Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

TESTOSTERONE, SMILING, AND

FACIAL APPEARANCE
James M. Dabbs, Jr.

ABSTRACT: In a study of possible links between testosterone and dominance, 119


men and 114 women provided saliva samples for testosterone assay and posed
smiling and not smiling for portrait photographs. Expert judges viewing the photo-
graphs found smaller smiles among high than low testosterone men, with less zy-
gomatic major (raising the corners of the mouth) and orbicularis oculi (raising the
cheeks and crinkling around the corners of the eyes) muscle activity. Naive judges
viewing individual photographs gave higher potency ratings to smiling high tes-
tosterone men than smiling low testosterone men. Naive judges viewing photo-
graphs grouped into high and low testosterone sets gave higher potency and lower
goodness ratings to high than to low testosterone men, regardless of whether they
were smiling. Among women, judges found only slight relationships between tes-
tosterone and facial appearance. The pattern among men of less smiling with higher
testosterone levels fits with research linking testosterone to face-to-face dominance.

Research on testosterone and social behavior has emphasized antiso-


cial activity, including criminal violence (Dabbs, Carr, Frady, & Riad,
1995), delinquency (Dabbs & Morris, 1990), and marital disturbance
(Booth & Dabbs, 1993). It has also emphasized aggression (Archer, 1991),
although testosterone is probably related more closely to interpersonal
dominance than to aggression (Albert, Walsh, & Jonik, 1993; Mazur,
1985).
A problem with interpreting research findings in terms of dominance is
that dominance itself is not clearly defined. Dominance among animals
includes winning in competitive encounters, often through fighting. Suc-
cessful fighting is associated with testosterone. When two male rhesus

This research was supported National Institute of Mental Health grant MH42525 and
National Science Foundation grant BNS-9021393. I thank Paul Ekman and Mark Frank for
help in scoring the photographs; Paula Williams for making the photographs; Charles Cum-
mins, Denise de La Rue, Debira Plyler, and Ellen Steinberg for collecting data; and Earl Ander-
son, Aruna Rao, and Suzanne M. Bell for conducting assays.
Address correspondence to the author at Department of Psychology, Georgia State Uni-
versity, Atlanta, GA 30303, or at e-mail jdabbs@gsu.edu.

laurral of Nonverbal Behavior 2tm, Spring 1997


« 1997 Human Sciences Press. Inc. 45
46
JOURNAL OF NONVERBAL BEHAVIOR

monkeys fight, the higher testosterone one usually wins, and winning fur-
ther increases his testosterone level. The losing animal drops in tes-
tosterone (Bernstein, Rose, & Gordon, 1974). Human male athletes react
similarly, increasing in testosterone before a contest and increasing further
when winning (Booth, Shelley, Mazur, Tharp, & Kittok, 1989). Sports fans
increase in testosterone when their champions win, in a kind of hormonal
basking in reflected glory (Bernhardt, Dabbs, Fielden, & Lutter, 1996). In
the present paper, we regard dominance as a quality that helps one win
whatever one wants to win in one-on-one interpersonal encounters. The
winning can come through using overt force or subtle body movements
and expressions that evoke deference from others.
Social life has many moments of conflict, making it a good place in
which to study testosterone and dominance. Because increases and de-
creases in smiling are easily produced and often appear around moments
of potential conflict, they could serve as mediators between testosterone
and dominance. Research in the area intersecting testosterone, smiling,
and dominance involves empirical and theoretical questions. Empirical
questions ask whether in fact smiling does vary with testosterone, and the-
oretical questions ask what mechanisms might link testosterone to smiling,
and smiling to dominance.
Two studies report less smiling in high than in low testosterone indi-
viduals. One found less frequent smiling in fraternities whose members had
higher mean levels of testosterone (Dabbs, Hargrove, & Heusel, 1996). The
other found less frequent smiling in higher testosterone women engaged in
group discussions (Cashdan, 1995). These studies did not describe or ex-
amine smiles in detail.
Regarding mechanisms, there is no known link between testosterone
and activation of the muscles that produce smiles. Testosterone does not
affect behavior directly. It builds body mass and affects development and
functioning of organ systems, including the central nervous system. Tes-
tosterone molecules bind to receptors in the amygdala, hypothalamus, and
preoptic area of the brain. Through this route testosterone might affect the
brain, influencing thoughts and feelings and leading to changes in smiling.
The relationship of smiling to dominance depends to some extent
upon the kind of smile and the setting in which it appears. Smiles of confi-
dence should relate positively to dominance, as should condescending,
ridiculing, or sneering smiles. But although one can smile and be a villain,
we usually associate smiles with pleasantness, friendliness, graciousness,
and a desire to please. This gentler smiling conveys positive feelings to
others (Hess, Banse, & Kappas, 1995). It reduces conflict by showing we
are friendly or at least benign. This view leads us to expect more smiling
47

JAMES M. DABBS, JR.

among individuals who are less dominant, with lower levels of tes-
tosterone.
Most researchers treat smiling as a behavior that fluctuates with mood,
thought, and influence tactics, but smiling is also characteristic of individ-
uals. Hall's (1984) work on gender and Seaford's (1976) paper on the
Southern smile point to group differences in smiling. Dominance can be
conveyed by chacteristic low smiling, as it is by glances (Exline, 1972) and
facial expressions of contempt (Ekman & Heider, 1988). Dominance can
also be conveyed by physiognomy, as found in studies predicting future
leadership among West Point cadets and Army officers from their yearbook
photographs (Mazur, Mazur, and Keating, 1984; Mazur & Mueller, 1996).
Testosterone levels are relatively constant across days, weeks, and years
(Dabbs, 1990), and thus if smiling is correlated with low testosterone
levels, it is correlated with a relatively stable individual difference charac-
teristic.
In the present study we examined testosterone, smiles, and facial ap-
pearance among male and female college students. We assayed testos-
terone from saliva and scored smiling and facial appearance from portrait
photographs.

Method

Subjects were 119 male and 114 female undergraduate students, mean age
20.6 years (SD = 3.8, range 17-40) for males and 20.8 years (SD = 4.7,
range 17-44) for females. Subjects had earlier provided saliva samples for
testosterone assay and posed for portrait photographs in a study of the
reliability of salivary testosterone measurements (Dabbs, 1990). The pres-
ent study compared subjects' testosterone levels with their appearances in
the photographs.

Photographs
We took two photographs of each subject, one smiling and one not
smiling. The photographer was a friendly and attractive female under-
graduate. She took the photographs in an alcove in a social psychology
laboratory, always taking the non-smile photograph first. The subject stood
in front of a wall, with an identification number beside his or her head. The
photographer asked the subject to look at the camera, relax, and not smile,
and she took a color photograph from 6 ft. away with a hand-held 35-mm
camera with a portrait lens, framing the subject's face. Then she lowered
48

JOURNAL OF NONVERBAL BEHAVIOR

the camera, paused, smiled at the subject, said, "Okay, now smile," and
raised the camera and took a second photograph.

Testosterone Measurements
Subjects provided all saliva samples for testosterone assay at 10:00
a.m., to control for diurnal fluctuations in testosterone level (Dabbs, 1990).
We measured testosterone from saliva rather than serum because salivary
and serum testosterone levels are highly correlated, and it is much easier to
get subjects to participate using saliva (Dabbs, 1993; Navarro, Juan, &
Bonnin, 1986). Saliva samples were stored frozen and assayed in duplicate
using a radioimmunoassay (RIA) procedure with 125l-Testosterone tracer,
ether extraction, and charcoal separation. Mean within-assay coefficient of
variation (CV) between sample duplicates was less than 10% for each sex.
Between-assay CV, based upon mean values of control pool samples in
each assay, was 12% for males and 21% for females. There were two
groups of subjects, one in which subjects provided samples on a single
day, and the other in which they provided samples on two consecutive
days. Subjects with samples from two days received testosterone scores
representing the mean of the two days. We used a logarithmic transforma-
tion to normalize the scores, because the raw distribution of testosterone,
like that of many hormones, is negatively skewed. A mean difference be-
tween the two groups of subjects (which commonly appears in hormone
research, due to variations in laboratory procedure or materials) was re-
moved by standardizing scores within each group.

Smiles and Action Unit Scoring


Informal examination of the photographs suggested high testosterone
men looked less friendly and more tense around the mouth than low tes-
tosterone men, and high testosterone women had bolder and more direct
gaze than low testosterone women. The difference between non-smile and
smile photographs was scored systematically by an expert judge using the
Emotion Facial Action Coding System (Friesen, 1986), a variation of Ekman
and Friesen's (1978) Facial Action Coding System. The judge did not know
the subjects' testosterone levels. The action units scored were AU12 (zy-
gomatic major activity, moving the corners of the mouth upward and out-
ward), AU6 (orbicularis oculi activity, raising the upper cheeks and pro-
ducing crinkling around the outer corners of the eyes), AU25 (lips parted),
and AU26 (lips parted and jaws slightly open). AU12 was scored on a five-
point scale with 1 = zero or barely perceptible muscle contraction, 2 =
49
JAMES M. DABBS, JR.

slight contraction barely meeting minimal requirement for scoring, 3 =


contraction definitely meeting requirements for scoring, 4 = very strong
contraction, and 5 = full contraction. AU12 was also scored on a two-
point scale indicating presence or absence of a very strong or full contrac-
tion. Asymmetry of smile was scored when zygornatic activity (AU12) was
at least one point stronger on one side of the mouth than on the other.
AU6, AU25, AU26, and asymmetry were scored as present or absent on
2-point scales.

Overall Judgments
There is more information in the face than just that revealed by action
unit scores alone. Expressions change continuously, and many expressions
are very faint. Judges sometimes see muscle activity which they regard as
not strong enough to warrant scoring. Other scoring systems might use the
information missed by action unit scoring. Such information might reside
in slight and poorly delineated expressions of an individual, or in the com-
bined expressions of many individuals in a group, where the impact of the
group impression transcends the impact of the impression produced by a
single group member alone. In trying to go beyond the action unit scoring,
we divided the photographs into four sets, comprising smiling men, smiling
women, non-smiling men, and non-smiling women. Judges were college
students who did not know the purpose of the study. Different judges per-
formed one of three tasks, as follows.
First, 72 judges rated individual photographs on the semantic differen-
tial dimensions of potency, activity, and goodness (Osgood, Suci, & Tan-
nenbaum, 1957). Each judge sorted one of the four sets of photographs into
piles marked "least," "average," and "most" in response to one of the fol-
lowing questions: "Who looks strong and dominant?" "Who looks active
and energetic?" and "Who looks good and friendly?" Each photograph re-
ceived mean potency, activity, and goodness scores based on the ratings of
six judges.
Second, 40 judges rated groups of photographs. They rated groups
because we found observers could sometimes detect group differences
more easily than individual differences. For example, a visitor who noticed
the photographs of high and low testosterone men laid out in two groups
on a table, with no identification, remarked that the smiles in one group
(the high testosterone one) looked "hard." We grouped together photo-
graphs of the 12 highest and 12 lowest testosterone subjects from each of
the four sets (smiling men, smiling women, non-smiling men, and non-
smiling women), and we placed these high and low testosterone groups at
SO

JOURNAL OF NONVERBAL BEHAVIOR

the top and bottom of 18 X 24-in. posters. Half the judges viewed posters
with high testosterone subjects at the top, and half viewed posters with low
testosterone subjects at the top. Each judge viewed four posters, represent-
ing the four sets of photographs, in one of four different orders. Judges
studied each poster for two minutes and indicated whether photographs at
the top or the bottom looked more "strong and dominant," more "active
and energetic," and more "good and friendly."
Third, 41 judges rated photographs of male subjects on facial maturity
(Zebrowitz, Olson, & Hoffman, 1993). They rated only males, because
most testosterone effects in the semantic differential ratings had appeared
among males. Seventeen judges rated non-smile photographs, and 24 rated
smile photographs. They viewed photographs one at a time and rated each
on a 7-point scale anchored at 1 = "Baby faced" and 7 = "Mature face"
(Zebrowitz et al., 1993). We assigned a facial maturity score to each sub-
ject based on the mean of his non-smile and smile ratings. Ratings were
limited to subjects 18-23 years old (N = 100), to reduce the effects of any
confounding of age with facial maturity and testosterone level.

Results

Smiles and Action Units


Among men, the smiles of higher testosterone subjects showed less
AU6 crinkling around the eyes, r(117) = -.19, p < .05, and less AU12
upward and outward movement of the corners of the mouth, r(117) =
— .17, p < .07. When AU12 was reduced to a two-point scale indicating
presence or absence of a very strong or full zygomatic muscle contraction,
its negative correlation with testosterone increased from r = — .17 to r =
— .25 (p < .01). Testosterone was not related to other facial action unit
scores among males, and it was not related to any action unit scores in the
smiles of women. Among women, the correlation of testosterone with AU6
was r(112) = .04, ns, and with AU12 was r (112) = -.05, ns.

Overall Judgments
Table 1 shows correlations between testosterone level and judges' rat-
ings of the four sets of photographs on strong/dominant, active/energetic,
and good/friendly dimensions. The one significant finding in these data
was that among smiling men, those higher in testosterone looked more
strong/dominant than those lower in testosterone.
Table 2 shows the percent of judges choosing high testosterone photo-
graphs grouped together as more potent, active, and good than low tes-
51

JAMES M. DABBS, JR.

TABLE 1

Correlation of Subjects' Testosterone Levels with Judges' Semantic


Differential Ratings in Four Sets of Photographs

Semantic differential dimension

Photograph set Strong/dominant Active/energetic Good/friendly

Smiling males .19* .00 –.03


Non-smiling males .00 .17 .02
Smiling females –.02 –.15 .00
Non-smiling females –.05 –.02 .00

Note. Cell entries are product-moment correlations between subjects' testosterone levels
and judges' mean ratings of "strong/dominant," "active/energetic," and "good/friendly." There
were 119 male subjects, 114 female subjects, and 72 judges. Each entry is a correlation based
upon mean ratings of 6 judges judging one set of photographs on one semantic dimension.
* p < .05

TABLE 2

Percent of Judges Choosing High Testosterone Group as more


Strong/dominant, Active/energetic, and Good/friendly Than Low
Testosterone Group in Four Sets of Photographs

Semantic differential dimension

Photograph set Strong/dominant Active/energetic Good/friendly

Smiling males 82*** 68* 18***


Non-smiling males 80*** 55 18***
Smiling females 65 65 48
Non-smiling females 65 68* 65
Note. Judges chose between two groups of photographs displayed together, comprising
the 12 highest and 12 lowest testosterone subjects in a particular set of photographs. Cell
entries show the percent of judges (n = 40) choosing high testosterone groups over low
testosterone groups on each semantic differential dimension. Judges would be expected by
chance to choose the high testosterone group 50% of the time, and significance of departures
from 50% was tested using the normal approximation of the binomial distribution. Ratings in
different cells are not independent, because all judges chose between lows and highs in all
cells.
* p < .05
*** p < .001
52

JOURNAL OF NONVERBAL BEHAVIOR

tosterone photographs grouped together. Here judges chose high tes-


tosterone men, smiling or not, significantly more often as strong/dominant
and less often as good/friendly than low testosterone men. Judges chose
groups of high testosterone women more often than groups of low tes-
tosterone women on all three dimensions, but the effect was significant
only for the active/energetic dimension.
Facial maturity, which we scored only for male subjects, was related
significantly to age, r<98) = .52, p < .001, but not testosterone, r(98) =
.13, ns. Age in this sample was not significantly related to testosterone,
t(98) = .12, ns. Facial maturity was correlated with semantic differential
ratings of strong/dominant, r(98) = .20, p < .05, but not of active/ener-
getic, r(98) = -.11, ns, or good/friendly, r(98) = -.03, ns.

Discussion

High testosterone men were judged more potent and less good than low
testosterone men. Low testosterone men had larger smiles, indicated by
more AU12 and AU6 muscle activity and suggestive of more true enjoy-
ment (Ekman, Davidson, & Friesen, 1990). Low testosterone men looked
more friendly and less dominant than high testosterone men. These results
fit with an earlier finding that ministers, who are more benevolent than
many other men, have low levels of testosterone (Dabbs, de La Rue, &
Williams, 1990).
High and low testosterone women did not differ in smiling, although
other studies of women have found testosterone related to libido (Sherwin,
Gelfand, & Brender, 1985), violence (Dabbs, Ruback, Frady, Hopper, &
Sgoutas, 1988), and occupational choice (Schindler, 1979), and Cashdan
(1995) has reported high testosterone women smile less often than low
testosterone women. Women generally smile more often than men (Hall &
Halberstadt, 1986). Perhaps being willing to smile, along with feeling
obliged to smile for a photographer, overwhelmed individual differences
among women in the present setting. Cashdan had observed women sub-
jects in informal group discussions, where they could respond to their own
tendencies rather than to the demands of a photography session.
We do not know whether high testosterone men tend always to smile
less, or whether they just smiled less in the present setting, perhaps in
reaction against being told to smile. We would not expect a strong link
between testosterone and smiling, because testosterone levels are relatively
stable (Dabbs, 1990; Vermeulen & Verdonck, 1992) and smiles have a
fleeting presence affected by the mood of the individual and the social
demands of the setting (Hess et al., 1995). Both testosterone and smiling
53

JAMES M. DABBS, JR.

are likely related to dominance, however, and their relationship to each


other should become increasingly apparent as subjects are observed over
longer periods of time. The correlation between high testosterone and low
smiling may be greater in settings where confrontation and dominance are
salient issues.
Semantic differential ratings distinguished between high and low tes-
tosterone subjects viewed in groups, but not between low and high tes-
tosterone subjects viewed individually. This is reminiscent of a fraternity
study in which small mean differences in testosterone were associated with
sizeable differences in behavior (Dabbs et al., 1995). Group ratings may
reveal effects that accumulate to the point where they are noticed by an
observer, although the effects are too slight to be noticed in individuals one
at a time.
In focusing upon smiles, we should not ignore information available in
non-smiling faces. Even when our subjects tried to show no expression at
all, judges reported those higher in testosterone looked more potent and
less good. The judges may have responded to faint facial expressions or to
physical features of the face. Mazur and his colleagues found that the facial
structure of non-smiling West Point cadets predicted the behavior of these
men decades later (Mazur et al., 1984; Mazur & Mueller, 1996).
Although high testosterone levels are heritable (Meikle, Stringham,
Bishop, & West, 1988) and appear correlated with low levels of smiling,
this does not mean testosterone causes low levels of smiling. Smiling could
be heritable, as suggested by the finding that happiness is heritable (Lykken
& Tellegen, 1996), or testosterone and smiling could be both determined
by other variables. Highly male-identified men may smile less frequently
than others (LaFrance & Carmen, 1980) because high testosterone levels
lead them to identify strongly with male stereotypes, which in turn limits
how much they smile.
The present study presents an empirical relationship between tes-
tosterone and smiling and a theoretical link that suggests smiling is more
appropriate for individuals who are less dominant. We plan to extend the
research in a quasi-experimental direction, observing smiling among peo-
ple for whom testosterone levels are raised or lowered for medical pur-
poses and among people with naturally low or high testosterone levels
who encounter threatening or challenging social settings.

References

Albert, D.)., Walsh, M. L., & Jonik, R. H. (1993). Aggression in humans: What is its biological
foundation? Neuroscience and Biobehavioral Reviews, 17, 405-425.
54
JOURNAL OF NONVERBAL BEHAVIOR

Archer, J. (1991). The influence of testosterone on human aggression. British Journal of Psy-
chology, 82, 1-28.
Bemhardt, P. C., Dabbs, J. M., Jr., Fielden, J. A., & Lutter, C. (1996). Testosterone changes
during vicarious experiences of victory and defeat in spectators of sporting events. Manu-
script under editorial review.
Bernstein, I. S., Rose, R. M., & Gordon, T. P. (1974). Behavioral and environmental events
influencing primate testosterone levels, journal of Human Evolution, 3, 517-525.
Booth, A., & Dabbs, J. M., Jr. (1993). Testosterone and men's marriages. Social Forces, 72,
463-477.
Booth, A., Shelley, C., Mazur, A., Tharp, G., & Kittok, R. (1989). Testosterone, and winning
and losing in human competition. Hormones and Behavior, 23, 555-571.
Cashdan, E. (1995). Hormones, sex, and status in women. Hormones and Behavior, 29, 354-
366.
Dabbs, J. M., Jr. (1990). Salivary testosterone measurements: Reliability across hours, days,
and weeks. Physiology and Behavior, 48, 83-86.
Dabbs, J. M., Jr. (1993). Salivary testosterone measurements in behavioral studies. Annals of
the New York Academy of Sciences, 694, 177-183.
Dabbs, J. M., Jr., Carr, T. S., Frady, R. L., & Riad, J. K. (1995). Testosterone, crime, and mis-
behavior among 692 male prison inmates. Personality and Individual Differences, 18,
627-633.
Dabbs, J. M., Jr., de La Rue, D., & Williams, P. M. (1990). Testosterone and occupational
choice: Actors, ministers, and other men. Journal of Personality and Social Psychology,
59, 1261-1265.
Dabbs, J. M., Jr., Hargrove, M. F., & Heusel, C. (1996). Testosterone differences among college
fraternities: Well-behaved vs. rambunctious. Personality and Individual Differences, 20,
157-161.
Dabbs, J. M., Jr., & Morris, R. (1990). Testosterone, social class, and antisocial behavior in a
sample of 4,462 men. Psychological Science, 1, 209-211.
Dabbs, J. M., Jr., Ruback, R. B., Frady, R. L., Hopper, C. H., & Sgoutas, D. S. (1988). Saliva
testosterone and criminal violence among women. Personality and Individual Differ-
ences, 9, 269-275."
Ekman, P., Davidson, R., & Friesen, W. V. (1990). Emotional expression and brain physiology
II: The Duchenne smile. Journal of Personality and Social Psychology, 58, 342-353.
Ekman, P., & Friesen, W. V. (1978). Facial action coding system: A technique for the measure-
ment of facial movement. Palo Alto, CA: Consulting Psychologists Press.
Ekman, P., & Heider, K. G. (1988). The universality of a contempt expression: A replication.
Motivation and Emotion, 12, 303-308.
Exline, R. V. (1972). Visual interaction: The glances of power and preference. In J. K. Cole
(Ed.), Nebraska Symposium on Motivation: Vol. 19. Current theory and research in mo-
tivation, (pp. 163-206). Lincoln: University of Nebraska Press.
Friesen, W. (1986). Recent developments in FACS—EMFACS. face Value: Facial Measure-
ment Newsletter, 1(1), 1-1.
Hall, J. A. (1984). Nonverbal sex differences: Communication accuracy and expressive style.
Baltimore: Johns Hopkins University Press.
Hall, J. A., & Halberstadt, A. G. (1986). Smiling and gazing. In J. S. Hyde & M. C. Linn (Eds.),
The psychology of gender. Advances through meta-analysis (pp. 136-158). Baltimore:
Johns Hopkins University Press.
Hess, U., Banse, R., & Kappas, A. (1995). The intensity of facial expression is determined by
underlying affective state and social stimuli. Journal of Personality and Social Psychology,
69, 280-288.
LaFrance, M., & Carmen, B. (1980). The nonverbal display of psychological androgyny. Jour-
nal of Personality and Social Psychology, 38, 36-49.
Lykken, D., & Tellegen, A. (1996). Happiness is a stochastic phenomenon. Psychological
Science, 7, 186-189.
55
JAMES M. DABBS, JR.

Mazur, A. (1985). A biosocial model of status in face-to-face primate groups. Social Forces,
64, 377-402.
Mazur, A., Mazur, J., & Keating, C. (1984). Military rank attainment of a West Point class:
Effects of cadets' physical features. American Journal of Sociology, 90, 125-150.
Mazur, A., & Mueller, U. (1996). Channel modeling: From West Point cadet to general. Public
Administration Review, 56, 191-198.
Meikle, A. W., Stringham, J. D., Bishop, D. T., & West, D. W. (1988). Quantitating genetic and
nongenetic factors influencing androgen production and clearance rates in men. Journal
of Clinical Endocrinology and Metabolism, 67, 104-109.
Navarro, M. A., Juan, L, & Bonnin, M. R. (1986). Salivary testosterone: Relationship to total
and free testosterone in serum. Clinical Chemistry, 32, 231-232.
Osgood, C. E., Suci, C. J., & Tannenbaum, P. H. (1957). The measurement of meaning. Ur-
bana, IL: University of Illinois.
Schindler, C. L. (1979). Testosterone concentration, personality patterns, and occupational
choice in women. Dissertation Abstracts International, 40, 1411 A. (University Microfilms
No. 79-19,403).
Seaford, H. W., Jr. (1976). Cultural facial expression. In E. Giles & J. S. Friedlaender (Eds.), The
measures of man (pp. 616-639). Cambridge, MA: Peabody Museum.
Sherwin, B. B., Gelfand, M. M., & Brender, W. (1985). Androgen enhances sexual motivation
in females: A prospective, crossover study of sex steroid administration in the surgical
menopause. Psychosomatic Medicine, 47, 339-351.
Vermeulen, A., & Verdonck, G. (1992). Representativeness of a single point plasma tes-
tosterone level for the long term hormonal milieu in men. journal of Clinical Endocrinol-
ogy and Metabolism, 74, 939-942.
Zebrowitz, L. A., Olson, K., & Hoffman, K. (1993). Stability of babyfaceness and attractiveness
across the life span. Journal of Personality and Social Psychology, 64, 453-466.

You might also like