Selective auditory attention in 3- to 5-year-old children: An event-related potential study

https://doi.org/10.1016/j.neuropsychologia.2005.10.007Get rights and content

Abstract

Behavioral and electrophysiological evidence suggests that the development of selective attention extends over the first two decades of life. However, much of this research may underestimate the attention abilities of young children. By providing strong, redundant attention cues, we show that sustained endogenous selective attention has similar effects on ERP indices of auditory processing in adults and children as young as 3 years old. All participants were cued to selectively attend to one of two simultaneously presented stories that differed in location (left/right), voice (male/female), and content. The morphology of the ERP waveforms elicited by probes embedded in the stories was very different for adults, who showed a typical positive-negative-positive pattern in the 300 ms after probe onset, and children, who showed a single broad positivity during this epoch. However, for 3- to 5-year-olds, 6- to 8-year-olds, and adults, probes in the attended story elicited larger amplitude ERPs beginning around 100 ms after probe onset. This attentional modulation of exogenously driven components was longer in duration for the youngest children. In addition, attended linguistic probes elicited a larger negativity 300–500 ms for all groups, indicative of additional attentional processing. These data show that with adequate cues, even children as young as 3 years old can selectively attend to one auditory stream while ignoring another and that doing so alters auditory sensory processing at an early stage. Furthermore, they suggest that the neural mechanisms by which selective attention affects auditory processing are remarkably adult-like by this age.

Introduction

By labeling selective auditory attention “the cocktail party effect,” Cherry (1953) might have implied that it is something in which only adults fully engage. Indeed, evidence gathered over the past 50 years has indicated that the ability to differentially process relevant and irrelevant information is not fully adult-like until at least the teenage years (e.g., Doyle, 1973, Hiscock and Kinsbourne, 1980; Pearson & Lane, 1991). However, young children certainly have some ability to selectively attend (McKay, Halperin, Schwartz, & Sharma, 1994; Pritchard & Neumann, 2004; Rueda et al., 2004), and attention may play a role in the development of other cognitive skills. Designing tasks to accurately measure selective attention in young children renders it possible to gain a better understanding of how attention mechanisms are established during development and how the maturity of attention systems correlates with the growth of other cognitive and perceptual abilities.

Evidence from behavioral studies of visual and auditory selective attention using traditional tasks suggests that adult-like attentional control is not fully developed until at least after puberty. In dichotic listening tasks, the abilities to correctly recall and recognize information presented to the attended ear while avoiding intrusions from the unattended ear continue to develop beyond the age of 12 years (Asbjørnsen & Bryden, 1998; Doyle, 1973; Hiscock & Kinsbourne, 1980; Pearson & Lane, 1991). In the visual modality as well, the ability to attend to specific features within stimuli, as measured by slowed reaction times for detecting targets when irrelevant features vary, continues to develop until at least the age of 10 years (Shepp, Barrett, & Kolbet, 1987). Finally, in a cross-modal Stroop task, 11-year-olds are worse than adults at suppressing auditorally presented color names when asked to identify the color of a visually presented rectangle (Hanauer & Brooks, 2003). Thus, from studies employing classic adult attention tasks across modalities, there is evidence for a prolonged period of development for selective attention systems.

Studies that have more directly manipulated the degree of executive control necessary to complete a task suggest that adult-like control of selective attention can be accomplished at an earlier age under conditions in which external cues direct attention more strongly (Klenberg, Korkman, & Lahti-Nuuttila, 2001; McKay et al., 1994; Shepp & Barrett, 1991; Zukier & Hagen, 1978). For example, in a visual target detection task with various set sizes, 9- and 10-year-old children were able to ignore distractors at peripheral locations to a similar extent as adults only if the complexity of the foveally presented set demanded more attention (Huang-Pollock, Carr, & Nigg, 2002). Studies such as this suggest that with adequate external cues young children can overcome difficulty with sustaining attention over an extended period of time and switching among competing tasks. Overall, the ability to ignore irrelevant information across a variety of more child-directed tasks appears fully developed by the age of 11 (see Lane & Pearson, 1982 for a review).

Studies designed specifically for young children have shown that children can demonstrate adult-like competence in some aspects of attention at even earlier ages. For example, adapting a traditional Posner visual cuing paradigm to be more interesting and child-friendly by employing fish on colorful backgrounds yielded the finding that attention orienting is fully developed before the age of 6 years (Rueda et al., 2004). When attention is indexed by a more implicit measure such as negative priming (unattended items are more difficult to process on subsequent trials) rather than by a more explicit measure such as reaction time, children as young as 5 years old have been shown to perform similarly to adults (i.e., they demonstrate as much inhibition of irrelevant items as do adults; Pritchard & Neumann, 2004).

The results of these studies highlight the importance of designing appropriate tasks in developmental research; certainly questions about the maturation of selective attention systems are tractable—with sensitive measurement tools. Many of these developmental studies, particularly those that have employed more traditional tasks, may have inadvertently confounded or conflated at least two aspects of selective attention: the initial selection of the to-be-attended channel and the subsequent perception and further processing of stimuli within (and outside of) that channel. The results from the more child-accessible paradigms that provide support for selection suggest that some of the effects of attention on perception and further processing of stimuli may be remarkably mature in very young children. Reports of poor selective attention skills in older children may thus be a reflection of broad, inflexible, or poorly tuned selection criteria in children rather than a lack of differential processing for attended and unattended items after that selection has been made. There is some behavioral evidence to support this hypothesis: on visual attention tasks, 5- and 8-year-olds benefit more than adults from cues directing their attention to the correct items (Heinbuck & Hershberger, 1989).

Employing neurocognitive measures in addition to behavioral measures may help to clarify what sorts of attentional skills are developing when. There is a lengthy history of event-related potentials (ERPs) being used to determine the stages at which selective attention affects processing in adults. In a classic ERP study of selective attention, adult listeners were asked to attend to one ear to detect rare high-frequency tones in rapid series of standard and target tones presented to both ears (Hillyard, Hink, Schwent, & Picton, 1973). Standard tones presented to the attended ear elicited larger N100s (negativity between 80 and 120 ms) than the same tones when unattended. Targets presented to the attended ear elicited larger P300s (positivity between 250 and 400 ms) than unattended targets. The finding that some portion of the attention effect mirrors the distribution of underlying ERP components suggests that attention modulates exogenously driven neural activity. However, attention effects in other time windows can have distributions distinct from the sensory ERP responses (Hansen & Hillyard, 1980). These effects are considered to index additional endogenous processing of attended stimuli rather than a simple gating of sensory information like that indexed by the amplitude of the N100 (Hansen & Hillyard, 1980). Since the initial report by Hillyard et al. (1973), this classic dichotic listening paradigm has been adapted to test many specific hypotheses about the neural systems important in selective attention, the timing of selective attention, and the impact of selective attention on the processing of syllables (Hink, Hillyard, & Benson, 1978), linguistic stimuli that vary by different features (Hansen, Dickstein, Berka, & Hillyard, 1983), and irrelevant probe stimuli that share many selection criteria with attended items (Hink & Hillyard, 1976).

A few similar ERP studies have been conducted with children over 5 years of age. Using a dichotic listening paradigm, Berman and Friedman (1995) showed that when 8-year-old, 14-year-old, and adult listeners attended to either specific pitch ranges or specific syllables to detect longer duration targets while ignoring other stimuli, a negative attention effect was evident between about 200 and 400 ms in all groups. However, the amplitude of this attention effect increased and its latency decreased from childhood to adulthood. In a cross-modal paradigm, children as young as 6 years old showed a larger negativity in response to auditory stimuli when they attended to the auditory as opposed to the visual modality, with the latency of this attention effect decreasing to adult levels by the age of 8 (Satterfield, Schell, Nicholas, Satterfield, & Freese, 1990). In a less canonical paradigm, listeners ages 8–22 were asked either to read a book or to actively attend to a rapid series of standard and rare tones. Typical auditory onset responses (P1, N1, P2, N2, and P3) were reported, but there was no evidence of a negative processing difference in children younger than 17 years of age (Oades, Dittmann-Balcar, & Zerbin, 1997). Overall, it appears that ERP selective auditory attention effects can be elicited in children at least as young as the age of 6 under certain conditions, consistent with the behavioral findings reviewed above.

In a review of both behavioral and ERP studies of the development of selective attention, Ridderinkhof and van der Stelt (2000) proposed that the abilities to select among competing stimuli and preferentially process more relevant information are essentially available in very young children, but that the speed and efficiency of these behaviors and the systems contributing to these abilities improve as children develop. To test this hypothesis more directly, the ERP paradigm employed by Hink and Hillyard (1976) was adapted to make dichotic listening more interesting and engaging for 6- to 8-year-old children (Coch, Sanders, & Neville, 2005). When these listeners were asked to selectively attend to one of two simultaneously presented stories that differed in position (left/right), voice (male/female), and content, children as young as age 6 years showed an auditory selective attention effect on ERPs to probe stimuli with similar onset latency to that observed in adults (100 ms). This finding suggests that, if given strong attentional cues, children as young as 6 years old can selectively attend to auditory information and that the nature and timing of these effects on processing auditory information are similar to those found in adults. Interestingly, the children in this study did not show the typical negative ERP attention effect; instead, irrelevant probe stimuli presented at the same location as the attended story elicited a larger positivity than those sharing the location of the unattended story. Some indication of a similar positive processing difference for the youngest children among 5-, 7-, and 9-year-olds in another selective attention study can be found in data presented in a figure but not otherwise analysed or discussed (Bartgis, Lilly, & Thomas, 2003, Fig. 1, Fig. 2). Positive attention effects in children may also relate to the positive mismatch response reported in other studies with young children (Kushnerenko, Ceponiene, Balan, Fellman, & Näätänen, 2002; Maurer, Bucher, Brem, & Brandeis, 2003).

The results of these studies raise an important issue concerning ERP measures of auditory selective attention effects in young children: to accurately interpret attention effects as modulations of exogenously driven components or added endogenous components, it is necessary to understand the development of the basic underlying auditory evoked potentials (AEPs). Several studies have compared the AEP response to clicks or simple tones preceded by silence in children and adults. Although the distinct positive-negative-positive oscillations (P1, N1, P2) recorded from adults can be identified in at least some children as young as 6 years of age, P1 and N1 responses do not achieve mature amplitudes, latencies, or distributions until close to 20 years of age (Albrecht, Suchodoletz, & Uwer, 2000; Bruneau, Roux, Guérin, Barthélémy, & Lelord, 1997; Martin, Barajas, Fernandez, & Torres, 1988; Ponton, Eggermont, Khosla, Kwong, & Don, 2002; Ponton, Eggermont, Kwong, & Don, 2000). With an increased latency of the first positive component in the waveform and little or no N100 apparent until after the age of 12 (Ponton et al., 2002, Ponton et al., 2000), it could be expected that young children would show a broad positivity rather than an N100 at the onset latency of auditory selective attention effects, as described above (Coch et al., 2005). Moreover, there is some evidence that the impact of a crowded auditory environment (as in these complex dichotic listening paradigms) differs in young children and adults. In an MEG study with children ages 3 months to 15 years and adults, the age at which an adult-like N1m (magnetic equivalent of the N100) emerged was older with shorter interstimulus intervals (Paetau, Ahonen, Salnonen, & Sams, 1995); similar effects have been reported in ERP studies (Ceponiene, Cheour, & Näätänen, 1998).

Considering both the evidence for extended development of the AEP (P1, N1, and P2) and the complexity and rapid presentation of stimuli in auditory selective attention paradigms suitable for young children, it can be expected that the ERPs at a specific latency might be very different for young children and adults. However, it is not yet clear how immature AEPs (i.e., a broad positivity) might interact with potentially more mature ERP attention effects, particularly in very young children. One possibility is that an adult-like processing negativity (Nd) might be superimposed on children's AEPs, as reported for older children (Berman & Friedman, 1995; Satterfield et al., 1990). Alternatively, AEP amplitude in children might be modulated by selective attention in a manner similar to that in adults, resulting in larger responses to attended than unattended stimuli reflected in an Nd in adults but a positive processing difference in children (Coch et al., 2005). From the few studies available it might be hypothesized that there is a developmental progression from attentional modulation of a positivity in very young children to a superimposition of an adult-like Nd on that positivity in older children to a full-fledged, adult-like Nd in older adolescents.

In the present study, we explored the early development of these attentional and perceptual systems by modifying the procedure employed previously with 6- to 8-year-olds (Coch et al., 2005) to engage even younger children. As in the previous study, multiple, redundant attention cues were employed to direct attention to one of two simultaneously presented auditory stories. Children and adults were informed of the location the attended story would be presented from, the voice the attended story would be read in, and the topic of the story they should attend. The present study also used images that corresponded only to the attended story as an additional cue and multiple, related, shorter stories to help listeners sustain attention. We predicted that under these conditions of redundant attention cues and entertaining stimuli, children as young as 3 years old could successfully selectively attend to one story and that the effects of this selective attention on the processing of irrelevant probe stimuli would be an increase in the amplitude of the neural responses to probes from the attended story. More specifically, we hypothesized that probe stimuli would elicit a broad positivity similar to that found in older children and that attention would modulate this positivity such that a comparison between probes played from the attended and unattended sides would yield a positive processing difference.

Section snippets

Participants

The final sample included 13 adults between the ages of 18 and 35 years (eight women), 14 children between the ages of 6 and 8 years (range = 74–106 months, mean = 87 months [7 years, 3 months], seven girls), and 39 children between the ages of 3 and 5 years (range = 40–71 months, mean = 57 months [4 years, 9 months], 21 girls). All participants were right-handed (Oldfield, 1971), monolingual English speakers with no history of neurological or language disorders. The volunteers were paid for their

Adults: nonlinguistic probes

Although the two simultaneously presented stories resulted in almost continuous sound, the onset of nonlinguistic probe stimuli elicited typical auditory evoked potentials in adults as shown in Fig. 3. An early positive component (P1) peaking around 90 ms was largest at medial and anterior electrode sites (Laterality: F(1,12) = 60.26, p < 0.001, Anterior/Posterior: F(5,60) = 23.45, p < 0.001, Laterality × Anterior/Posterior: F(5,60) = 19.93, p < 0.001), but was not affected by attention (p's > 0.40, latency: M

Discussion

The paradigm employed in this study involved an extremely dense auditory environment: two different stories presented simultaneously as continuous natural speech and two types of slightly louder, 100 ms duration probe stimuli presented with variable ISIs at an average of 1.5 probes per second. Even with a complex auditory scene, adults showed typical auditory onset ERP components in response to the probe stimuli. Additionally, adults were able to parse this auditory scene and selectively attend

Acknowledgements

This research was supported by NIH, NIDCD grant DC00481 to H.J.N. D.C was supported by NIH/NICHD grant HD08598, C.S by an NSF Graduate Research Fellowship, and L.D.S. by NIH/NIDCD DC005291. We also wish to thank the many researchers and technicians involved in collecting extensive behavioral and ERP data from young children: Annika Andersson, Paul Compton, Jessica Fanning, Petya Ilcheva, Brittni Lauinger, Nicole Makarenco, David Paulsen, Lisa Stewart, Ray Vukcevich, Brad Wible, and Yoshiko

References (48)

  • D. Pearson et al.

    Auditory attention switching: A developmental study

    Journal of Experimental Child Psychology

    (1991)
  • C. Ponton et al.

    Maturation of human central auditory system activity: Separating auditory evoked potentials by dipole source modeling

    Clinical Neurophysiology

    (2002)
  • C. Ponton et al.

    Maturation of human central auditory system activity: Evidence from multi-channel evoked potentials

    Clinical Neurophysiology

    (2000)
  • K. Ridderinkhof et al.

    Attention and selection in the growing child: Views derived from developmental psychophysiology

    Biological Psychology

    (2000)
  • M. Rueda et al.

    Development of attentional networks in childhood

    Neuropsychologia

    (2004)
  • J. Satterfield et al.

    Ontogeny of selective attention effects on event-related potentials in attention-deficit hyperactivity disorder and normal boys

    Biological Psychiatry

    (1990)
  • B. Shepp et al.

    The development of perceived structure and attention: Evidence from divided and selective attention tasks

    Journal of Experimental Child Psychology

    (1991)
  • B. Shepp et al.

    The development of selective attention: Holistic perception versus resource allocation

    Journal of Experimental Child Psychology

    (1987)
  • J. Bartgis et al.

    Event-related potential and behavioral measures of attention in 5-, 7-, and 9-year-olds

    Journal of General Psychology

    (2003)
  • N. Bruneau et al.

    Temporal prominence of auditory evoked potentials (N1 waves) in 4–8-year-old children

    Psychophysiology

    (1997)
  • R. Ceponiene et al.

    Interstimulus interval and auditory event-related potentials in children: Evidence for multiple generators

    Electroencephalography and Clinical Neurophysiology

    (1998)
  • E.C. Cherry

    Some experiments on the recognition of speech, with one and two ears

    Journal of the Acoustical Society of America

    (1953)
  • E. Clark

    I love you, Blue Kangaroo

    (1998)
  • E. Clark

    Where are you, Blue Kangaroo?

    (2000)
  • Cited by (0)

    View full text