Disruption of attention by irrelevant stimuli in serial recall

Abstract

In four experiments the behavioral consequences of an involuntary attentional distraction concerning memory performance was investigated. The working memory model of Cowan (1995) predicts a performance deficit for memory representations that are held in an active state when the focus of attention is distracted by a change in physical properties. In the first experiment, the distraction was realized by an irrelevant tone, which was repeatedly presented and exchanged by another tone unexpectedly. Further experiments explored an analogous effect of irrelevant visual–spatial stimuli. The experiments demonstrated a domain-specific distraction effect: an irrelevant tone change impaired performance in verbal but not in spatial serial recall, whereas a change of position of an irrelevant object disrupted spatial but not verbal memory performance. The results are discussed in the context of several memory models that offer accounts for the effect of irrelevant changing state sounds.

Introduction

Our thinking is often disrupted by noises in the environment, for instance car horns, bells, or people chatting on cell phones. There exists much research about the effect of irrelevant information on several cognitive tasks, in the verbal as well as in the visual–spatial domain. Different designs investigate possible effects of involuntary distraction by irrelevant auditory or visual information. Some emphasize memory impairment, e.g., irrelevant speech disrupting verbal serial recall (Colle & Welsh, 1976); others investigate the distraction of an attentional selection mechanism, e.g., the detection of a visual target surrounded by other visual items (Folk, Remington, & Johnston, 1992). Empirically it is difficult to distinguish between an attentional conflict and interference between contents or processes (e.g., see Cowan, 1995), complicating the development of theory regarding the possible interplay. The working memory model of Cowan, 1995, Cowan, 1999 integrates attention and memory and makes some strong predictions about the interplay of attentional distraction by irrelevant stimuli while focusing the relevant items.

In the model of Cowan, 1995, Cowan, 1999 memory representations are hierarchically organized in three faculties: first, there are elements in long-term memory that are not activated. Second, a subset of long-term memory representations is activated but the activation is susceptible to time-based decay. Third, a subset of the activated part of long-term memory is held in the focus of attention. Items in the focus of attention are activated continuously and are directly available in working memory. The focus of attention is voluntarily controlled by a central executive but can be involuntarily attracted by specific stimuli. Stimuli of personal relevance and changes in physical properties tend to capture the focus of attention. The following prediction can be derived from the model: if the focus is drawn away from the items to be remembered, the representations of memory items will drop out of the focus of attention and lose their activation benefit, leading to memory impairment. One of the main motivations of the following experiments was to test this hypothesis. This was realized by combining two experimental paradigms: a classically serial recall task and the attentional distraction by a sudden change of physical properties of an irrelevant stimulus. It was hypothesized that the sudden change would distract the focus of attention resulting in memory impairment.

Behaviorally the recruitment of attention by irrelevant items could be observed in the visual as well as in the auditory domain. In the context of visual search tasks irrelevant visual deviants lead, e.g., to an increase of reaction time (Theeuwes, 1994), or to an eye movement towards the singleton (Theeuwes, Kramer, Hahn, & Irwin, 1998). There is still a debate whether the distraction by a singleton (especially an abrupt visual onset) is rather automatic and stimulus driven (e.g., Theeuwes, 1994, Yantis and Jonides, 1990), or is contingent on the attentional set and the task demands (e.g., Atchley, Kramer, & Hillstrom, 2000).

The occasionally presentation of an irrelevant auditory deviant or novel tone in a series of irrelevant repeated identical tones results in a slowing of reaction time or decrease of accuracy in a visually presented forced-choice task (e.g., Escera, Alho, Winkler, & Näätänen, 1998). More analogous to visual search, the irrelevant deviant in a tone sequence makes the search for a target in this sequence more difficult (Dalton & Lavie, 2004). Event-related potentials (for a summary, see Näätänen, 1992) have revealed that change detection is associated with a mismatch negativity (MMN) as early as 150 ms after the deviant. The MMN is interpreted as a pre-attentive process (Näätänen, 1992, Schröger, 1997). The deviant detection can be followed by the orienting of attention associated with the ERP component P3a (Escera et al., 1998). The reorientation of attention towards the task-relevant information might be reflected by a late negativity, called RON (Schröger & Wolff, 1998). In the context of neuropsychological research this detection system was already joined with the working memory model of Cowan (1995), specifying the mechanism which elicits the involuntary attraction of the focus by a sudden auditory deviant (Berti, 2002).