 2590 (90.97) 1983 (76.16) 0 (0) 257 (9.03) 0.77
190.5 80 (3.07) 2243 (78.78) 1994 (76.57) 0 (0) 604 (21.22) 0.89
193.04 53 (2.04) 1929 (67.76) 2001 (76.85) 0 (0) 918 (32.24) 1.04
195.58 17 (0.65) 1536 (53.95) 2025 (77.76) 0 (0) 1311 (46.05) 1.32
198.12 10 (0.38) 1185 (41.62) 2063 (79.21) 0 (0) 1662 (58.38) 1.74
200.66 4 (0.15) 1016 (35.69) 2085 (80.06) 0 (0) 1831 (64.31) 2.05
203.2 1 (0.04) 959 (33.68) 2096 (80.50) 0 (0) 1888 (66.32) 2.18
213.36 2 (0.08) 844 (29.65) 2110 (81.03) 0 (0) 2003 (70.35) 2.50
Numbers in parentheses are percentages.
*
Height in cm (originally reported in inches).
y
The number of women who included that particular male height in their preferred height range.
z
The average number of men liked by all the women who included that particular male height in their preferred height range.
x
The average number of competitors was a function of how many women would accept a man of a given height and the average number of men that were accepted by these
women: that is, the average number of other men accepted by the women (average number preferred minus 1) divided by the number of women who would accept them.
G. Stulp et al. / Animal Behaviour xxx (2013) 1e10 9
Please cite this article in press as: Stulp, G., et al., The height of choosiness: mutual mate choice for stature results in suboptimal pair formation
for both sexes, Animal Behaviour (2013), http://dx.doi.org/10.1016/j.anbehav.2013.03.038

Table A4
Calculating the average number of competitors per female height (see Fig. 2)
Female
height
*
Frequency No. of men would
accept
y
Average no. of women
accepted by these men
z
Considered too short
by no. of men
Considered too tall
by no. of men
Average no. of
competitors
x
144.78 1 (0.04) 863 (33.18) 2760 (96.77) 1738 (66.82) 0 (0) 3.20
147.32 9 (0.32) 1000 (38.45) 2736 (95.92) 1601 (61.55) 0 (0) 2.73
149.86 17 (0.6) 1070 (41.14) 2725 (95.53) 1531 (58.86) 0 (0) 2.55
152.4 85 (2.98) 1740 (66.9) 2706 (94.89) 861 (33.1) 0 (0) 1.55
154.94 124 (4.35) 1887 (72.55) 2695 (94.49) 714 (27.45) 0 (0) 1.43
157.48 277 (9.71) 2181 (83.85) 2666 (93.47) 420 (16.15) 0 (0) 1.22
160.02 304 (10.66) 2303 (88.54) 2644 (92.72) 297 (11.42) 1 (0.04) 1.15
162.56 455 (15.95) 2446 (94.04) 2606 (91.37) 151 (5.81) 4 (0.15) 1.06
165.1 363 (12.73) 2532 (97.35) 2570 (90.1) 58 (2.23) 11 (0.42) 1.01
167.64 380 (13.32) 2532 (97.35) 2564 (89.9) 21 (0.81) 48 (1.85) 1.01
170.18 305 (10.69) 2468 (94.89) 2574 (90.26) 8 (0.31) 125 (4.81) 1.04
172.72 219 (7.68) 2329 (89.54) 2592 (90.88) 3 (0.12) 269 (10.34) 1.11
175.26 169 (5.93) 2099 (80.7) 2614 (91.64) 1 (0.04) 501 (19.26) 1.24
177.8 100 (3.51) 1862 (71.59) 2628 (92.13) 0 (0) 739 (28.41) 1.41
180.34 32 (1.12) 1537 (59.09) 2641 (92.61) 0 (0) 1064 (40.91) 1.72
182.88 10 (0.35) 1356 (52.13) 2660 (93.27) 0 (0) 1245 (47.87) 1.96
185.42 1 (0.04) 949 (36.49) 2682 (94.03) 0 (0) 1652 (63.51) 2.82
187.96 1 (0.04) 859 (33.03) 2704 (94.82) 0 (0) 1742 (66.97) 3.15
Numbers in parentheses are percentages.
*
Height in cm (originally reported in inches).
y
The number of men who included that particular female height in their preferred height range.
z
The average number of women liked by all the men who included that particular female height in their preferred height range.
x
The average number of competitors was a function of how many men would accept a woman of a given height and the average number of women that were accepted by
these men, that is, the average number of other women accepted by the men (average number preferred minus 1) divided by the number of men who would accept them.
G. Stulp et al. / Animal Behaviour xxx (2013) 1e1010
Please cite this article in press as: Stulp, G., et al., The height of choosiness: mutual mate choice for stature results in suboptimal pair formation
for both sexes, Animal Behaviour (2013), http://dx.doi.org/10.1016/j.anbehav.2013.03.038