- Мозги обезьян синхронизировались при наблюдении друг за другом [2018-04-02]
- Ученые впервые описали детоубийство у косаток [2018-04-02]
- Самцы белок убили чужих детенышей перед большим урожаем [2018-04-02]
- Что брезгливость к поту говорит о ваших взглядах [2018-03-15]
- Два вида шимпанзе говорят на одном языке жестов [2018-03-15]
- Серые дельфины спланировали заплыв за едой [2018-03-15]
- Почему хозяева похожи на своих собак? [2018-03-15]
- Когда пригорает: «социологизаторы» против «биологизаторов» [2016-06-13]
- Открывашка / Что такое этология и антропология? [2018-03-15]
- Введение в поведение. Этология человека [2017-04-11]
In the 2004 Olympic Games, opponents in four combat sports (boxing, taekwondo, Greco-Roman wrestling and freestyle wrestling) were randomly assigned red or blue athletic uniforms (complete outfits or body protectors). It was found that for all four competitions there is a consistent and statistically significant pattern in which contestants wearing red were more likely to win (Hill & Barton, 2005a). Two alternative hypothesizes were proposed to explain this color-associated winning bias: (1) red color had a special psychological effect on human behavior in competitions (Hill & Barton, 2005a); and (2) there is nothing psychologically special about the color red. Instead, red contestants were favored by perceptual differences, such as visibility against the background (Rowe, Harris, & Roberts, 2005). In a subsequent analysis of the Olympic data, it was found that the red-associated winning bias was only apparent in men (Hill & Barton, 2005b). Such a gender-dependent effect suggested that sexual selection may have influenced the evolution of human response to red color in competitions, thus arguing for the behavioral hypothesis (Hill & Barton, 2005a, Hill & Barton, 2005b).
Color of the athletic uniform can affect the behavior of the competitors (Frank & Gilovich, 1988, Mills & French, 1996) and artificial signals are able to trigger innate behavioral responses (Pryke, Andersson, Lawes, & Piper, 2002). In men, ruddiness of complexion seems to reflect testosterone levels (Edwards & Duntley, 1939, Edwards et al., 1941, Frost, 2005). Furthermore, anger is associated with a reddening of the skin due to increased blood flow, whereas fear is associated with increased pallor in similarly threatening situations (Drummond, 1997, Drummond & Quah, 2001). Hence, it was reasonable to suspect that redness of the athletic uniform enforced the evolutionary and/or cultural psychological associations of red with aggression and dominance (Hill & Barton, 2005a, Hill & Barton, 2005b). It remains unclear to which extent such a psychological effect of red during combat competition may have acted on the wearer and/or the opponent.
We hypothesized that “seeing red” may act as distractor, particularly for men in competition. The color–word Stroop test (Stroop, 1935) is a simple and reliable tool for investigation of human selective attention. During a Stroop test, the subjects are presented with a series of colored word stimuli and are asked to name the color in which the words are written while disregarding the actual meaning of the words. Stroop interference (SI) is quantified as the prolongation in response times (RTs) to incongruent stimuli (e.g., “BLUE” in red ink) as compared to neutral (meaningless word of any color) or congruent (e.g., “RED” in red ink) stimuli. Thus, the Stroop test could serve as a direct method to distinguish the effect of colors on perception (RT) from the effect on attention (SI).
Commonly used variants of the Stroop test (MacLeod, 1991) are based on block designs (subjects are timed after reading lists of words) and the color effects are not measurable. Investigation of the differences in SI for different colors is possible in computerized single-trial designs (every color naming is timed). Nevertheless, computerized tasks are usually more complex than the block-reading tasks and yield reduced amounts of interference (Salo, Henik, & Robertson, 2001) that do not favor the distinction of color effects. When such effects were investigated, complex interactions between ink color and interference were reported (Izawa & Silver, 1988, Laeng et al., 2005), although these studies used a large number of colors and did not differentiate the particularities of red color.
The aim of this study was to investigate in a group of medical students the distractor effect of red color naming during a Stroop task. To limit the number of colors used and account for both precortical (McKeefry, Parry, & Murray, 2003) and cortical (Laeng et al., 2005) opponent color effects, we compared RT and SI during naming of equiluminant, red, green and blue colors. To challenge perception (McKeefry et al., 2003, Rowe et al., 2005), we further tested the effect of halving the luminosity of the colored stimuli against the black background. Since gender differences during a Stroop test were previously reported in both RT (Mekarski, Cutmore, & Suboski, 1996) and SI (Palmer & Folds-Bennett, 1998, Sarmany, 1977), a comparable number of men and women were included in the study. To further distinguish gender from personality effects, a psychometric evaluation was carried out using the SCL-90-R test (Derogatis, 1983).
The investigations were carried out on volunteer medical students (27 men and 23 women) taking part in a “Brain awareness week” workshop (http://www.edab.net) at “Carol Davila” University of Medicine and Pharmacy in Bucharest. The age range varied narrowly between 19 and 26 years (Uttl & Graf, 1997). All participants reported good health, no major biological life events during the last year, no color blindness or uncorrected visual deficits, and no unusual stressful events or sleep deprivation during the week prior to the experiments (MacLeod, 1991). All subjects reported to be right-handed (Simon, Paullin, Overmyer, & Berbaum, 1985).
The experiments were carried out with the approval of the committee for experimental research of Carol Davila University of Medicine and Pharmacy (Bucharest, Romania) in accordance with international regulations regarding human experimentation.
2.2. Computerized color–word Stroop test
Subjects were comfortably seated in a quiet room with dim light and instructed to name the color of the words appearing on a 15-in. LCD screen (8 ms, 250 cd/m2) placed ∼1 m away. Verbal responses were captured with a unidirectional microphone attached to headphones fully covering the ears to reduce background noise and enhance vocal feedback (Breslow, Grand, & Freedman, 1980). Stimulus presentation and recording were controlled by custom-made software developed in MATLAB (R14, MathWorks Inc.).
The Romanian language equivalents of red (RO-SU), green (VER-DE) and blue (AL-BAS-TRU) were used as color words and YYYYY as a meaningless neutral word (letter “Y” is not used in the Romanian language). The words were presented with a large font (Arial bold, 100 points), centered on a dark background. During the test, a total of 54 word stimuli (of congruent, incongruent and neutral types) were presented. The number of word/color combinations was chosen so that the number of stimuli per type (N=18) would be equal to the number of stimuli of each color (N=18), thus allowing a balanced design per type per color.
2.3. Stroop tasks
Prior to the test, students attended a lecture on the psychophysiology of SI where it was deliberately suggested that responses to the Stroop test may be related to “intelligence”. To further induce a competitive situation, volunteers for the Stroop experiments were told that their performance would be ranked by sex (Palmer & Folds-Bennett, 1998). The subjects were, however, instructed to focus on accuracy rather than speed to reduce the rate of errors and facilitation (decreased RT in congruent stimulus conditions) (Chen & Johnson, 1991).
To ensure easy reproducibility of our colors, we chose the “pure” red (#FF0000), green (#00FF00) and blue (#0000FF) colors in the red-green-blue (RGB) computer color model. Translated to the hue–saturation–luminosity color model, these colors had the hue of 0°, 120° and 240°, respectively, full saturation and the same luminosity (Foley, van Dam, Feiner, & Hughes, 1990).
To challenge the perception of the colored stimuli (McKeefry et al., 2003), the subjects took part in a second Stroop test where the luminosity was halved for all colors, thus halving the contrast against the black background (the corresponding RGB colors were changed to #800000, #008000 and #000080, respectively). The dim color test was performed within 15 min after the full color test.
In the Stroop test, priming is an unwanted but unavoidable effect that reflects altered color naming by prior word and/or color activation (McClain, 1983). Priming decreases SI especially in single trial designs (Salo et al., 2001). To minimize “priming,” the inter-stimulus interval was set to 3 s (just enough to allow full articulation) (Henik, Friedrich, Tzelgov, & Tramer, 1994). Nevertheless, our preliminary reports using this software (Sandulache et al., 2004) indicated that even the 3-s interval may induce a small degree of priming. To further reduce the priming effect, we randomized the order of stimulus presentation for each subject.
2.5. Quantification of SI
Verbal responses triggered by each stimulus were digitized at 44 kHz for off-line analysis. The RTs were manually measured (Fig. 1). Stroop interference was quantified as the prolongation in RT between incongruent stimuli (e.g., word “BLUE” in red ink) as compared to neutral stimuli (“YYYYY” word of any color). Only valid responses (the correct color was named) were considered.
Large variation in red-green color vision exists among non-color-deficient subjects (Deeb, 2005). To reduce this source of variability, responses to incongruent, congruent and neutral stimuli were averaged for red, green and blue colors for each test. Hence, the RT (mean responses to incongruent stimuli) and SI (the difference between corresponding incongruent and neutral color means) could be calculated per color per subject. All measurements were stored in a SQL database for statistical analysis.
After the Stroop test, the subjects were asked to fill out a Romanian translation of the Symptom Checklist 90 R (SCL-90-R) self-administered personality questionnaires (Derogatis, 1983, Derogatis et al., 1973). Each of the items was rated on a five-point scale of distress (0–4). The nine primary symptom dimensions were labeled as somatization, obsessive-compulsive, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation and psychotics. Raw scores were evaluated by dividing the sum of scores for a dimension by the number of items in the dimension.
Results in numbers are given as mean±S.E.M. Mean and regression comparisons were carried out using SPSS (version 14, SPSS Inc.). The statistical tests and level of significance are mentioned when used. If the test statistic is conventionally quoted with degrees of freedom, these are presented as a subscript to the test statistic.
3.1. Color naming task
All responses to neutral stimuli were valid and overall accuracy of responses was higher than 98%.
In the pooled material from all colors (N=150), the mean RT was 676±7 ms. A prolongation of RT when naming incongruent stimuli (positive SI) was observed in 43/50 subjects. The mean SI was 133±7 ms. Both RT (Fig. 2A) and SI (Fig. 2B) appeared remarkably similar between the different colors (ANOVA: n=50, Bonferroni post hoc comparisons).
3.2. Gender differences during Stroop task
Naming accuracy was similarly high in both men and women. Nevertheless, we found that gender induced important differences. Men had a mean RT ∼40 ms faster than women (independent samples t-test, equal variances by Levene's test: t148=2.9, p<.005). Furthermore, men experienced ∼25 ms more interference than women (independent samples t-test, equal variances by Levene's test, t148=2, p<.05). In fact, six of seven subjects who did not show a positive SI were women.
We tested the interactions between color and gender in an ANOVA model (Type III sum of squares). In agreement with the general effects, we found gender effects as a single factor for both RT (F1=7.9, p<.005) and SI (F1=4.3, p<.05). Furthermore, we found a strong interaction between gender and color effects for SI (F2=7.4, p<.001) but not for RT.
To further elucidate the interactions between gender and color, we distinguished the gender effect for each color (Fig. 3). We found that the gender differences in SI were confined to red color naming (Fig. 3B, independent samples t-test, equal variances by Levene's test, t48=4.5, p<.001). Since 4 of 23 women did not show interference during red color naming, we repeated the comparison only with the subjects who showed positive SI. We found that the gender difference in SI during red color naming was maintained (independent samples t-test, equal variances by Levene's test: t44=3.5, p<.001).
3.3. Distractor effect of red during Stroop task
The SI was similar when naming green and blue colors (Fig. 3B). To increase the statistical power, we compared the SI during red color naming with the pooled SI during green and blue color naming. In men, we found that during red color naming the SI for red was ∼50 ms increased, as compared to other colors (independent samples t-test, equal variances by Levene's test: t79=2.8, p<.01).
Quite surprisingly, the distractor effect of red color was not apparent in women. In fact, for red SI was ∼55 ms less in women (independent samples t-test, equal variances by Levene's test: t67=2.6, p<.01). Thus, women seemed to experience less interference during red color naming.
3.4. Effect of color luminosity
Half of the subjects (15/27 men and 10/23 women) were subjected to a second color naming task where the luminosity of the stimuli was halved for all colors (the hue and color saturation were preserved). The decreased contrast with the black background made this task more challenging. The pooled RT for dim color naming was ∼40 ms increased (independent samples t-test, equal variances by Levene's test: t223=−3.2, p<.001). The increase in RT was similar for all colors (Fig. 2A).
Reducing color luminosity had no effect on SI (Fig. 2B). Similar to the full-color task, men experienced ∼75 ms longer SI during dim red color naming than did women (Fig. 3D, independent samples t-test, equal variances by Levene's test: t23=2.7, p=.013). Furthermore, in men, SI for red was ∼50 ms longer than for green and blue colors (independent samples t-test, equal variances by Levene's test: t43=2.7, p<.01).
Similar to the bright color task, women appeared to have longer RT than men (Fig. 3C) and shorter SI for red than for other colors (Fig. 3D), but these differences did not reach the level of statistical significance.
3.5. Divergent effect of red color in men and women
The distractor effect of red color observed during Stroop task appeared to diverge between men and women. It was possible that gender-dependent psychological differences, contributed to this divergence.
In our study, SCL-90-R questionnaires were returned by 22/27 men and 16/23 women. The women scored almost double the men's score (independent samples t-test, equal variances by Levene's test) on depression (t36=3, p<.005), interpersonal sensitivity (t36=2.7, p<.01) and phobic anxiety (t36=2.2, p<.05) SCL-90-R symptom subscales (Fig. 4A). These subscales seemed to co-vary since a strong correlation was found between depression and either interpersonal sensitivity (Pearson: r=0.8, n=38, p<.01) or phobic anxiety (Pearson: r=0.7, n=38, p<.01) scores.
To further distinguish gender from personality effects, we tested the relationship between SI during bright red color naming and SCL-90-R symptom subscales. In men, we found a strong positive correlation between depression SCL-90-R symptom subscale score and SI for red (Fig. 4B, Pearson r=0.8, n=22, p<.01). In women, a negative correlation was observed between depression SCL-90-R symptom subscale score and SI for red (Pearson: r=−0.3, n=22) but the effect did not reach the level of statistical significance (Fig. 4B).
3.6. Effects of menstrual cycle on variability
During both Stroop tasks and psychometric evaluation, SI variability appeared larger in women. Previous studies have indicated that a major cause of variability of SI in women is induced by fluctuations of the monthly cycle (Lord & Taylor, 1991). Women's perception of male facial color has been reported to fall into two phases according to the menstrual cycle: an estrogen-dominant phase (first 2/3 of cycle) and a progesterone-dominant phase (last 1/3) (Frost, 1994). It was therefore possible that phases of the menstrual cycle contributed to the heterogeneity of responses in women. We found that SI for red was 55±29 ms in the estrogen-dominant phase (10/23 women) and 103±23 ms in the progesterone-dominant phase (13/23 women). The ∼50 shorter SI in the estrogen-dominant phase may indicate a predominantly “estrogenic” contribution to the divergent effects of red in women, although the differences remained below the level of statistical significance, most likely due to the small sample size.
We compared color-naming, word-reading interference for red, green and blue in a group of medical students subjected to a computerized Stroop test. We found that, in men, naming red produced more SI than naming other colors. Furthermore, in men SI for red positively correlated to scores on the depression subscale of SCL-90-R questionnaire. Our data suggest that “seeing red” distracts men through a psychological rather than perceptual mechanism. Such a mechanism would associate red with aggression or dominance and may have a long evolutionary history, as indicated by behavioral evidence from nonhuman primates and other species.
4.1. Particularities of red color naming during Stroop task
We report here a strong gender and color interaction on a Stroop test given to a group of medical students. Clinically applicable Stroop tests focus on the overall interference effects and do not distinguish color-specific effects (MacLeod, 1991). Previous studies using different experimental paradigms found no difference in SI between colors (Izawa & Silver, 1988) or genders (Daniel, Pelotte, & Lewis, 2000). Nevertheless, the double interaction between gender and color was not addressed. Since our study indicates that gender differences in SI may be confined to red color and men and women seem to experience opposing effects, it is likely that the gender differences may not be observed in the pooled data, as also indicated in our study (Fig. 2B).
In agreement with an extensive study on the effect of color and luminosity on RT (McKeefry et al., 2003), we found that reducing the colored stimulus luminosity (thus decreasing the contrast with the black background) increased RT during the Stroop task. Although the dim color test was carried out after the full color test, the luminosity effect overpowered any confounding effects of experience (Jorgenson, Davis, & Opella, 1983). It has also been reported that RT should be longer for blue color as determined by cone-opponency mechanisms (McKeefry et al., 2003). We did not find any significant differences between RT to different colors in either luminosity conditions, although naming the blue color did appear to take longer than naming either red or green colors (Fig. 2, Fig. 3). Blue stimulus hue had the lowest contrast with the black background (Foley et al., 1990), and blue word (AL-BAS-TRU in Romanian) was also the longest word (Griffin, 2003). Such subtle differences in RT to color naming did not appear to converge into a significant effect in our Stroop task. We did find that men had shorter RT than women, probably due to the fact that men were more willing to sacrifice accuracy for speed (von Kluge, 1992) due to increased impulsivity (Dickman & Meyer, 1988).
Stroop interference occurs at the “output stage” of information processing (Atkinson, Drysdale, & Fulham, 2003) and it is unlikely that slowness in naming the red color in incongruent condition but not in neutral condition could be explained by differences in perception (Deeb, 2005). Previous studies have reported that nonopponent word–color color pairs (e.g., BLUE in red) induce more SI than opponent color pairs (e.g., GREEN in red) (Laeng et al., 2005). Such color–word interactions were unlikely to contribute to the increased red SI observed in our study since we used a balanced study design per color and found similar interference when naming green or blue colors. It is therefore reasonable to believe that increased SI during red color naming resulted from more complex information processing in men.
4.2. Red as an evolutionary-selected signal of male quality/dominance
Where individuals contest access to a resource, the requirement for mechanisms of conflict management has led to the selection of a variety of signals that act to reduce the frequency of life-threatening physical aggression. Although many colors are present in animal displays, it is specifically the presence and intensity of red coloration that has been found to signal male dominance in some species of fish (Boughman, 2001, Guderley & Couture, 2005), birds (Cuthill et al., 1997, Pryke et al., 2002) and primates (Setchell & Dixson, 2001, Waitt et al., 2003).
Several physiological factors may predispose red coloration to become a signal of male quality or male dominance (Frost, 2005): (1) the red pigment hemoglobin is used to transport oxygen in the blood (vertebrates and some invertebrates); (2) blood circulation responds more to male hormones than to female hormones (in species where metabolic demands on the male are greater than those on the female, particularly for short bursts of effort); and (3) cutaneous blood circulation is clearly visible on some portions of the body surface (e.g., anogenital areas of most primates). The signal could then become amplified by sexual selection, according to the intensity of competition among males for mates, i.e., the more intense the competition, the greater the reward for those males who emit the strongest and clearest signal. Male ruddiness would thus evolve into a specialized visual cue.
Among nonhuman primates, male mandrills (Mandrillus sphinx) possess a testosterone-dependent red coloration of the face, rump and genitalia (Setchell & Dixson, 2001). In a recent behavioral study of mandrills (Setchell & Jean Wickings, 2005), direct conflicts were significantly more frequent between similarly colored males, while clear submission of the pale males was more frequent where color differences were large. It was concluded that male mandrills may use relative brightness of red coloration to facilitate the assessment of individual differences in fighting ability (Setchell & Jean Wickings, 2005). Similarly, adult male rhesus macaques undergo a hormonally regulated increased reddening of facial and anogenital skin during their mating season (Rhodes et al., 1997). When female macaques were presented with computer-manipulated pale and red versions of male faces, they showed a clear preference for the red versions, suggesting that in macaques male coloration may also signal male quality in sexual selection (Waitt et al., 2003). Thus, at least in some primates, red coloration is a powerful signal of male quality regulating both intra-sexual (male-male competition) and inter-sexual (female choice) components of the sexual competition.
In humans, the adult male is ruddier in complexion than the adult female (Edwards & Duntley, 1939, Jablonski & Chaplin, 2000) and male hormones greatly increase blood circulation in the skin's outer layers (Edwards et al., 1941). Testosterone influences erythropoiesis during male puberty (Shahidi, 1973) and a decline of testosterone with aging increases the risk of anemia (Ferrucci et al., 2006). Furthermore, men with hypogonadism (Fonseca, Rajkumar, White, Tefferi, & Hoagland, 1998) or those taking anti-androgenic drugs (Ornstein, Beiser, & Andriole, 1999) frequently have anemia. These data are consistent with a testosterone-dependent ruddiness of the male complexion, as seen in many other species where red coloration acts as a signal of male dominance. Nevertheless, it remains unclear to what extent ruddiness has evolved as a behavioral signal in human competition.
4.3. Red as a behavioral signal in human competition
In the 2004 Olympic competition, the red-associated winning bias was apparent only in men, and this gender specificity was interpreted as an argument for the signaling effect of red color (Hill & Barton, 2005b). We enforced an intra-sexual and inter-sexual competitive situation and found that men experienced more SI for red, which was in agreement with a previous report where men experienced more global interference when they were told that performance will be “ranked by sex” (Palmer & Folds-Bennett, 1998). Furthermore, in men, SI appeared to positively correlate with scores on the SCL-90-R depression subscale. It was previously suggested that there is a positive relationship between depressive symptoms reflected in SCL-90-R and submissive behaviors including strategies for conflict de-escalation (Allan & Gilbert, 1997). It is therefore possible that “seeing red” had a greater distracting effect in men with depressive/submissive personality traits. Thus, it is reasonable to suspect that the gender specificity and personality dependency of the distracting effect of the red color may reflect a behavioral signal selected to regulate conflict management in humans.
An unexpected finding of our study was that women seemed to experience less interference when naming the red color than when naming the other colors (Fig. 3B,D). It was suggested that red color preference may differ with personality (Hafner & Corotto, 1980, Singg & Whiddon, 2000). We found that in our group of medical students, women scored higher than men for depression on the SCL-90-R questionnaire. We do not think that depression personality traits contributed to gender differences in red color SI because diverging relationships between depression score and red SI were found in men and women (Fig. 4B). Furthermore, we found that the SI for red color naming appeared to be even lower in the estrogen-dominant phase of the menstrual cycle (Frost, 1994). In a recent neurobiological study, it was found that SI correlates inversely with the levels of N-acetyl aspartate in the brain (Grachev, Kumar, Ramachandran, & Szeverenyi, 2001), which was found to depend on estrogens (Celik, Erdem, Hascalik, Karakas, & Tamser, 2005). Therefore, it seems more plausible to suspect that the divergent effect of gender on SI for red color naming reflected different aspects of sexual competition: in men, redness of the opponent signaled relative dominance (explaining increased interference of the opponent) while in women increased redness signaled increased attention (thus reducing interference).
To further investigate the apparent gender divergence in red signaling, we reanalyzed the publicly available data from the 2004 Olympic taekwondo competition where the strongest red wining bias was reported (Hill & Barton, 2005a, Hill & Barton, 2005b). To compensate for discrepancies in strength and skill, we included only the fights of at least quarter-final phase (n=94). We found that men wearing red athletic uniform won 66.7 % (χ2=5.3, p<.05) of matches, while women won only 43.5 %. Thus women did appear to perform less well when wearing red athletic uniform than when wearing blue athletic uniform but the effect was not significant. In our study, the effect of red color appeared to be more variable during the Stroop task and it failed to reach the level of statistical significance in the dim color experiments (Fig. 3C). Thus it is possible that the gender discrepancy was not statistically apparent in the Olympic data because the signaling effects were more variable in women.
Taken together with the behavioral evidence in several species including nonhuman primates, the striking similarity between the red winning bias in Olympic competition (Hill & Barton, 2005a, Hill & Barton, 2005b) and our report on SI suggests that red color may act as a distractor for men through a psychological effect that evolved in response to sexual competition. Since it is not known which wavelength of “red” may be the optimal signal to trigger such complex behavioral responses in humans, it is counterintuitive to suspect that artificial red color might affect SI in a simple computerized test. Nevertheless, while the efficacy of using RGB colors as animal signals has been debated (Stevens & Cuthill, 2005), the nonhuman primates may categorize the visible spectrum in a similar fashion with humans (Sandell, Gross, & Bornstein, 1979) and male redness was shown to influence sexual selection in primates (Waitt et al., 2003).
We acknowledge student Smaranda Dinescu for helping us test the reproducibility of data measurements. We would like to thank the editor and three anonymous reviewers for their very constructive comments. The project was supported by Carol Davila University of Medicine and Pharmacy Bucharest and the National Neuroscience Society of Romania.
Allan & Gilbert, 1997 1.Submissive behaviour and psychopathology. . British Journal of Clinical Psychology. 1997;36(Pt 4):467–488.
Atkinson et al., 2003 2.Event-related potentials to Stroop and reverse Stroop stimuli. . International Journal of Psychophysiology. 2003;47:1–21.
Boughman, 2001 3.Divergent sexual selection enhances reproductive isolation in sticklebacks. . Nature. 2001;411:944–948.
Breslow et al., 1980 4.Effect of vocal feedback on Stroop Color–word interference. . Perceptual and Motor Skills. 1980;50:447–451.
Celik et al., 2005 5.Magnetic resonance spectroscopic comparison of the effects of resveratrol (3,4′,5-trihydroxy stilbene) to conjugated equine estrogen, tibolone and raloxifene on ovariectomized rat brains. . European Journal of Obstetrics, Gynecology, and Reproductive Biology. 2005;120:73–79.
Chen & Johnson, 1991 6.The Stroop congruency effect is more observable under a speed strategy than an accuracy strategy. . Perceptual and Motor Skills. 1991;73:67–76.
Cuthill et al., 1997 7.Colour bands, dominance, and body mass regulation in male zebra finches (Taeniopygia guttata). . Proceedings of the Royal Society of London. 1997;264:1093–1099.
Daniel et al., 2000 8.Lack of sex differences on the Stroop color–word test across three age groups. . Perceptual and Motor Skills. 2000;90:483–484.
Deeb, 2005 9.The molecular basis of variation in human color vision. . Clinical Genetics. 2005;67:369–377.
Derogatis, 1983 10.SCL-90-R: Administration, scoring & procedures manual-II for the revised version. . Towson, MD: Clinical Psychometric Research; 1983;.
Derogatis et al., 1973 11.SCL-90: An outpatient psychiatric rating scale — preliminary report. . Psychopharmacology Bulletin. 1973;9:13–28.
Dickman & Meyer, 1988 12.Impulsivity and speed–accuracy tradeoffs in information processing. . Journal of Personality and Social Psychology. 1988;54:274–290.
Drummond, 1997 13.Correlates of facial flushing and pallor in anger-provoking situations. . Personality and Individual Differences. 1997;23:575–582.
Drummond & Quah, 2001 14.The effect of expressing anger on cardiovascular reactivity and facial blood flow in Chinese and Caucasians. . Psychophysiology. 2001;38:190–196.
Edwards & Duntley, 1939 15.The pigments and color of living human skin. . American Journal of Anatomy. 1939;65:1–33.
Edwards et al., 1941 16.Cutaneous vascular and pigmentary changes in castrate and eunuchoid men. . Endocrinology. 1941;28:119–128.
Ferrucci et al., 2006 17.Low testosterone levels and the risk of anemia in older men and women. . Arch Intern Med. 2006;166:1380–1388.
Foley et al., 1990 18.Computer graphics: Principles and practice. . 2nd ed.. Boston, MA: Addison-Wesley Longman Publishing Co. Inc; 1990;.
Fonseca et al., 1998 19.Anemia after orchiectomy. . Am J Hematol. 1998;59:230–233.
Frank & Gilovich, 1988 20.The dark side of self- and social perception: Black uniforms and aggression in professional sports. . Journal of Personality and Social Psychology. 1988;54:74–85.
Frost, 1994 21.Preference for darker faces in photographs at different phases of the menstrual-cycle — preliminary assessment of evidence for a hormonal relationship. . Perceptual and Motor Skills. 1994;79:507–514.
Frost, 2005 22.Fair women, dark men. The forgotten roots of color prejudice. . Christchurch, New Zealand: Cybereditions; 2005;.
Grachev et al., 2001 23.Grachev, I. D., Kumar, R., Ramachandran, T. S., & Szeverenyi, N. M. (2001). Cognitive interference is associated with neuronal marker N-acetyl aspartate in the anterior cingulate cortex: An in vivo (1)H-MRS study of the Stroop Color–word task. Molecular Psychiatry 6, 496, 529–496, 539.
Griffin, 2003 24.A reversed word length effect in coordinating the preparation and articulation of words in speaking. . Psychonomic Bulletin & Review. 2003;10:603–609.
Guderley & Couture, 2005 25.Stickleback fights: Why do winners win? Influence of metabolic and morphometric parameters. . Physiological and Biochemical Zoology. 2005;78:173–181.
Hafner & Corotto, 1980 26.Age, sex, race, and the Luscher color test. . Perceptual and Motor Skills. 1980;50:1144–1146.
Henik et al., 1994 27.Capacity demands of automatic processes in semantic priming. . Memory & Cognition. 1994;22:157–168.
Hill & Barton, 2005a 28.Psychology: Red enhances human performance in contests. . Nature. 2005;435:293.
Hill & Barton, 2005b 29.The Hill & Barton reply to “Rowe C et al., Nature 2006, 437:E10”. . Nature. 2005;437:E10–E11.
Izawa & Silver, 1988 30.Response variations to Stroop color–word stimuli. . Genetic, Social, and General Psychology Monographs. 1988;114:211–255.
Jablonski & Chaplin, 2000 31.The evolution of human skin coloration. . Journal of Human Evolution. 2000;39:57–106.
Jorgenson et al., 1983 32.Hemispheric asymmetry in the processing of Stroop stimuli: A balanced design. . International Journal of Neuroscience. 1983;21:171–174.
Laeng et al., 2005 33.Reduced Stroop interference for opponent colors may be due to input factors: Evidence from individual differences and a neural network simulation. . Journal of Experimental Psychology. Human Perception and Performance. 2005;31:438–452.
Lord & Taylor, 1991 34.Monthly fluctuation in task concentration in female college students. . Perceptual and Motor Skills. 1991;72:435–439.
MacLeod, 1991 35.Half a century of research on the Stroop effect: An integrative review. . Psychological Bulletin. 1991;109:163–203.
McClain, 1983 36.Color priming affects Stroop interference. . Perceptual and Motor Skills. 1983;56:643–651.
McKeefry et al., 2003 37.Simple reaction times in color space: The influence of chromaticity, contrast, and cone opponency. . Investigative Ophthalmology & Visual Science. 2003;44:2267–2276.
Mekarski et al., 1996 38.Gender differences during processing of the Stroop task. . Perceptual and Motor Skills. 1996;83:563–568.
Mills & French, 1996 39.Assertive behavior and jersey color in the national hockey and football leagues: The black uniform myth and reality. . Journal of Human Movement Studies. 1996;31:47–60.
Ornstein et al., 1999 40.Anaemia in men receiving combined finasteride and flutamide therapy for advanced prostate cancer. . BJU International. 1999;83:43–46.
Palmer & Folds-Bennett, 1998 41.Performance on two attention tasks as a function of sex and competition. . Perceptual and Motor Skills. 1998;86:363–370.
Pryke et al., 2002 42.Carotenoid status signaling in captive and wild red-collared widowbirds: Independent effects of badge size and color. . Behavioral Ecology. 2002;13:622–631.
Rhodes et al., 1997 43.Effects of administration of testosterone, dihydrotestosterone, oestrogen and fadrozole, an aromatase inhibitor, on sex skin colour in intact male rhesus macaques. . Journal of Reproduction and Fertility. 1997;111:51–57.
Rowe et al., 2005 44.Sporting contests: Seeing red? Putting sportswear in context. . Nature. 2005;437:E10–E11.
Salo et al., 2001 45.Interpreting Stroop interference: An analysis of differences between task versions. . Neuropsychology. 2001;15:462–471.
Sandell et al., 1979 46.Color categories in macaques. . Journal of Comparative and Physiological Psychology. 1979;93:626–635.
Sandulache et al., 2004 47.Priming and facilitation in a Stroop task. . International Journal of Psychophysiology. 2004;54:158.
Sarmany, 1977 48.Different performance in Stroops interference test from aspect of personality and sex. . Studia Psychologica. 1977;19:60–67.
Setchell & Dixson, 2001 49.Arrested development of secondary sexual adornments in subordinate adult male mandrills (Mandrillus sphinx). . American Journal of Physical Anthropology. 2001;115:245–252.
Setchell & Jean Wickings, 2005 50.Dominance, status signals and coloration in male mandrills (Mandrillus sphinx). . Ethology. 2005;111:25–50.
Shahidi, 1973 51.Androgens and erythropoiesis. . New England Journal of Medicine. 1973;289:72–80.
Simon et al., 1985 52.Reaction time to word meaning and ink color of laterally-presented Stroop stimuli: Effects of handedness and sex. . International Journal of Neuroscience. 1985;28:21–33.
Singg & Whiddon, 2000 53.Relationship between preference for red and locus of control. . Perceptual and Motor Skills. 2000;91:84–86.
Stevens & Cuthill, 2005 54.The unsuitability of HTML-based colour charts for estimating animal colours — a comment on Berggren and Merila (2004). . Frontiers of Zoology. 2005;2:14.
Stroop, 1935 55.Studies of interference in serial verbal reactions. . Journal of Experimental Psychology. 1935;18:643–662.
Uttl & Graf, 1997 56.Color–word Stroop test performance across the adult life span. . Journal of Clinical and Experimental Neuropsychology. 1997;19:405–420.
von Kluge, 1992 57.Trading accuracy for speed: Gender differences on a Stroop task under mild performance anxiety. . Perceptual and Motor Skills. 1992;75:651–657.
Waitt et al., 2003 58.Evidence from rhesus macaques suggests that male coloration plays a role in female primate mate choice. . Proc Biol Sci. 2003;270(Suppl 2):S144–S146.
a Center for Excellence in Neuroscience, Faculty of Medicine, “Carol Davila” University of Medicine and Pharmacy, Bucharest, Romania
b Anatomy and Physiology Department, Medical School, Laval University, Québec, Canada
c Division of Neuroscience and Pharmacology, Panum Institute, University of Copenhagen, Copenhagen, Denmark
© 2007 Elsevier Inc. All rights reserved.