Fig. 1: Experimental paradigm, Rapid Invisible Frequency Tagging (RIFT), and search performance.

a Trials were presented in a blocked design. Each block contained 20 target absent and 20 target present trials. Set sizes (16 or 32) were the same within each block. At the start of a block in the guided search condition, a “T” was presented in yellow or cyan, revealing the target colour for the following 40 trials. A block in the unguided search condition began with the presentation of a white “T”, and the target colour was randomized over trials. Note that the search displays are not true to scale; the eccentricity of the search array amounted to 10° visual angle, 5° on either side of the fixation dot. b RIFT at 60 and 67 Hz was applied to the colour of the stimuli by modulating the luminance sinusoidally. In this example, yellow stimuli were tagged at 67 Hz and cyan stimuli were tagged at 60 Hz. c Search performance decreases for more difficult searches. A hierarchical regression approach reveals a significant main effect for set size (β = 0.180) and guided/unguided (β = −0.138). Indicating that larger set sizes are associated with slower responses, while guided searches are faster than unguided searches. Pairwise comparisons did not reveal any significant difference in reaction time between unguided search set size 16 and guided search set size 32 (V = 134, z = 2.24, r = 0.4, p = 0.148), suggesting that participants focused their search on task-relevant items in the guided search condition. d Analogously, for accuracy (as measured by d’), hierarchical regression reveals a significant main effect for set size (β = −0.74) and guided/unguided (β = 0.56), indicating that accuracy is higher in guided searches and for set size 16 compared to 32. Again, there is no significant difference in sensitivity for unguided search set size 16 and guided search set size 32 (t(30) = 2.2, d = 0.2, p = 0.23).