When small is left and large is right : Behavioural evidence for attentional shifts due to irrelevant numerical cues

Numerous behavioural studies indicate the existence of a link between numerical representations and visuo-spatial processes (for review see DeHevia et al., 2008). A striking demonstration of this link was provided by Fischer and colleagues (2003), who reported that participants detect a target faster in the left hemifield, if preceded by a small number (e.g. 2 or 3) and faster in the right hemifield if preceded by a large number (e.g. 8 or 9). This is strong evidence that numbers orient visuo-spatial attention to the left or right hemifield, depending on their magnitude (e.g., small and large, respectively) (see also Galfano et al., 2006; Ristic et al., 2006).
We designed a modified version of this target detection paradigm, by replacing the detection task with a target discrimination task (cf. Hommel et al., 2001). The participants (n=16) were presented 1 task irrelevant digit (1,2 vs. 8,9) for 400ms. After a variable inter-stimulus interval (500, 1000 or 2000ms), they had to discriminate the colour of a brief (100ms) lateral target. We hypothesized that the centrally presented numbers would induce an orientation of attention, in the same direction as the initial observations by Fischer et al. (2003). The current results indicate a significant effect, but only for the shortest digit-target interval (500ms). We observed a significant interaction between number magnitude (small/large) and side of target presentation (left/right) (F1,15 =7.784, p<0.014). These findings indicate that the attentional shifts induced by irrelevant numerical material are independent of the exact nature of target processing (discrimination vs. detection).