ARTICLE IN PRESS

Evolution and Human Behavior xx (2009) xxx – xxx

Original Article

Partner wealth predicts self-reported orgasm frequency in a sample of

Chinese women
Thomas V. Pollet⁎, Daniel Nettle
Centre for Behaviour and Evolution, Newcastle University, Henry Wellcome Building for Neuroecology, Framlington Place,
Newcastle upon Tyne NE24HH, UK
Initial receipt 17 April 2008; final revision received 13 November 2008

Abstract

There has been considerable speculation about the adaptive significance of the human female orgasm, with one hypothesis being that it
promotes differential affiliation or conception with high-quality males. We investigated the relationship between women's self-reported
orgasm frequency and the characteristics of their partners in a large representative sample from the Chinese Health and Family Life Survey.
We found that women report more frequent orgasms the higher their partner's income is. This result cannot be explained by possible
confounds such as women's age, health, happiness, educational attainment, relationship duration, wealth difference between the partners,
difference between the partners in educational attainment, and regional location. It appears consistent with the view that female orgasm has an
evolved adaptive function.
© 2009 Published by Elsevier Inc.

Keywords: Orgasm; Mate choice; Humans; Male quality; Evolutionary psychology

1. Introduction The frequency of orgasm has been found to be an

important component of sexual satisfaction, which in turn is
Female orgasm in humans is a topic of continued debate a predictor of relationship satisfaction, for Chinese women
and speculation but has relatively little solid empirical (Parish et al., 2007). In American women, age and religiosity
evidence (Dunn, Cherkas, & Spector, 2005; Meston, Levin, are negative predictors of orgasm frequency (Laumann,
Sipski, Hull, & Heiman, 2004). Although there have been a Gagnon, Michael, & Michaels, 1994; Meston et al., 2004),
multitude of definitions and physiologists, sexologists, and the frequency of masturbatory orgasms but not orgasms
doctors, and sociologists continue to debate the exact nature with a partner increases with increasing education (Laumann
of the experience (Levin, 2004), there is general agreement et al., 1994). Twin evidence also suggests a moderate
that female orgasm is “a variable, transient peak sensation of heritable component (Dunn et al., 2005). Beyond these,
intense pleasure, creating an altered state of consciousness, however, there is a relative paucity of evidence on the
usually with an initiation accompanied by involuntary, psychological and social factors influencing orgasmic
rhythmic contractions of the pelvic striated circumvaginal function (Meston et al., 2004) since research has mainly
musculature, often with concomitant uterine and anal focused on the physiological mechanisms involved (Levin &
contractions and myotonia that resolves the sexually induced Wagner, 1985; Masters & Johnson, 1966).
vasocongestion (sometimes only partially), generally with an Evolutionists have taken opposing positions on the
induction of well-being and contentment.” (Meston et al., function of female orgasm. On one hand, it has been seen
2004, p. 174). as a functionless by-product of the ejaculatory response in
males (Gould, 1987; Symons, 1979). An alternative view is
that women's capacity for orgasm is an adaptation that serves
⁎ Corresponding author. Department of Social and Organizational to discriminate between males on the basis of their quality
Psychology, University of Groningen, Groningen, The Netherlands. (Alcock, 1987; Smith, 1984; Thornhill, Gangestad, &
E-mail address: [email protected] (T.V. Pollet). Comer, 1995), leading to either enhanced conception
1090-5138/$ – see front matter © 2009 Published by Elsevier Inc.
doi:10.1016/j.evolhumbehav.2008.11.002
 ARTICLE IN PRESS
2 T.V. Pollet, D. Nettle / Evolution and Human Behavior xx (2009) xxx–xxx

probability (Baker & Bellis, 1993) or selective emotional status in China as it is elsewhere (Chen, Campbell, Li, &
bonding with high-quality sires (Alexander, 1979; Fisher, Peto, 1990; Chen & Zhou, 2007; Samaras, Elrick, & Storms,
1992; Hrdy, 1996). While there has been considerable debate 2003). Therefore, it is possible that height is a sought-after
surrounding these hypotheses, there are only a few studies characteristic in a man for Chinese women as it is for women
that set out to test the predictions of adaptive versus by- in Western populations.
product hypotheses. If the adaptive view of female orgasm is correct, then we
If female orgasm is adaptively designed for discriminat- predict that women will report more frequent orgasms the
ing male quality, then it should be more frequent in females richer their partners are and the taller their partners are.
paired with high-quality males. There is evidence consistent However, there is the danger that associations between
with such an effect in Japanese macaques, where females partner characteristics and orgasm frequency might be due to
display the orgasm-like clutching reaction more often when confounding third variables. For example, the partners of
mating with high-ranking males (Troisi & Carosi, 1998). richer men might be younger, be more educated, be
Two studies on humans have also found comparable effects. physically healthier, or have higher levels of psychological
Thornhill et al. (1995) found that young women reported well-being, or be more westernized, than partners of poorer
copulatory orgasm more frequently if their partners were men. In our analysis, therefore, we needed to control for a
more symmetrical. Shackelford et al. (2000) found that large number of potential confounding variables of this type.
women paired with more attractive males were more likely to
report orgasm at their last intercourse than women paired
with less attractive males. These studies are interesting but 2. Methods
rely on modest samples (n=86 and 388) of student
volunteers, which means that the ranges of ages and male The Chinese Health and Family Life Survey sampled
qualities may be quite limited. They also found effects only 60 villages and urban neighborhoods chosen in such a way
of symmetry and physical attractiveness as proxies of male as to represent the full geographical and socioeconomic
quality. In a larger population cross-section, other qualities range of contemporary China excluding Hong Kong and
such as income and height may prove to be important. Tibet. Eighty-three individuals were chosen at random for
In this study, we investigated for the first time in a large each location from official registers of adults aged between
population cross-section the relationship between women's 20 and 64 years to target a sample of 5000 individuals in
self-reported orgasm frequency and the characteristics of total. The response rate was around 75%. We only
their partners. The data are from the Chinese Health and included in our analysis women with current male partners
Family Life Survey, which was a large-scale investigation of for whom no information was missing, leading to a sample
sexual attitudes, behavior, and sexually transmitted diseases of 1534 women.
in a nationally representative sample of the Chinese The survey was carried out by structured computer-
population, carried out in 1999–2000 (Parish et al., 2003; administered interview. In view of the sensitive nature of the
Parish, Laumann, & Mojola, 2007). For male quality, we questions, additional measures to ensure privacy were
investigated two parameters not only because they were adopted (Parish et al., 2003). The interview took place
available but also because previous population-level surveys away from each respondent's home, and for sensitive
have found them to be important determinants of male questions, the respondent could enter her responses directly
reproductive success. The first is income: Wealth is seen as into the computer.
desirable in potential partners by women all over the world The dependent variable we analyzed was self-reported
(Buss, 1989), and studies in diverse populations have found orgasm frequency with current partner (“When having sex
that increasing wealth increases male marriage success with your current partner, how often did you have an
(Borgerhoff Mulder, 1990; Fieder & Huber, 2007; Hopcroft, orgasm?”). There are five response categories, which we
2006; Nettle & Pollet, in press; Pollet & Nettle, 2008). Thus, treated as an ordinal scale (1=never, n=61; 2=rarely, n=182;
other things being equal, richer men are preferred to poorer 3=sometimes, n=762; 4=often, n=408; 5=always, n=121).
ones as mates. Our two independent variables of interest, partner income
Our second independent variable is height: Several and partner height, are based on the woman's report, not the
studies have found either a preference expressed by partner's. However, for a subsample of 57 couples, the
women for taller men or an actual increase in marriage interrater reliability of respondent and partner reports has
success in taller-than-average men (Hensley, 1994; Mueller been calculated and is substantial (income, κN0.6; height,
& Mazur, 2001; Nettle, 2002; Pawlowski, Dunbar, & κN0.48) (Parish et al., 2003). Partner income was as directly
Lipowicz, 2000) in Western populations. Little is known reported for 90% of the respondents and was imputed from a
about male height and reproductive success in China. A categorical response for the remainder, as the respondents
recent study suggested that male height is not significantly did not give an exact figure. Figures are in yuan; 1 yuan is
positively related to offspring count, after controlling for worth around 15 U.S. cents at international currency
potential confounds (Fielding et al., in press), yet height is conversion rates but closer to 50 U.S. cents in terms of
generally positively related to health and socioeconomic local purchasing power (World Bank, 2006). Incomes higher
 ARTICLE IN PRESS
T.V. Pollet, D. Nettle / Evolution and Human Behavior xx (2009) xxx–xxx 3

than 10,000 yuan were recoded as 10,000. Partner height was (Armstrong & Sloan, 1989; McCullagh, 1980). This is the
an estimate given in centimeters. Partner height and partner appropriate technique for a dependent variable consisting of
income were virtually uncorrelated (r=0.06). ranked outcome categories. Our analytical strategy was as
Our control variables were the following: age (years, follows: Model 1 included only partner income and partner
continuous); duration of current relationship (years, contin- height as independent variables, while model 2 omitted any
uous); self-rated health (five categories); self-rated happiness independent variable not significant in model 1. In
(four categories); education (six categories); difference in subsequent models, we added the control variables one at a
income between partner and self (yuan, continuous); and time, with the order determined by trying all possible
difference in education between partner and self (based on variables and using those that gave the best model fit.
the six education categories, continuous). In addition, it is Variables were retained for subsequent models if they
possible that orgasm frequency, along with other sexual improved model fit. Model fit is ascertained using informa-
behavior variables, is influenced by degree of westernization tion from the Bayesian information criterion (BIC) and
(Parish et al., 2007). The control for education will partly Akaike information criterion (AIC), which are higher-is-
deal with differing degrees of westernization; however, we worse model fit statistics (Burnham & Anderson, 2004;
also included the regional location within China (six Kuha, 2004). As a rule of thumb, one model generally has an
categories) as different regions are westernized to differing unacceptably worse fit compared with another if that model
extents. Table 1 shows a summary of all the variables and the scores more than 10 units higher in the AIC or BIC
categories of response available within them. (Burnham & Anderson, 2004; Raftery, 1996).
For the analysis, we employed ordinal regression using Parameters for individual variables in the models were
the PLUM (polytomous logit universal model) algorithm estimated by maximum log likelihood. The parameter
estimates are logit coefficients (λ) that can be converted to
odds ratios [exp(λ)] for ease of interpretation. We also report
Table 1 the Nagelkerke R2 for each model. This is a pseudo-R2
Descriptive statistics for the variables analyzed (frequencies or mean±S.D.
measure giving an indication of the explained variance in
values, as appropriate)
orgasm frequency in each model (Nagelkerke, 1991). It
Male income (yuan) 986.69±1195.96
should be borne in mind, however, that model selection is
Male height (m) 1.71±0.059
Sexual relationship duration (years) 14.99±9.69 always based on model fit statistics such as AIC and BIC
Age (years) 38.99±9.57 rather than Nagelkerke R2. For all analyses, there was
Difference in education (category) −0.28±0.98 absolute parameter and log likelihood convergence.
Difference in wealth (yuan) −369.69±1070.81
Happiness
Very happy 280
Relatively happy 1055 3. Results
Not too happy 185
Very unhappy 14 The descriptive statistics for the variables in the models
Region are summarized in Table 1, while the results of the ordinal
Coastal South 319
regression models are summarized in Table 2. Model 1 tested
Coastal East 331
Inland South 156 whether partner income and partner height predicted orgasm
North 241 frequency. Partner income proved to be a highly significant
North East 279 predictor of female orgasm frequency, while the estimate for
Central West 208 partner height was not significant (although there was a
Health
trend, .5bpb.1). The odds ratio for partner income [exp(λ)]
Excellent 342
Good 582 was 1.28. This means that an increase by 1 S.D. in partner
Fair 461 income increases the odds of a positive shift by a category in
Not good 139 orgasm frequency by a factor of 1.28.
Poor 10 Given that partner height was not a significant predictor at
Education
the 5% level, it was excluded from model 2, resulting in an
No school 90
Primary 267 improved model fit and a slightly increased parameter
Lower middle 583 estimate for partner income (Table 2, model 2).
Upper middle 425 Model 3 additionally included self-reported happiness.
Junior college 125 This improved model fit (in comparison with model 1) and
University 44
some parameter estimates for happiness were significant,
Orgasm frequency (dependent variable)
Always 121 with “very happy” women more likely to have a positive
Often 408 shift in orgasm frequency relative to “very unhappy” women
Sometimes 762 (Table 2, model 3). However, partner income remained
Rarely 182 highly significant in this model, with the odds ratio [exp(λ)]
Never 61
only slightly reduced at 1.26.
4

Table 2
Summaries of ordinal regression models
Model 1 2 3 4 5 6 7 8 9 10
Male wealth (1 S.D.) 0.243⁎⁎⁎⁎ 0.247⁎⁎⁎⁎ 0.23⁎⁎⁎⁎ 0.132⁎⁎ 0.198⁎⁎⁎⁎ 0.18⁎⁎⁎⁎ 0.224⁎⁎⁎⁎ 0.228⁎⁎⁎⁎ 0.27⁎⁎⁎ 0.172⁎⁎⁎
Male height (1 S.D.) 0.082† – – – – – – – – –
Happiness
Very happy – – 1.09⁎ 1.36⁎⁎ 1.10⁎ 0.87† 0.73 1.08⁎ 1.08⁎ 0.84
Relatively happy – – 0.84† 1.05⁎ 0.84† 0.61 0.52 0.84† 0.83 0.57
Not too happy – – 0.22 0.47 0.20 −0.02 0.05 0.22 0.22 −0.07
Very unhappy – – 0 0 0 0 0 0 0 0
Education
No school – – – −1.50⁎⁎⁎⁎ – – – – – –
Primary – – – −0.52 – – – – – –
Lower middle – – – −0.06 – – – – – –
Upper middle – – – 0.43 – – – – – –
Junior college – – – 0.29 – – – – – –
University – – – 0 – – – – – –
Region
Coastal South – – – – 0.69⁎⁎⁎⁎ – – – – –
Coastal East – – – – 0.12 – – – – –
Inland South – – – – 0.15 – – – – –
North – – – – 0.09 – – – – –
North East – – – – 0.42⁎ – – – – –
Central West – – – – 0 – – – – –
Age woman (1 S.D.) – – – – – −0.38⁎⁎⁎⁎ – – – –
Health
Excellent – – – – – – 1.64⁎⁎ – – –
Good – – – – – – 1.66⁎⁎ – – –
Fair – – – – – – 1.50⁎ – – –
Not good – – – – – – 1.04⁎ – – –
Poor – – – – – – 0 – – –
Education difference (1 S.D.) – – – – – – – 0.07 – –
Wealth difference (1 S.D.) – – – – – – – – 0.05 –
Relationship duration (1 S.D.) – – – – – – – – – −0.47⁎⁎⁎⁎

n parameters 2 1 4 9 9 6 9 6 6 6
Nagelkerke R2 0.019 0.017 0.035 0.095 0.053 0.074 0.053 0.036 0.035 0.076
−2LL 1868.1 405.55 752.4 1438.42 1623.27 2894.12 1366.09 1448.83 1887.57 2864.44
AIC 1872.1 407.55 760.4 1456.42 1641.27 2906.12 1384.09 1460.83 1899.57 2876.44
BIC 1882.77 412.89 781.73 1504.44 1689.29 2938.13 1432.11 1492.844 1931.58 2908.45
Parameter estimates are logit coefficients. −2LL indicates −2 log likelihood.
⁎ pb.05.
⁎⁎ pb.01.
⁎⁎⁎ pb.001.
⁎⁎⁎⁎ pb.0001.
†
.1bpb.05 based on Wald test.

Model 3 cannot be further improved in terms of model associations between orgasm frequency and partner wealth
fit by adding more control variables (Table 2, models without controlling for any other variable.
4–10; each additional variable introduced worsened the
model fit by more than 650 units of AIC/BIC). In all of
models 4–10, the estimates for male wealth remained 4. Discussion
highly significant and relatively unchanged (odds ratios
from 1.14 to 1.31), while the effects of self-reported In a large representative sample of the Chinese popula-
happiness varied according to which other variables were tion, we found evidence that women's self-reported orgasm
present in the analysis. frequency increases with the income of their partner. The
To summarize the results, increasing partner income had a effect of partner income is not an artifact of female age,
highly significant positive effect on women's self-reported educational attainment, happiness, health, relationship dura-
frequency of orgasm, and this effect remained robust when tion, regional differences, and differences between partners
all the control variables were entered in the model. The effect in educational attainment and wealth. Thus, the predictions
can be visualized from Fig. 1, which is based on simple of the functional hypothesis of human female orgasm were
 ARTICLE IN PRESS
T.V. Pollet, D. Nettle / Evolution and Human Behavior xx (2009) xxx–xxx 5

from the respondents' home, without their partner present and

with the respondents able to input their responses directly into
the computer if they so wished (Parish et al., 2003). Thus, the
perceived anonymity and confidentiality were high.
Second, there may be assortative mating of desirable men
with women susceptible to be highly orgasmic. There is a
heritable component to female orgasmic function (Dunn et
al., 2005), but nothing is known about whether there are
genetic correlations with other phenotypic attributes that
might be involved in mate choice. If there were assortative
mating, this would be consistent with an evolved function for
female orgasm, albeit a different one from that discussed so
far, in terms of signaling female quality. However, we note
that the data show that the more frequent orgasms of women
paired to high-income men are not explicable by those
women being healthier or happier overall.
The third interpretation is that more desirable mates cause
women to experience more orgasms. This is the interpretation
most consistent with the functional view of the human female
orgasm discussed in Section 1. Given that orgasmic function
can be influenced by desire to conceive (Singh, Meyer,
Zambarano, & Hurlbert, 1998) and by phase of the menstrual
cycle (Matteo & Rissman, 1984) and that it is closely tied to
sexual satisfaction (Parish et al., 2007), it is plausible that it
would also be enhanced by sought-after partner character-
istics. Interpreted in this way, our results confirm on a much
larger scale the conclusions drawn from samples containing
predominantly students (Shackelford et al., 2000; Thornhill
et al., 1995). However, Shackelford et al. (2000) investigated
only rated physical attractiveness, while Thornhill et al.
(1995) found an effect of partner physical symmetry but no
effect of partner financial prospects. We were not able to
Fig. 1. Relationship between partner income and self-reported orgasm replicate the findings with regard to physical symmetry and
frequency illustrated in two ways. (A) Mean and 95% confidence interval attractiveness since these data were not gathered in the
values of partner income for women reporting different frequencies of Chinese Health and Family Life Survey, and the differing
orgasm. (B) Proportion of women reporting orgasm frequency as “never” or
“always” by quartile of partner income.
conclusion with regard to financial status may stem from the
modest size and limited range of financial status in their data.
However, we do not know the specific proximate mechan-
met with respect to partner income but not partner height. isms behind the relationship between male financial status
One potential explanation for this is that there is no strong and female orgasm frequency. Moreover, we cannot
positive relationship between male height and reproductive definitively rule out a by-product explanation: it remains
success in contemporary China (Fielding et al., in press); possible that wealth relates to other traits that do not serve any
thus, apparently, male height is not a sought-after character- adaptive function and hence give rise to a spurious relation-
istic in this population. ship between male wealth and female orgasm frequency. At
There are three possible interpretations of how the effect face value, however, these data appear consistent with a
for partner income arises. First, the results might be an artifact functional view of the human female orgasm.
of reporting bias. Reporting bias can act in two ways, either by The data produced so far, while apparently consistent with
women with frequent orgasms overestimating their partner's an adaptive role for female orgasm, are far from definitive.
income or by women with high-quality mates overreporting Moreover, even if consistent with an adaptive role for female
their orgasm frequency. These possibilities are difficult to orgasm, these data do not allow conclusive testing between
exclude from data of this kind (Shackelford et al., 2000). two alternative proposed functions—namely, that female
However, self-report is the only practical source of data on orgasm differentially promotes emotional bonding with high-
human female orgasm in large samples and is normally taken quality males or that it differentially promotes conception
at face value in this literature (Dunn et al., 2005; Meston et al., with such males under conditions of sperm competition
2004; Thornhill et al., 1995). While we cannot rule out (Smith, 1984; Thornhill et al., 1995). Income is both an
reporting bias, we note that the interviews took place away indicator of value as a long-term investing partner and
 ARTICLE IN PRESS
6 T.V. Pollet, D. Nettle / Evolution and Human Behavior xx (2009) xxx–xxx

potentially an indicator of underlying genetic characteristics, Hrdy, S. B. (1996). The evolution of female orgasms: Logic please but no
so either hypothesis could predict that it would be important. atavism. Animal Behaviour, 52, 851−852.
Kuha, J. (2004). AIC and BIC: Comparisons of assumption and
Further research is needed to elucidate which of these, or if performance. Sociological Methods & Research, 33, 188−229.
both, would be the most important function of the human Laumann, E. O., Gagnon, J. H., Michael, R. T., & Michaels, S. (1994). The
female orgasm. In addition, a more thorough investigation social organization of sexuality: Sexual practices in the United States.
into the proximal mechanisms relating to orgasm frequency, Chicago: Chicago University Press.
as well as demonstrating a close relationship between female Levin, R. J. (2004). An orgasm is… who defines what an orgasm is? Sexual
and Relationship Therapy, 19, 101−107.
orgasm frequency and tangible fitness benefits, is necessary Levin, R. J., & Wagner, G. (1985). Orgasm in women in the laboratory—
in order to completely rule out a by-product explanation for quantitative studies on duration, intensity, latency and vaginal blood
the human female orgasm. flow. Archives of Sexual Behaviour, 14, 439−449.
Masters, W. H., & Johnson, V. E. (1966). Human sexual response. Boston:
Acknowledgments Little, Brown.
Matteo, S., & Rissman, E. F. (1984). Increased sexual activity in the
We thank the Population Center at NORC and the midcycle portion of the human menstrual cycle. Hormones and
Behavior, 18, 249−255.
University of Chicago for providing public access to these
McCullagh, P. (1980). Regression models for ordinal data. Journal of the
data. Information about the Chinese Health and Family Life Royal Statistical Society, 42, 109−142.
Survey, as well as due acknowledgement of its original Meston, C., Levin, R. J., Sipski, M. L., Hull, E. M., & Heiman, J. R. (2004).
funding, can be found at http://www.spc.uchicago.edu/prc/ Women's orgasm. Annual Review of Sex Research, 15, 173−257.
chfls.php. We also thank the editor and two anonymous Mueller, U., & Mazur, A. (2001). Evidence of unconstrained directional
selection for male tallness. Behavioral Ecology and Sociobiology, 50(4),
referees who provided feedback that greatly improved a
302−311.
previous version of the manuscript. Nagelkerke, N. J. D. (1991). A note on a general definition of the coefficient
of determination. Biometrika, 78, 691−692.
References Nettle, D. (2002). Height and reproductive success in a cohort of British men.
Human Nature—an Interdisciplinary Biosocial Perspective, 13(4), 473−491.
Alcock, J. (1987). Ardent adaptationism. Natural History, 96, 4. Nettle, D., & Pollet, T. V. (2008). Natural selection for male wealth.
Alexander, R. D. (1979). Sexuality and sociality in humans and other primates. In American Naturalist, 172, 658−666.
A. Katchadourian (Ed.), Human sexuality: A comparative and develop- Parish, W. L., Laumann, E. O., Cohen, M. S., Pan, S. M., Zheng, H. Y.,
mental perspective (pp. 81−97). Berkeley: University of California Press. Hoffman, I., et al. (2003). Population-based study of chlamydial
Armstrong, B. G., & Sloan, M. (1989). Ordinal regression for epidemio- infection in China—A hidden epidemic. JAMA—Journal of the
logical data. American Journal of Epidemiology, 129, 191−204. American Medical Association, 289(10), 1265−1273.
Baker, R. R., & Bellis, M. (1993). Human sperm competition: Ejaculate Parish, W. L., Luo, Y., Stolzenberg, R., Laumann, E. O., Farrer, G., & Pan, S.
manipulation by females and a function for female orgasm. Animal (2007). Sexual practices and sexual satisfaction: A population based
Behaviour, 46, 887−909. study of Chinese urban adults. Archives of Sexual Behavior, 36, 5−20.
Borgerhoff Mulder, M. (1990). Kipsigis women's preferences for wealthy Parish, W. L., Laumann, E. O., & Mojola, S. A. (2007). Sexual behavior in
men: Evidence for female choice in mammals. Behavioral Ecology and China: Trends and comparisons. Population and Development Review,
Sociobiology, 27, 255−264. 33(4), 729−756.
Burnham, K. P., & Anderson, D. R. (2004). Multimodel inference: Pawlowski, B., Dunbar, R. I. M., & Lipowicz, A. (2000). Evolutionary fitness
Understanding AIC and BIC in model selection. Sociological Methods —Tall men have more reproductive success. Nature, 403(6766), 156.
& Research, 33, 261−304. Pollet, T. V., & Nettle, D. (2008). Driving a hard bargain: Sex ratio and male
Buss, D. M. (1989). Sex differences in human mate preferences: marriage success in a historical US population. Biology Letters, 4(1), 31−33.
Evolutionary hypotheses tested in 37 cultures. Behavioral and Brain Raftery, A. E. (1996). Approximate Bayes factors and accounting for model
Sciences, 12, 1−49. uncertainty in generalized linear regression models. Biometrika, 83,
Chen, J., Campbell, T. C., Li, J., & Peto, R. (1990). Diet, life-style and 251−266.
mortality in China. New York: Cornell University Press. Samaras, T. T., Elrick, H., & Storms, L. H. (2003). Is height related to
Chen, Y., & Zhou, L. -A. (2007). The long-term health and economic longevity? Life Sciences, 72, 1781−1802.
consequences of the 1959–1961 famine in China. Journal of Health Shackelford, T. K., Weekes-Shackelford, V. A., LeBlanc, G. J., Bleske, A.
Economics, 26, 659−681. L., Euler, H. A., & Hoier, S. (2000). Female coital orgasm and male
Dunn, K. M., Cherkas, L. F., & Spector, T. D. (2005). Genetic influences on attractiveness. Human Nature, 11, 299−306.
variation in female orgasmic function: A twin study. Biology Letters, 1, Singh, D., Meyer, W., Zambarano, R. J., & Hurlbert, D. F. (1998). Frequency
260−263. and timing of coital orgasm in women desirous of becoming pregnant.
Fieder, M., & Huber, S. (2007). The effects of sex and childlessness on the Archives of Sexual Behavior, 27, 15−29.
association between status and reproductive output in modern society. Smith, R. L. (1984). Human sperm competition. In R. L. Smith (Ed.), Sperm
Evolution and Human Behavior, 28(6), 392−398. competition and the evolution of animal mating systems (pp. 601−660).
Fielding, R., Schooling, C. M., Adab, P., Cheng, K. K., Lao, X. Q., Jiang, C. London: Academic Press.
Q. et al. (in press). Are longer legs associated with enhanced fertility in Symons, D. (1979). The evolution of human sexuality. Oxford: Oxford
Chinese women? Evolution and Human Behavior, doi:10.1016/j. University Press.
evolhumbehav.2008.06.003. Thornhill, R., Gangestad, S. W., & Comer, R. (1995). Human female orgasm
Fisher, H. (1992). The anatomy of love. New York: Norton. and mate fluctuating asymmetry. Animal Behaviour, 50, 1601−1615.
Gould, S. J. (1987). Freudian slip. Natural History, 96, 14−21. Troisi, A., & Carosi, M. (1998). Female orgasm rate increases with male
Hensley, W. E. (1994). Height as a basis for interpersonal attraction. dominance in Japanese macaques. Animal Behaviour, 56, 1261−1266.
Adolescence, 29(114), 469−474. World Bank (2006). World development indicators 2006. http://devdata.
Hopcroft, R. L. (2006). Sex, status, and reproductive success in the contemporary worldbank.org/wdi2006/contents/cover.htm accessed on 22 September
United States. Evolution and Human Behavior, 27(2), 104−120. 2008.