Observations on facilitation and IOR in the Posner paradigm

In the Posner paradigm (Posner & Cohen, 1984), participants respond to a visual target preceded by a peripheral cue, either for a short or long stimulus onset asynchrony (SOA), and either at the same or a different location (Validity). Participants respond more quickly to targets when the target is cued by a stimulus at the same location for short SOAs (facilitation). This effect reverses for longer SOAs (Inhibition of Return, IOR). Previous research reported that the magnitude of facilitation and IOR was affected by physical characteristics of the cues (e.g., Lambert & Hockey, 1991; Pratt, Hillis, & Gold, 2001) and the targets (e.g., Reuter-Lorenz, Jha, & Rosenquist, 1996) in the COVAT paradigm. This thesis investigated the effect of physical characteristics of cues and targets and double cueing on facilitation and IOR. Exp. 1 aimed to examine the cue-target discriminability by manipulating the novelty of cues and targets. I found IOR but no facilitation. In addition, the cue-target discriminability did not affect the IOR, which was in line with the conclusion suggested by Pratt and colleagues (2001). In Exp. 2 I examined the effect of presenting onset and offset cues on facilitation and/ or IOR in different sequences of appearing and disappearing cues, labelled on-off and off-on cues. For on-off cueing, I found IOR but no facilitation. For off-on cueing, I found facilitation and no IOR for off-on cueing. This suggests that compared to on-off cues, off-on cues work more effectively in terms of capturing attention and that off-on cues can eliminate the otherwise robust IOR effect. In addition, for short SOA, the cue delayed the target detection for on-off cues but not for on-off cues. In Exp. 3A I examined the effects of the cue and target size on facilitation and IOR. IOR was found for both the small and large condition. Size did therefore not affect IOR for the long SOA. Facilitation was not found for either the small or large condition and instead, early IOR was found for the short SOA in the small size condition. In Exp. 3B I investigated the effect of changes in cue size on facilitation and IOR, using static cues similar to Exp. 3A as controls. I found no facilitation but observed IOR. Increased IOR was found for small cues compared to large cues in the static cue conditions. In Exp. 4A and Exp. 4B I used a double-cueing paradigm. By systematically varying SOA-D and the relative location between cue and target I studied the characteristics of facilitation as well as IOR. No matter whether single and double cueing was presented in intermixed or blocked trials, facilitation and IOR was found for single cueing, but not for double cueing. I speculate that the occurrence of facilitation for single cueing may be due to the temporal overlap of cues and targets between trials and blocks. The results for double cueing in Exp. 4A showed that IOR occurred at more than one location, that the most recently cued location produced the strongest IOR, and that IOR accumulated at the same location. When single and double cueing trials were presented in separate blocks, the second cue generated facilitation for short SOA between second cue and target and produced weaker IOR for long SOA between second cue and target. These results suggest that the second cue works effectively in terms of attracting attention, which is possibly due to the expectation of observers for target onset of the second cue. This thesis strongly supports the finding in previous research that in detection tasks facilitation is more difficult to occur whereas IOR is a robust effect (see, e.g., Collie et al., 2000; Mele et al., 2008; Tassinari et al., 1994; Tassinari & Berlucchi, 1995). This was observed when the experiment included only single and on-off cueing. Nevertheless, I found that the absence and presence of facilitation and IOR can be affected by the cue type and the number of cues. That is, for cue type, off-on cue triggered facilitation for the short SOA but eliminated IOR for the long SOA. For the number of cues, when only two cues preceded targets in a block, the second cue generated facilitation for short SOA between second cue and target and produced weaker IOR for long SOA between second cue and target, compared to when either one or two cues preceded targets in a block. The random effects of mixed models explained sufficient variance across subjects to improve estimates of facilitation and IOR and variability was mainly due to individual mean RTs (random intercepts) across conditions.