Seeing red

Red is one of those colours that anecdotally carries a degree of psychological significance. The problem is that the significance varies. Red has been linked with danger or intimidation (stop signs, traffic lights, red flag, red pens used for marking), romance, passion and sex (red roses, red lips, red-light districts, hearts, genital perfusion), salience (the word rubric means red chalk in Latin and referred to letters or sections in medieval psalmic manuscripts that were highlighted in red), or fear and horror (Grand Guignol, surgery, haemoglobin, the devil). And, of course, Santa Claus.

Carl Jung squirrelled his confrontations with the unconscious away in a red leather notebook – the famous Liber Novus, 'The Red Book' – the reasons for which have escaped science and, probably, Jung. (Another, more famous, Red Book was proffered to an unsuspecting worthy almost weekly via Thames Television in the 1970s and 80s.) In 1980, emotion theorist Robert Plutchick created a Munsell-type wheel to represent emotions by colour, assigning 'anger' and 'rage' to red. And George M. Stratton from the University of California published a paper entitled 'The colour red, and the anger of cattle' in which he sought to answer a perennially salient question: Is a red rag to a bull like a red rag to a bull? As Marc Abrahams reported in these pages in June 2015, Stratton concluded: 'Of 66 such persons who have favoured me with careful replies, I find that 38 believe that red never excites cattle to anger.' Three people were rigidly of the view that red does something to cattle. According to one of Stratton's respondents: 'I have found that waving anything before a bull is a dangerous business.'

Everywhere you look there are associations, from biblical associations with sin to the fate of the Enterprise crew in Star Trek (79 per cent of Enterprise crew who were killed-off wore red shirts). In stock markets, red is used in different countries to denote either the rise or fall of stocks. In Japan, a red kimono symbolises happiness and good luck.

This gallimaufry of different connotations is clearly incoherent. A stop sign is not consistent with an object designed to induce approach-behaviour such as seduction and romantic engagement, or even engagement online: one study from Yahoo labs and Georgia Tech published in PLoS One in 2015 found that images in red, purple and pink on Pinterest were significantly more likely to be shared than were those in green, black, blue or yellow.

It is this inconsistency and heterogeneity that has led to criticism of research in the field of colour psychology, particularly the part of the field that has sought to demonstrate a relationship between exposure to the colour red and behaviour change. The most controversial propositions suggest that the colour red has the ability to inhibit cognitive performance, enhance sexual attraction and enhance sporting prowess and judgements of that prowess. As I will demonstrate, however, red appears to create none of these effects consistently. In fact, there is evidence that other colours can induce similar behaviours – that cachet of red is not entirely exclusive.

The evidence for…

Red's ostensible but surprising importance to behaviour was highlighted in a famous paper by Russell Hill and Robert Barton from Durham University, in 2005. They examined the success of contact sport winners from the 2004 Olympic Games and found that contestants in the combat sports – boxing, tae kwon do, Greco-Roman wrestling and freestyle wrestling – were more likely to win if they wore a red than blue outfit. The same group, in 2008, found that players in football's Euro 2004 were also more likely to win if they wore red than other colours, and other researchers extended this advantage to Australian rugby players. And then Martin Attrill and colleagues' 2008 study of the English football league since 1947 found that when controlling for various biases (e.g. access to more resources), teams with red shirts won substantially more often than expected.

To address the argument that in sports where judges made assessments of winning, red may have made the contestants appear more salient, Norbert Hagemann and his team at Kassel University asked 42 tae kwon do referees to watch five different sparring contests in which the contestants' heads and trunks were covered in red or blue. Players wearing red were awarded 13 per cent more points than were players wearing other colours. But then, ingeniously, the researchers switched the colours using Photoshop and asked referees to make judgements of the same contestants. The contestants remained the same; the only variable that changed was the colour of the clothing. Those wearing red were awarded more points than those wearing blue, suggesting to the authors that these sports 'need a change to the rules (i.e. to forbid red sports attire) and support referees by providing electronic-decision making only'.

When the Burton and Hill group, in a study published in 2015, asked participants to rate how aggressive and dominant a series of digitally manipulated images of men were, men in red were judged by men and women to be more aggressive and angrier than were men dressed in blue or grey. Only men rated the men in red as more dominant. The effect extends to non-human stimuli: researchers from Sunderland and Northumbria Universities found that symbols coloured red were perceived as more dominant, eye-catching, 'riskier' and 'powerful' than were blue ones. Wearing a red tie, however, according to a study by Robin Kramer at the University of York, has no significant effect on ratings of the wearer's dominance or competence if the wearer is a familiar or unfamiliar politician.

People who used red chips in an online poker game were considered to be more intimidating than were players using blue or white chips, according to a 2012 study from the University of Amsterdam. The players also felt themselves to be more dominant and competitive. A red background in a betting game was associated with greater predicted losses than gains compared to one in green, according to researchers from the University of Haifa, and in one study by a team from Oxford and the US men (but not women) perceived greater savings when prices were presented in red than black in a series on online adverts. However, this effect disappeared when men were asked to pay attention to the products and evaluate them.

Sex differences also emerge in another field of research: sexual attraction. In one of the more dramatic demonstrations, in 2010 the University of Rochester's Andrew Elliott and his colleagues reported across seven experiments that women perceived men to be more attractive and sexually desirable when men wore red clothing or were standing in front of a red background. The effect, however, is likely to be attributable to the status of the men. When the effect was found it was found for men described as a professional worker earning a salary of $50,000, compared with a man described as a cleaner on a salary of $12,000. A more serious problem is that this study lacked sufficient power and that the study may be reporting a series of false positives. Some, such as Gregory Francis from Purdue University, have argued that 'either some experiments with null findings were not reported or reported experiments were run improperly' and that these findings are 'nonscientific or anecdotal'.

Other studies by the Elliot group found that women who were expecting to converse with an attractive man were more likely to select a red (rather than a green or blue) shirt to wear, and that women perceived other women who wore red to be more sexually receptive. In the last experiment, participants were asked to look at a photograph (of a woman) and asked to imagine that she was 'competing against you for the attentions of an attractive man'. Women were more likely to guard their partner from the women wearing red (compared to those wearing green). Women have been found to wear more 'reddish clothing' when at peak fertility, are more likely to wear red if they are sexually active and – in a study that features the first recorded use of the validity check 'using the Farmer's Almanac, we confirmed that these perceptions were accurate' – Jessica Tracy and Alec Bell from the University of British Columbia found that women wore red-coloured clothes significantly more often in cold weather than warm. (That the red clothing may have felt warmer because of the colour wasn't considered.) Men have been reported to sit closer to women wearing red t-shirts than blue, and prefer to offer lifts to hitch-hikers who wear red.

What is of potentially greater importance for psychology, and perhaps more important than the effect of the colour of a sporting jersey or a gambling chip, is the possibility that exposure to the colour red can influence cognitive performance. This is a consequence of the colour activating schema of threat or avoidance, according to Andrew Elliot and his colleagues. If red activates feelings of avoidance, people may take longer completing tests. Participants in their 2007 study performed less well on a test if they were exposed to the colour red beforehand. This effect was found even when participants did not express awareness of the colour and when the number was written in red ink at the top of a sheet of paper. Based on the proposition that the colour red signifies threat, 10- to 16-year-old girls in a 2011 study by Elliot and Aarts were asked to open a metal clasp as wide as they could on a sheet of white paper that also included their participant number in red, blue or grey. The girls used significantly more force when the clasp was next to a red number than numbers of another colour. To further highlight the link between red and inhibition/avoidance, Elliott and his team published a study in 2009 in which they found that participants moved their body away from an IQ test with a red cover more often than they did one coloured grey or green. The same authors also found that people presented with a test with a red or green cover and told that they would either be completing the test or rating its 'likeability', were less likely to knock on the door of an adjacent lab to complete the test if they were the ones told they would need to complete the test and the cover of this test was red.

The evidence against…

The red advantage for sport is not a robust phenomenon. For example, no association between wearing red and competition success has been found in a range of countries with competitors in rugby, football and boxing when corrections have been made for competitor ability. One study by Candy Rowe and her colleagues published in Nature in 2005 even reported a blue advantage in judo at the Athens 2004 Olympics, the same event as the one studied by Hill and Barton. Another study from 2008 by Peter Dijkstra and Paul Preenen from the Universities of Groningen and Amsterdam, examined 71 tournaments and controlled for possible confounds. It found no effect of blue.

Interestingly, the first experimental study of sporting performance and kit colour was published by Mark Frank and Thomas Gilovich from Cornell University in 1988 and found an effect of wearing black, not red. The effect was not a positive one. Their study of NHL and NFL players found that those kitted in black were more likely to be punished for aggression during play. However, as is common in this field, a study of Turkish football players, in 2005, failed to replicate this effect.

If there is a bias for red, it is likely to be with the observer, rather than the actor. That is, wearing red does not make a competitor a better competitor, but the colour influences the perception of the opponent and this opponent is perceived as being dominant. A competitor facing someone wearing red may be intimidated by the colour and this intimidation leads to poorer performance, blocking the motivation to achieve and activating a 'desire to avoid'. Yet this is inconsistent with those studies described earlier in which men wearing red were considered to be more attractive and desirable. The effect of red is fiercely context-specific.

The Attrill findings with English football league players were not replicated in Polish or German football players, in a study by Piotr Sorokowski and Andrej Szmajke. One hypothesis argues that the effect of red may be due to the disruption of the perception of a moving object so that an opponent makes errors of judgement about a red competitor's speed and distance. Supporting a salience model of red, one of Sorokowski's studies published in 2011 asked 225 participants to play nine arcade games in which they had to hit, avoid or outmanoeuvre blue, red or black objects. People hit the red objects better than they did black or blue, but there was no effect of colour on the ability to avoid objects. The study of game-playing with coloured symbols referred to earlier also found no effect of colour on actual task performance.

The greatest criticism, however, has been reserved for studies of cognition. The criticism concerns the ecological validity of the experiments. In a review of the field, Winifred Arthur and her colleagues from Texas A&M University and Universidad Adolfo Ibanez argued, in a 2016 issue of Psychonomic Bulletin and Review, that studies have produced inconsistent or non-significant findings because methodologies have been heterogeneous (information processing ability vs. knowledge-based exercises, familiar vs. unfamiliar content, a variety of colours used as comparators) and, crucially, differ in terms of the stakes involved. When colour produces significant effects, the authors argue, these are typically found in experiments where the stakes are low: materials may be unfamiliar, specific abilities are tested such as the ability to create anagrams, and the outcomes are inconsequential. Ravi Mehta, in a 2009 issue of Science, found that priming with red led to more productive anagram formation, but a replication in 2014 by Kenneth Steele from Appalachian State University, with three times the number of participants, failed. When stakes are high, the experiment is knowledge-based, the content of the exercise is familiar (course material) and where the outcomes matter (getting a grade), the red effect may disappear. One of these studies, published in 2014 by Adnan Smajic and colleagues from the University of Missouri, found that papers with a red cover (compared with green) had no significant effect on levels of anxiety expressed by students sitting examinations. Could it be that preparation, motivation and investment eliminate any effect of colour?

To test the effect of colour directly in a context where the outcomes were real, Arthur and colleagues accessed undergraduate outcome data to assess achievement when the cover of the examination booklet was printed in red or green; and any effect of the whole examination booklet printed in one colour. In both experiments, colour had no significant effect on the students' performance.

Issues

There are several issues that are relevant to this literature. One is that context is important. Red is not a universal signifier of an exclusive trait. It does not represent danger/threat only or romance only. It exerts its effects in a way that is context-dependent. In a romantic context, it is perceived as a colour encouraging approach. In a cognitive context, where the stakes are low, it is considered inhibitory and encourages avoidant behaviour. In a cognitive context where the stakes are high, its effects are negligible. There is no universal effect of red on multifarious behavioural dimensions

The second issue is that the ecological validity is questionable: studies have been limited to anagram formation, creation of analogies, assessment of photographs and so on in an artificial and constrained experimental environment that does not reflect the reality and heterogeneity of cognition in environments outside the laboratory. In addition, the trend has been for colour stimuli to be presented once, rather than several stimuli presented a number of times and participants' responses averaged for each colour stimulus.

The third issue is that few of the significant effects in the cognitive domain have been replicated by a number of independent groups; the successful replications have primarily been conducted by the same research groups or groups that share members. Given the reluctance in psychology to publish negative results, there may be more failures to replicate or produce significant results that go beyond the current literature.

Finally, and this is an issue that is barely addressed in these studies, there's the nature and composition of the colour itself. Red is not always red. It can vary according to hue, vividness, sharpness, lightness and other chromatic dimensions. It is unclear from the studies using red whether there is consistency in the kind of red stimulus used. Deviations from a particular chromatic dimension might exert effects on the results of an experiment because these dimensions are what contribute to these effects rather than the colour represented by the blanket label red.

What the literature suggests is that if there is an effect of the colour 'red' (with the 'red flag' caveat concerning what this phenomenologically means), this may be an epiphenomenon or an effect that can be accounted for by alternative explanations. This is especially true in those studies that have reported a beneficial effect of wearing red in sports contests. The inhibitory effect of red (or colour) in cognitive contexts can also be eliminated by enhancing ecological validity or increasing the power of the experiment. Where red may exert an effect on cognition it is in tasks that are low-stakes and where the outcomes are inconsequential – and this effect is potentially attributable to the distraction associated with the brightness or vividness of the colour, rather than specifically with the colour red and the schema it might activate.

More generally, the heterogeneity of contexts with which red has been associated is clearly problematic. How can red be equated with intimidation or withdrawal, when it is also considered to be a colour that attracts a potential partner? No account has yet fully explained this disparity.

We're left, therefore, with a perennially interesting field. But, as we plough and plant in this field, sighting green shoots in the development of reliable findings with external validity would probably require the application of rose- (if not red-) tinted glasses.

- Professor G. Neil Martin is Head of Psychology at Regent's University London
[email protected] Twitter: @thatneilmartin

Key sources

Allen, M.S. & Jones, M.V. (2014). The home advantage over the first 20 seasons of the English Premier League. International Journal of Sport and Exercise Psychology, 12(1), 10–18.
Arthur Jr, W., Cho, I. & Muñoz, G.J. (2016). Red vs. green: Does the exam booklet color matter in higher education summative evaluations? Not likely. Psychonomic Bulletin & Review, 23(5),1596–1601.
Beall, A.T. & Tracy, J.L. (2013). Women are more likely to wear red or pink at peak fertility. Psychological Science, 24, 1837–1841.
Elliot, A.J., Kayser, D.N., Greitemeyer, T. et al. (2010). Red, rank and romance in women viewing men. Journal of Experimental Psychology: General, 139, 399–417.
Elliot, A.J., Maier, M.A., Moller, A.C. et al. (2007). Color and psychological functioning: The effect of red on performance attainment. Journal of Experimental Psychology: General, 136(1), 154.
Francis, G. (2013). Publication bias in 'Red, rank, and romance in women viewing men,' by Elliot et al. (2010). Journal of Experimental Psychology: General, 142, 292–296.
Frank, M.G. & Gilovich, T. (1988). The dark side of self-and social perception: Black uniforms and aggression in professional sports. Journal of Personality and Social Psychology, 54(1), 74.
Hill, R.A. & Barton, R.A. (2005). Psychology: Red enhances human performance in contests. Nature, 435(7040), 293–293.
Prokop, P. & Hromada, M. (2013). Women use red in order to attract mates. Ethology, 119(7), 605–613.
Rowe, C., Harris, J.M. & Roberts, S.C. (2005). Sporting contests: Seeing red? Putting sportswear in context. Nature, 437(7063), E10–E11.
Smajic, A., Merritt, S, Banister, C. & Blinebury, A. (2014). The red effect, anxiety, and exam performance: A multidisciplinary examination. Teaching of Psychology, 41, 37–43.
Sorokowski, P. & Szmajke, A. (2011). The influence of the 'red win' effect in sports: A hypothesis of erroneous perception of opponents dressed in red – Preliminary test. Human Movement, 12(4), 367–373.
Steele, K.M. (2014). Failure to replicate the Mehta and Zhu (2009) color-priming effect on anagram solution times. Psychonomic Bulletin & Review, 21(3), 771–776.