Memory for colours: a reaction time experiment

We used simultaneous and delayed match to sample tasks to investigate memory for 5 colour tests (green, yellow, purple, pink and orange) in men and women. Stimuli were emulated Munsell colour samples displayed on a CRT monitor. Colour tests were presented with distracters that could vary either in hue or in saturation. Our results indicate that: 1) over the five colours, women were more accurate than men in remembering colours (p=0.025). This advantage was significant for pink and purple colours and, in the later case, could be explained by the high women agreement in their categorical perception; 2) better memory for hue than for saturation, or distortion of the remembered colour towards more saturated samples depend on the colour of the test; 3) support for category effect in memory is provided by faster response times to the green colour test when presented with its cross-category hue distracter (p=0.025); 4) the best remembered colours were yellow, green and purple and the worst remembered colours were orange and pink. Introduction Several factors have been reported as influencing performances of colour memory. First, Pérez-Carpinell et al. (1998) [1] reported gender differences in a study using Munsell samples with retention interval varying from 15 s to 24 hours. In overall, women were more accurate than men in remembering colours for 15 s and 15 min delays, although, there was no significant effect when colour tests were considered individually. Second, compared to simultaneous colour matching, in delayed matching task, the comparison involves the mnemonic representation of the colour and discrimination performances are deteriorated. Different degree of impairment can be observed depending on the colour attribute of hue, saturation or lightness. For instance, Newhall et al., (1957) [2] reported an increase in saturation of remembered colours, and Ling & Hurlbert (2002) [3] found for three colour tests with different retention intervals a better memory for hue than for saturation. Finally, colour memory appears to be categorical. In a same/different task with a 10-second inter-stimuli interval, Boynton et al, (1989) [4] observed that the percentage of errors (omissions) increased as the categorical colour difference index between stimuli (OSA samples) was reduced, suggesting that categorisation occurs when colours must be remembered. In the present study, we investigate gender differences by testing men and women in simultaneous and delayed match-to-sample tasks, using the simultaneous matching task as a base-line to account for colour discriminability. To each colour test corresponds 4 distracters; two of them vary in hue (in two opposite directions from the test) while the other two vary in saturation (with an increase or a decrease from the test). This design will allow us to examine differences in performances when the colour test is presented with a hue or a saturation distracter. To determine the category membership of the hue distracters, a subset of subjects was also requested to perform a colour naming task allowing us to compare performances between crossand within-category hue distracters.