Content uploaded by Lee Ellis
All content in this area was uploaded by Lee Ellis on Oct 06, 2019
Content may be subject to copyright.
Color preferences according to gender and sexual orientation
Lee Ellis *, Christopher Ficek
Minot State University, Minot, ND 58707, USA
Received 30 June 2000; received in revised form 13 November 2000; accepted 30 November 2000
The present study was undertaken to verify ®ndings surrounding sex dierences in color preferences, and
to extend this realm of inquiry by looking for possible dierences in color preferences associated with sexual
orientation. Based on a large sample of North American college students, signi®cant gender dierences were
found, with the main dierence being a greater preference for shades of blue by males than by females.
Females, on the other hand, tended to be more evenly divided than males between preferring both green
and blue. Regarding sexual orientation, we found no signi®cant dierences between heterosexuals and
homosexual/bisexuals of either gender. In other words, male and female homosexuals/bisexuals exhibited
essentially the same con®guration of color preferences as did their heterosexual counterparts. #2001
Elsevier Science Ltd. All rights reserved.
Studies have arrived at dierent conclusions regarding the existence of gender dierences in color
preferences. In a study of United States college students, Silver et al. (1988) found that greater
proportions of males chose blue as their favorite color than was true for females. In a subsequent
study of elderly persons in the United States, Silver and Ferrante (1995) reported four signi®cant
dierences in color preferences, with males again being more likely than females to prefer blue. Males
were also more likely to choose red, whereas females were more likely to prefer black and pink.
Three additional studies, all among elderly persons, concluded that there were no signi®cant
dierences between men and women in their preferences for various colors. Two of these studies
were conducted in the United States (Mather, Stare & Breinin, 1971; Tate & Allen 1985), and the
other was carried out in Sweden (Wijk, Berg, Sivik & Steen, 1999).
The present study was undertaken to verify the existence of gender dierences in color pref-
erence. In addition, we decided to look for evidence that color preference might also vary
according to sexual orientation.
0191-8869/01/$ - see front matter #2001 Elsevier Science Ltd. All rights reserved.
Personality and Individual Differences 31 (2001) 1375±1379
* Corresponding author. Tel.: +1-701-858-3241; fax: +1-701-839-6933.
E-mail address: email@example.com (L. Ellis).
The present study derived its data from a large investigation of gender-typical behavior patterns
in the United States and Canada conducted at 22 universities during the 1990s. A total of 1924
males and 3766 females provided answers to a question regarding their color preference. Their
ages ranged from 17 to 63, with a mean of 22.82 (S.D.=5.32) for the males and 22.54
(S.D.=5.34) for the females.
To measure sexual orientation, we asked subjects if they considered themselves to be ``hetero-
sexual, homosexual, bisexual, or undecided.'' Of those who provided answers to the question on
color preference, 53 (2.86%) of the males and 62 (1.71%) of the females declared themselves to be
either homosexual or bisexual.
To our question regarding the subjects' favorite color, subjects were given a brief space in
which to write their choice. For comparing males and females, we consolidated the responses into
11 dierent color categories, including one category for those who stated that they have no par-
Regarding our analysis of males and females according to sexual orientation, several of the
color categories were chosen by fewer than ®ve homosexuals/bisexuals, thus making meaningful
statistical comparison impossible. Therefore, when we compared sexes by sexual orientation, only
three speci®c color categories were used, with the remaining choices of colors subsumed under the
category of ``other'' (including no preference).
The color preferences expressed by males and females are shown in Table 1, with the colors
arranged roughly in the order of their appearance along the color spectrum. A chi-square test
revealed very signi®cant dierences between the proportional distributions of male and female
subjects in color preferences (
=435.202, d.f.=10, P=0.000).
The greatest dierences had to
do with preferences for blue and green. Whereas nearly half (45%) of all males chose various
shades of blue as their favorite color and less than 20% chose some shade of green, only 25% of
females chose blue, while about 28% chose green. Other noteworthy gender dierences were that
females were more likely to choose pink and purple, while males were more partial to black (this
latter ®nding is contrary to Silver and Ferrante (1995), who found females preferring black more
In the case of sexual orientation, Table 2 presents consolidated proportions of males and
females according to sexual orientation who chose each color (or shades thereof) as their favor-
ites. To assess statistical signi®cance, we again used chi square. However, because our samples of
homosexual/bisexuals of both genders were much smaller than the samples of heterosexuals of
both genders, there were fewer than ®ve observations for several of the colors as far as the
For details on the universities sampled for this study, see Ellis and Robb (2000).
The computer program utilized in this study (SPSS, Version 9) only carried its calculations of the degree of sta-
tistical signi®cance to three decimal points. Therefore, any probability with a 0.0001 or lower appears simply as 0.000.
1376 L. Ellis, C. Ficek / Personality and Individual Dierences 31 (2001) 1375±1379
homosexual/bisexual subjects were concerned. We compensated for this sampling shortfall by
collapsing all of the least frequently chosen color categories (plus ``no preference'') into an
``other'' category. This left the following four color categories for comparison: green, blue, black,
The resulting chi square for comparing heterosexual males and homosexual/bisexual males was
=1.959, d.f.=3, P=0.581. For the females, the chi square dierence between heterosexuals
and homosexual/bisexuals was w
=1.234, d.f.=3, P=0.745. Thus, in neither the case of males
nor females did sexual orientation have a signi®cant link to color preference. For comparison
purposes, we also performed a chi square in which all males were compared to all females. This
again yielded extremely signi®cant dierences (w
=310.955, d.f.=3, P=0.000)
, although not
quite as strong as when all eleven color categories were dierentiated.
Distribution of consolidated color preferences by gender and sexual orientation
Color preference Males Females
Heterosexual Homo/bisexual Heterosexual Homo/bisexual
Frequency % Frequency % Frequency % Frequency %
Green 340 18.9 14 26.4 999 28.0 17 27.4
Blue 817 45.4 21 39.6 896 25.1 13 21.1
Black 220 12.2 6 11.3 228 8.1 7 11.3
Other 424 23.5 12 22.6 1389 38.9 25 40.3
Total 180 100.0 53 99.9 3512 100.1 62 100.0
Distribution of consolidated color preferences by gender
Color preference Males Females
Frequency % Frequency %
No preference 19 1.0 95 2.5
Pink 9 0.5 199 5.3
Red 233 12.1 447 11.9
Orange 16 0.8 66 1.8
Yellow 19 1.0 100 2.7
Green 367 19.1 1051 27.9
Blue 866 45.0 938 24.9
Purple 98 5.1 459 12.2
Brown 13 0.7 19 0.5
Gray 22 1.1 7 0.02
Black 233 12.1 306 8.1
White 29 1.5 79 2.1
Total 1924 100.0 3666 99.9
L. Ellis, C. Ficek / Personality and Individual Dierences 31 (2001) 1375±1379 1377
The present study con®rms Silver and Ferrante's (1995) observation that males are signi®cantly
more likely than females to prefer variations on the color of blue. Given the very high degree of
statistical signi®cance for our ®ndings, we are surprised that three studies have failed to ®nd any
signi®cant gender dierences in color preference (Mather et al., 1971; Tate & Allen, 1985; Wijk et
al., 1999). This could possibly be due to the fact that all of these studies were of elderly popula-
tions, and perhaps gender dierences dissipate with age. We were unable to investigate this pos-
sibility since our sample was limited to college students.
Why would color preference vary by gender? Without ruling out any possibility at this point,
we are inclined to suspect the involvement of neurohormonal factors. Studies of rats have found
average sex dierences in the number of neurons comprising various parts of the visual cortex
(Reid & Juraska, 1992, 1995). Also, gender dierences have been found in rat preferences for the
amount of sweetness in drinking water (Lichtensteiger & Schlumpf, 1985; Shapiro & Gold-
man, 1973; Valenstein, Kakolewski & Cox, 1967). One experiment demonstrated that the sex
dierences in rat preferences for sweetness was eliminated by depriving males of male-typical
testosterone levels in utero (Lichtensteiger & Schlumpf, 1985). Perhaps, prenatal exposure to
testosterone and other sex hormones operates in a similar way to ``bias'' preferences for certain
colors in humans.
Focusing on some of the actual mechanisms that may be involved in humans, research by
Hoyenga and Wallace (1979) and McGuinness and Lewis (1976) found sex dierences in
retinal biochemistry and in how the brain processes color information. According to both
studies, females seemed to be more sensitive to pinks, reds, and yellows. In addition, the
Hoyenga and Wallace study identi®ed certain conditions under which males appeared to be
more sensitive to colors in the blue-green light spectrum. Nevertheless, caution must be exercised
in any attempts to explain preferences for various colors on the basis of sensitivity to those colors.
None of the above conjecture is meant to imply that cultural factors might not also
operate, although we are not aware of any cultural training or expectations regarding which
gender is supposed to prefer one color over another. There is, of course, a tendency in several
Western cultures to dress baby boys in blue and baby girls in pink, and one could argue that
dressing baby boys in blue might increase their preferences for blue later in life. If this were true,
we would expect to ®nd women exhibiting a corresponding bias toward shades of red. There
was a substantial tendency for greater number of women to choose pink as their favorite color
when compared to men, but women were actually less likely to prefer darker shades of red
Cross-cultural research could shed light on these issues by determining how varied gender dif-
ferences are in color preferences. Also, insight could come from comparing males and females of
other species as to any color preferences they might harbor.
We thank Katharine Hoyenga and anonymous reviewers for helpful suggestions on drafts of
1378 L. Ellis, C. Ficek / Personality and Individual Dierences 31 (2001) 1375±1379
Ellis, L., Robb, B. (2000). Sexual orientation in a large sample of United States and Canadian college students: toward
re®ned measurement (submitted for publication).
Hoyenga, K. B., & Wallace, B. (1979). Sex dierences in the perception of autokenetic movement of an afterimage.
Journal of General Psychology, 100, 93±101.
Lichtensteiger, W., & Schlumpf, M. (1985). Prenatal nicotine aects fetal testosterone and sexual dimorphism of sac-
charin preference. Pharmacology, Biochemistry, & Behavior, 23, 439±444.
Mather, J., Stare, C., & Breinin, S. (1971). Color preferences in a geriatric population. Gerontologist, 11, 311±313.
McGuinness, D., & Lewis, I. (1976). Sex dierences in visual persistence: experiments on the Ganzfeld and afterimages.
Perception, 5, 295±301.
Reid, S. N. M., & Juraska, J. M. (1992). Sex dierences in neuron numbers in the binocular area of the rat visual
cortex. Journal of Comparative Neurology, 321, 448±455.
Reid, S. N. M., & Juraska, J. M. (1995). Sex dierences in the number of synaptic junctions in the binocular area of the
rat visual cortex. Journal of Comparative Neurology, 352, 560±566.
Shapiro, B. H., & Goldman, A. S. (1973). Feminine saccharin preference in the genetically androgen insensitive male
rat pseudohermaphrodite. Horm Behavior, 4, 371±375.
Silver, N. C., & Ferrante, R. (1995). Sex dierences in color preferences among an elderly sample. Perceptual and
Motor Skills, 80, 920±922.
Silver, N. C., McCulley, W. L., Chambliss, L. N., Charles, C. M., Smith, A. A., Waddell, W. M., & Win®eld, E. B.
(1988). Sex and racial dierences in color and number preferences. Perceptual and Motor Skills, 66, 295±299.
Tate, F. B., & Allen, H. (1985). Color preferences and the aged individual: implications for art therapy. Arts in Psy-
chotherapy, 12, 165±169.
Valenstein, E. S., Kakolewski, J. W., & Cox, V. C. (1967). Sex dierences in taste preference for glucose and saccharin
solutions. Science, 156, 942±943.
Wijk, H., Berg, S., Sivik, L., & Steen, B. (1999). Color discrimination, color naming and color preferences in 80-year
olds. Aging: Clinical and Experimental Research, 11, 98±106.
L. Ellis, C. Ficek / Personality and Individual Dierences 31 (2001) 1375±1379 1379