![Sample our Behavioral Sciences journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/110801/TOC-Subject-Sample-2021-Behavioral-Sciences.jpg"})
ABSTRACT
Attentional effects are often inferred from keypress reaction time (RT) studies when two sequentially presented stimuli, appearing at the same location, generate costs or benefits. The universality of these attentional attributions is challenged by data from perceptual discrimination tasks, which reveal that location repetition benefits and costs depend on whether a prior response repeats or switches, respectively. According to dual-stage accounts, these post-attentional effects may be abolished by making responses in between two target stimuli or by increasing target location certainty, leaving only attentional effects. Here, we test these accounts by requiring responses to stimuli in between targets and by increasing target location certainty with 100% valid location cues. Contrary to expectations, there was no discernible effect of cueing on any repetition effects, although the intervening response diminished stimulus-response repetition effects while subtly reducing location-response repetition effects. Despite this, there was little unambiguous evidence of attentional effects independent of responding. Taken together, the results further highlight the robustness of location-response repetition effects in perceptual discrimination tasks, which challenge whether there are enduring attentional effects in this paradigm.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 We decided on a 1900–2100 ms response stimulus interval on trials without intervening events to roughly equate the TTOAs between trials with and without intervening response events, which obviously did not happen. We assumed, based on our prior work (Hilchey et al., Citation2017), that the M RT to the intervening response event would be roughly 300 ms. In those Experiments, there was always an intervening event. Here, where there was an intervening event on only half the trials, M RTs to the intervening event were almost double. This led to the CTOA being about 274 ms longer on trials with intervening events. This matter is discussed further in note 4.
2 An anonymous reviewer suggested that it would be interesting to examine whether Location Cueing altered inter-trial priming effects. That is, would the predictive cue alter the repetition effects from trial n − 1 to trial n, as measured by the RT to the first target stimulus on trial n? The inter-trial interval consisted of a 750 ms black screen. Trial sequences were excluded if errors were made on trial n − 1 or if the first target on trial n contained an error. Trial n − 1 and n could have been precued or not. As per usual, RTs > 2 s were first excluded and z-scores > 2.5 for the first target on trial n were excluded as outliers. To increase power, we pooled the central and peripheral target data and conducted a 4-way repeated measures ANOVA with the following factors: Location Repetition (repeat or switch) × Stimulus-Response Repetition (repeat or switch) × n − 1 Location Cueing (predictive or no cue) × n Location Cueing (predictive or no cue). N Location Cueing was significant, with faster RTs (∼24 ms) when the target location was predicted [F(1, 19) = 30.2, MSE = 1395, p < .001, h2 = 0.0301] but n Location Cueing did not interact with anything (all ps > .24). There was a main effect of Stimulus-Response Repetition [F(1, 19) = 12.81, MSE = 1277, p = .002, h2 = 0.0119], with generally faster (∼15 ms) RTs for stimulus-response repetitions than switches. The main effect of Location Repetition was marginal [F(1, 19) = 3.982, MSE = 508, p = .061, h2 = 0.0014], with generally faster RTs (∼5 ms) for location repetitions than alternations. As per usual, Location Repetition and Stimulus-Response Repetition interacted [F(1, 19) = 20.59, MSE = 671, p < .001, h2 = 0.0108], reflecting the location-response repetition effect. Nothing else was reliable (all ps > .10). Note that n Location Cueing does not reliably interact with these repetition effects even if we (1) ignore whether trial n − 1 contained a predictive cue or (2) analyse whether the first target on trial n appeared in central or peripheral vision separately.
3 “Switch” collapses across the levels “switch central” and “switch peripheral” because it is not possible for a second central target stimulus to mismatch the first target stimulus at the level of “switch central” but also because these two conditions typically yield similar RT effects. Note that all of the reported effects are significant even if “switch central” trials are removed from the analysis.
4 The interactions involving the Intervening Event Condition are obviously confounded by longer TTOAs when there are intervening events. Nevertheless, it seems reasonably clear that the intervening event, not the TTOA difference, is mainly responsible for the interactions. First, regarding the interaction between Intervening Event and Stimulus-Response repetition, Spadaro et al. (Citation2012; Experiment 1A) have already shown normal RT advantages for stimulus-response repetition without intervening events at TTOAs of 2–3 s and disadvantages for stimulus-response repetition with intervening events also at TTOAs of 2–3 s. Second, we re-analyzed the data in the “Follow-up analysis”, except we only included intervening event trials with 800 ms (not 1000 ms) intervals between the first target stimulus and intervening event and we only included non-intervening event trials with response stimulus intervals of 2100 ms. Thus, the TTOAs for trials with and without intervening events were on average within about one-tenth of a second of one other. Furthermore, we collapsed across Second Target Stimulus Location because this did not interact with the Intervening Event Condition. The interaction between Intervening Event and Stimulus-Response Repetition [F(1, 19) = 5.67, MSE = 529, p = .0279, h2 = 0.0047] and the interaction among Intervening Event, Stimulus-Response Repetition, and Location Repetition [F(1, 19) = 10.14, MSE = 292, p = .005, h2 = 0.0047] remained significant.