How is location defined? Implications for learning and transfer of location-specific control.

Abstract

Much research has explored location-specific proportion compatibility (LSPC) effects (i.e., how the appearance of a stimulus in certain locations can reactively trigger different attentional control settings) to elucidate mechanisms underlying reactive control. Recently, however, failures to reproduce key evidence showing transfer of LSPC effects (originally reported in Crump & Milliken, 2009) have called into question whether control per se supports these effects. Notably, Crump and Milliken (2009), and all studies attempting to reproduce their findings, presented stimuli in two locations, one above and one below fixation. Inspired by research on differences between horizontal and vertical meridians, we examined the consequences of defining location in this way compared with alternatives. Experiments 1 and 2 demonstrated that LSPC effects are robust when location is defined as left versus right and larger than when location is defined as upper versus lower, and additionally demonstrated LSPC effects for two locations within the same coarse spatial category (e.g., left vs. farther left). In Experiment 3, we aimed to reproduce Crump and Milliken's key findings using left and right locations for the first time. Critically, we found transfer of the LSPC effect to diagnostic items across two designs and the first evidence for a robust experiment wide LSPC effect for inducer items. Our findings support theories positing that LSPC effects reflect location-specific attentional control and more generally suggest that choosing a definition of location is not a minor methodological decision but critically impacts learning and transfer of location-specific attentional control. Public Significance StatementSuccessfully accomplishing daily tasks requires focusing on relevant, while ignoring irrelevant, stimuli, by employing attentional control. Prior research has shown that attentional control can be allocated differentially across different areas of space depending on where it is most needed (i. e., locations where there are more often irrelevant stimuli that must be ignored). Here we examine how relationships between the locations needing differential control (e.g., are they in different vertical categorical locations-upper vs. lower, different horizontal categorical locations-left vs. right, or within the same category-both within left) influence the allocation of attentional control and demonstrate that some spatial layouts naturally lend to more efficient flexible attentional control. In other words, the degree to which humans can vary the amount of attentional control they exert to accomplish certain tasks is dependent on where stimuli for these tasks appear.

Revealing object-based cognitive control in a moving object paradigm.

“…See the online article for the color version of this figure. et al, 2006Diede & Bugg, 2016, 2017Dreisbach et al, 2018;Gonthier & Blaye, 2021;Kuratomi & Yoshizaki, 2013;Pickel et al, 2019;Surrey et al, 2017Surrey et al, , 2019Weidler & Bugg, 2016;Weidler et al, 2020Weidler et al, , 2022Wendt et al, 2008). Bearing in mind these prior findings, in all experiments in the current study, participants were asked to count the number of flanker stimuli that appeared in a particular object, thereby encouraging participants to attend to the objects (e.g., Brosowsky & Crump, 2021).…”

“…The initial goal of CSPC researchers was to establish that context-specific control exists independent of known confounds (e.g., Corballis & Gratton, 2003;Crump & Milliken, 2009;Crump et al, 2008). More recently, researchers have turned to questions regarding the reliability (reproducibility) of context-specific control (e.g., location-based CSPC effects; Bugg et al, 2020Bugg et al, , 2022Crump et al, 2017;Hutcheon & Spieler, 2017;Weidler et al, 2022). The current study contributes to both goals by demonstrating that (a) objects can serve as a contextual signal for cognitive control in a novel paradigm that controls for known confounds and (b) object-based cognitive control is reproducible.…”

“…Indeed, location, more than other features such as shape or color, is believed to be obligatorily attended and encoded into episodic memory (Logan, 1998; but see Woodman, 2021). This may explain why a counting instruction directing participants to attend to shape or color, when they serve as contextual cues, has been needed to produce shape and color-based CSPC effects (e.g., Brosowsky & Crump, 2021; Crump et al, 2008) whereas there are many demonstrations of location-based CSPC effects without directing participants to count based on location (e.g., Colvett & Bugg, 2022; Corballis & Gratton, 2003; Crump & Milliken, 2009; Crump et al, 2006, 2008; Diede & Bugg, 2016, 2017, 2019; Dreisbach et al, 2018; Gonthier & Blaye, 2021; King, Korb, & Egner, 2012; Kuratomi & Yoshizaki, 2013; Pickel et al, 2019; Surrey et al, 2017, 2019; Weidler & Bugg, 2016; Weidler et al, 2020, 2022; Wendt et al, 2008). Bearing in mind these prior findings, in all experiments in the current study, participants were asked to count the number of flanker stimuli that appeared in a particular object, thereby encouraging participants to attend to the objects (e.g., Brosowsky & Crump, 2021).…”

Processing stage flexibility of the SNARC effect: Task relevance or magnitude relevance?

“…In addition, the Bayes factors (BF 10 ) were calculated for theoretically meaningful null interactions ( Weidler et al, 2022 ). A BF 10 less than 1/3 provides substantial evidence for the null while a BF 10 larger than 3 provides substantial evidence for the alternative, and any BF 10 between 1/3 and 1 indicates that the evidence is too weak to reject or accept the null interaction ( Wagenmakers et al, 2011 ).…”

A spatial version of the Stroop task for examining proactive and reactive control independently from non-conflict processes

“…In the context-specific PC paradigms, the two sets being contrasted are defined by a task-irrelevant and non-interfering feature (e.g., the positioning of colored words above or below fixation, with the words’ position having virtually no impact on color-naming performance). Although this paradigm also tends to produce a PC effect (Crump et al, 2006 ), there is currently considerable controversy as to whether it really involves conflict-induced control (Bugg et al, 2020 ; Hutcheon, 2022 ; Hutcheon & Spieler, 2017 ; Schmidt & Lemercier, 2019 ; Weidler et al, 2022 ). For this reason, in the following, we mainly focus on list-wide and item-specific PC paradigms.…”