Lecture 5: Model Answer

From PSY243 course wiki
Jump to: navigation, search


Hagemann et al (2008) set out to test experimentally an observation made by Hill & Barton (2005) - that sporting contests are more often won by contestants wearing red than wearing blue (even though these colours are allocated at random). Hill & Barton (2005) ascribe this effect to the evolved dominance associations of the colour red influencing the contestant wearing red. Rowe, Harris & Roberts (2005) disagree, claiming that the effect is due to the difference in visibility of the opponent. Hagemann et al's claim is that the effect is in fact located within the referees, who are biased by the colours worn by the contestants.


Hagemann et al (2008) recruited 42 taekwondo referees, with an average of 8 years’ experience. These referees watched videos clips of actual taekwondo fights (average duration 4.4 seconds), and awarded points as they would if they were adjudicating. Half of the fight contestants wore red, half blue. Digitally altered versions of the clips were created with the colours reversed, each participant saw half of the clips with the original colours, and half with the reversed colours (which half was counterbalanced).


The points awarded varied systematically with contestant colour - such that contestants wearing red were awarded more points. There was no significant effect of whether clips were in the original colours or reversed. There was no effect of order of presentation, total points awarded (between original vs reversed) or gender of the referee.


This study introduces convincing evidence that the effect of colour may bias referees’ decisions in sporting contests.

There is good ecological validity - the participants were experienced referees, watching clips of real taekwondo bouts, scoring the same way they would in a real tournament. It could conceivably be a problem that the clips were watched on a laptop, which will – for very fast movements – introduce limitations of spatial and temporal resolution that are not there when watching a contest in real life.

The statistical analysis is sub-optimal - rather than a series of 3 t-tests, the authors should have conducted a single ANOVA with two factors: clip type (original vs reversed) and contestant colour (red vs blue). This would have kept the type I (false positive) error rate constant whilst maximising statistical power. As it is, the main effect of colour - upon which the interpretation relies - is highly significant so it is not likely that a better analysis would have changed the interpretation.

Unlike Hill & Barton this study uses an experimental method - rather than observation - so the locus of the effect can be identified. The contestant behaviour is identical - they are identical clips - but changing the colours affects the points awarded by the referees. However, although this shows that referees can be effected by colour, it doesn't show that contestants are not affected, nor does it shed light on why the referees are affected.

Future work could focus on clarifying two separate questions about the effect of colour.

First, what is the cognitive mechanism of the effect? Hill & Barton (2005) suggest evolved dominance associations, Rowe, Harris & Roberts (2005) suggest opponent visibility and Hagenmann et al (2008) do not specify the nature of the effect other than that it is "psychological" (p. 769). Experiments to disentangle the first two hypotheses could manipulate dominance in some way other than by colour associations (perhaps priming of contestants using sounds), or manipulate visibility whilst keeping colour the same (bright and less bright shades of red, for example), or whilst altering colour but maintaining visibility (so, for example, if visibility is generating the effect participants wearing gloves of different colours but equal discriminability would have the same advantage/disadvantage).

Second, there is an independent question of through which actor in the scenario the effect plays out. Contestants may be affected by the colours they wear, or the colours their opponent wears, or the referees may be affected by the colours of both. Roberts, Owen and Havlicek (2010), in follow up research show that both those being observed and observers can be influenced by worn colours. They did this by manipulating both the colours worn (taking multiple photographs) and the colours seen by raters (using digital alternation). Overall, the study demonstrates that referee bias due worn colours is a strong potential candidate for generating the effect show in Hill & Barton (2005), but it does not speak to biases which may manifest on the part of the contestants, and it does not clarify how any effect may be operating psychologically.


Hagemann, N., Strauss, B., & Leißing, J. (2008). When the referee sees red…. Psychological Science, 19(8), 769-771.

Hill, R. A., & Barton, R. A. (2005). Psychology: red enhances human performance in contests. Nature, 435(7040), 293-293.

Roberts, S. C., Owen, R. C., & Havlicek, J. (2010). Distinguishing between perceiver and wearer effects in clothing color-associated attributions. Evolutionary psychology: an international journal of evolutionary approaches to psychology and behavior, 8(3), 350-364.

Rowe, C., Harris, J. M., & Roberts, S. C. (2005). Sporting contests: seeing red? Putting sportswear in context. Nature, 437(7063), E10-E10.