B.1.4.2. Results

The number of values excluded averaged 4 % of trials per participant (including outliers and errors). First, a general ANOVA was performed on correct RTs, with cue type (form, letter), target size (small, medium and large) and relevance (random, large-relevant, small-relevant) as within-subject factors. The target size effect was significant (F(2,74) = 48.49, MSE: 979,803, p < .001), as well as the effect of relevance (F(2,74) = 32.10, MSE: 203,247, p < .001). On the other hand, the effect of the cue type was not significant (F(1,37) = 1.17, MSE: 50,238, p > .28).The target size X relevance interaction was the only significant interaction (F(4,148) = 66.08, MSE: 951,593, p < .001). All the other ones involved the cue type, and none of them approached significance, neither the cue-type X relevance interaction (F(2,74) = 0.43, MSE: 1,715, p > .65), nor the cue-type X size (F(2,74) = 0.37, MSE: 1,803, p > .69), nor the three-way interaction cue-type X relevance X size (F(4,148) = 0.90, MSE: 3,777, p > .46). This suggested that the perceptual priming, as manipulated here, had no major influence on the efficiency of cueing in the present paradigm.

To better understand the target size X relevance interaction, we computed planned comparisons with two-tailed Student t-tests, and adjusted the alpha threshold according to the Bonferroni-type procedure proposed by Larzerele and Mulaik (1977). In the random condition, RTs for the small target (1056 ms) were significantly longer than those for the medium target (1027 ms, t(37) = 2.53; t'_crit = 2.34; p < .008), which were longer than those for the large target (936 ms, t(37) =6.69; t'_crit = 2.85; p < .001). In the large-relevant condition, RTs for the small target (1183 ms) were significantly longer than those for the medium target (1144 ms, t(37) = 2.54; t'_crit = 2.51; p < .008), which were longer than those for the large target (860ms, t(37) = 11.08; t'_crit = 2.90; p < .001). In the small-relevant condition, RTs for the small target (990 ms) were significantly shorter than those for the medium target (1067 ms, t(37) = 4.57; t'_crit = 2.63; p < .001), which did not differ significantly from those for the large target (1097 ms, t(37) = 1.4; t'_crit = 2.03; p > .082). Finally, the effect of relevance was calculated by subtracting the mean RTs for the target at hand in the condition where it was relevant, to the mean RTs for this target in the random condition. This relevance effect was significant for large (75 ms, SE: 14; t(37) = 6.44; t'_crit = 2.79; p < .001) and small targets (66 ms, SE: 10; t(37) = 5.39; t'_crit = 2.72; p < .001).