Processing time not modality dominates shift costs in the modality-shifting effect

Psychological Research https://doi.org/10.1007/s00426-019-01276-1 ORIGINAL ARTICLE Processing time not modality dominates shift costs in the modality‑shifting effect Hettie Roebuck1,2 · Kun Guo1 · Patrick Bourke1 Received: 26 November 2018 / Accepted: 2 December 2019 © The Author(s) 2019 Abstract Shifting attention between visual and auditory targets is associated with reaction time costs, known as the modality-shifting effect. The type of modality shifted from, e.g., auditory or visual is suggested to have an effect on the degree of cost. Studies report greater costs shifting from visual stimuli, yet notably used visual stimuli that are also identified slower than the auditory. It is not clear whether the cost is specific to modality effects, or with identification speed independent of modality. Here, to interpret whether the effects are due to modality or identification time, switch costs are instead compared with auditory stimuli that are identified slower than the visual (inverse of tested previously). A second condition used the same auditory stimuli at a low intensity, allowing comparison of semantically identical stimuli that are even slower to process. The current findings contradicted suggestions of a general difficulty in shifting from visual stimuli (as previously reported), and instead suggest that cost is reduced when targets are preceded by a more rapidly processed stimulus. ‘Modality-Shifting’ as it is often termed induces shifting costs, but the costs are not because of a change of modality per se, but because of a change in identification speed, where the degree of cost is dependent on the processing time of the surrounding stimuli. Abbreviations SART Sustained attention to response task RT Reaction time Introduction Having to switch attention between visual and auditory targets is associated with switching costs, known as the modality-shifting effect. When a previous target is of a different modality detecting a subsequent target is slower than when it is of the same modality (e.g., Zubin, 1975; Ferstl, Hanewinkel, and Krag, 1994; Spence, Nicholls, and Driver, 2001; Lukas, Phillip, and Koch, 2010). The frequency and recency of a given modality are thought to build up an * Hettie Roebuck Kun Guo Patrick Bourke 1 School of Psychology, University of Lincoln, Lincoln LN6 7TS, UK 2 University of Wisconsin, Madison 53706, USA underlying prediction of what the next target will be (Epstein & Rock, 1960). The Configuration–Execution model suggests that top–down factors (e.g., expectancy) affect the time to program central mental operations, and bottom–up factors (e.g., recency) independently affect the time to execute them (Ruthruff, Remington, & Johnston, 2001). Shifting costs represent the delay when a switch in modality conflicts with the underlying prediction generated from the set of previous stimuli. The size of the shift cost is not always the same. It is suggested that the type of modality that is being shifted from can have an effect on the degree of cost induced onto the subsequent target. For example, it has been suggested that there is either an underlying visual dominance, or a difficulty to shift away from visual stimuli compared to auditory stimuli (Posner et al., 1976). Visual stimuli are suggested to be not as attention capturing as auditory (e.g., Cohen & Rist, 1992); thus, visual dominance may result in a greater attentional tuning to the visual modality (Posner, Nissen, Klein, 1976). In the context of modality-shifting, a selective bias towards visual stimuli is thought to benefit responses to repeated trials, resulting in a greater cost if the modality switches (e.g., Miller & Cohen, 2001; Lukas et al., 2010). The distinction between the possibility of an underlying visual dominance or a difficulty to shift away from visual stimuli has been explored further comparing modality-shifting 13 Vol.:(0123456789) Psychological Research across multiple modalities (Spence, Nicholls, & Driver, 2001). Using tactile (finger stimulation), visual (light illumination) and auditory stimuli (pure tones), participants had to make a left/right discrimination for each modality. When the modality of discrimination shifted to another modality, there were even greater reaction time (RT) costs associated with shifting away from touch, compared to both visual and auditory stimuli. From these findings, they conclude that it is even more difficult to shift attention away from the tactile modality, rather than an underlying predisposition to attend to visual. Therefore, the authors interpret the findings observed by Posner et al. (1976) as a greater difficulty to shift away from visual stimuli, rather than an overall visual dominance, as tactile stimuli could induce even greater shifting delays compared to the visual stimuli. The debate over whether there are difficulties to shift away from a modality, or whether a modality dominates has also attracted interest in the interpretation of underlying difficulties associated with different clinical disorders. Both in terms of deficiencies, which may be specific to a modality (e.g., in Huntingtons Disease; Sprengelmeyer, Lange, & Hömberg, 1995), and sensory dominance to a modality (e.g., in Schizophrenia; Zahn, Pickart, & Haier, 1994). Despite the interest in modality specific effects, there has been little consideration to what extent switching costs are specifically inherent to modality verses a switch associated with the processing requirements of a differing stimulus. Studies suggest that task demands can influence the degree of shift cost, for example, if a task requires temporal processing or spatial processing, responses can be affected depending on which modality is given priority (Lukas, Philipp, & Koch, 2014). It is possible that differences between modality in attention shifting tasks may have less to do with whether they are auditory, visual or tactile and more to do with how quickly they can be processed. The impact of speed of stimulus identification has not previously been considered in the context of modality-shifting. However, research suggests that more caution is needed when attributing differences in performance to the modality of presentation more generally (Roebuck, Freigang, & Barry, 2016). The findings show that in a simple GO NOGO continuous performance task, different types of stimuli within a modality can induce different response rates and error propensities, suggesting that effects of stimulus type may be being misinterpreted as modality effects. Specifically, participants were presented with two stimuli, and had to press a button to one stimulus, and inhibit their responses to the other. Slower RT’s and greater errors were observed in the condition using two auditory tone stimuli (high and low tones), compared to the same task (and modality) using auditory spoken sounds (X and O). The stimuli were all presented for the same duration, but the linguistic stimuli may be easie (...truncated)