Elsevier

Consciousness and Cognition

Volume 60, April 2018, Pages 25-36
Consciousness and Cognition

Effects of perceptual load and socially meaningful stimuli on crossmodal selective attention in Autism Spectrum Disorder and neurotypical samples

https://doi.org/10.1016/j.concog.2018.02.006Get rights and content

Highlights

  • Both NT and ASD samples noticed the Socially Meaningful CAS at similar rates across both High and Low visual perceptual loads.

  • ASD participants had significantly greater detection rates of the Non-socially Meaningful CAS than NT participants under High Load.

  • Data suggests an inability to engage selective attention in ASD.

  • NT individuals may allocate a special status to socially meaningful stimuli.

Abstract

The present study examined whether increasing visual perceptual load differentially affected both Socially Meaningful and Non-socially Meaningful auditory stimulus awareness in neurotypical (NT, n = 59) adults and Autism Spectrum Disorder (ASD, n = 57) adults. On a target trial, an unexpected critical auditory stimulus (CAS), either a Non-socially Meaningful (‘beep’ sound) or Socially Meaningful (‘hi’) stimulus, was played concurrently with the presentation of the visual task. Under conditions of low visual perceptual load both NT and ASD samples reliably noticed the CAS at similar rates (77–81%), whether the CAS was Socially Meaningful or Non-socially Meaningful. However, during high visual perceptual load NT and ASD participants reliably noticed the meaningful CAS (NT = 71%, ASD = 67%), but NT participants were unlikely to notice the Non-meaningful CAS (20%), whereas ASD participants reliably noticed it (80%), suggesting an inability to engage selective attention to ignore non-salient irrelevant distractor stimuli in ASD.

Introduction

Selective attention is the process of focusing on, and reacting to certain stimuli when several occur simultaneously (Broadbent, 1958, Peterson and Posner, 2012, Treisman and Riley, 1969). The ability to ignore certain stimuli, whilst attending to other aspects of the environment is important to prevent overloading our sensory and perceptual systems. Research on selective attention in Autism Spectrum Disorder (ASD) is a large, complex, and expanding literature (see Fein, 2011, Just and Pelphrey, 2013, Marco et al., 2011, for example). ASD is a lifelong neurodevelopmental condition, defined in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5 American Psychiatric Association, 2013) by impaired communication, impaired social interactions, and repetitive behaviors (see also Landa, Holman, & Garrett-Mayer, 2007). Attentional deficits have been noted in the condition since Kanner’s (1943) and Asperger’s (cf. Frith, 1991) original definitions. Much literature suggests that people with ASD often have difficulties processing everyday sensory information and in focusing attentional resources which can be detrimental to social functioning (e.g., Baron-Cohen, 2008, Frith and Mira, 1992, Laurie, 2014). There have been a number of recent advances in the field of selective attention in ASD.

Much of the research on selective attention in ASD has provided contradictory findings. On one hand, for example, Burack (1994) suggested that children with ASD have a deficit in their selective attentional lens insofar as ASD participants were more distracted by peripheral visual stimuli, therefore, suggesting an inability to narrow their focus on a target stimulus. Participants in this study were asked to state whether they saw an ‘O’ or a ‘+’ in the centre of the screen, and their reaction time was recorded. Variables manipulated included the presence/absence of a window only showing the central target stimulus, the amount of non-target distractor stimuli, and the distance between the target and distractor stimuli. The results demonstrated that the presence of distractor stimuli negative impacted performance in the ASD group compared to the control group. These findings led to the conclusion that children with ASD had an overly broad attentional lens that they were unable to narrow efficiently (Burack, 1994; see also Adams and Jarrold, 2012, Christ et al., 2011, Ciesielski et al., 1990, Kanakri et al., 2017, Smith and Milne, 2009).

O’Riordan (2004) used visual search tasks to show that ASD participants had a superior ability to detect target stimuli compared to neurotypical (NT) participants. Furthermore, when distractor stimuli were more similar to target stimuli NT participants showed significantly reduced detection of target stimuli compared to ASD participants. This suggests that adults with ASD have enhanced selective attention abilities or a superior visual discrimination ability more specifically. While it could be argued that performance on the visual search task may be affected by memory for rejected distractors in search abilities, Joseph, Keehn, Connolly, Wolfe, and Horowitz (2009) accounted for the potential confounding influence of memory by using both a standard static and a dynamic search task, with target and distractor stimuli randomly changing position every 500 ms. Their findings showed that ASD participants had quicker reaction times and their performance accuracy was greater in the dynamic search task than that of NT participants. This suggests ASD participants have an enhanced ability to discriminate between targets and distractors at the locus of attention.

Both O’Riordan, 2004, Joseph et al., 2009 have argued that such performances observed in ASD samples is characterized by a superior selective attention coupled with an overly narrow attentional lens. The findings from these studies are supported by ASD individuals outperforming NT individuals in Stroop tests (Adams & Jarrold, 2009), and the weak central coherence theory of ASD (Frith, 1989, Happé, 1996, Happé and Frith, 2006). The weak central coherence theory suggests that individuals with ASD overly focus on the smaller parts of an overall picture, and are therefore more able to pick out finer detail due to selectivity in attention leading to enhanced focus and fixation on minor details. However, it could feasibly be argued that O’Riordan, 2004, Joseph et al., 2009 data could be accounted for by enhanced awareness of an overall visual scene (e.g., Kanakri et al., 2017).

Thus, it is apparent that there exists ambiguity in the literature on how selective attention functions in ASD. The discrepancies noted in previous selective attention research in ASD (e.g., for an impaired selective attention and an overly broad attentional lens; Burack, 1994, Smith and Milne, 2009; versus an overly narrow lens and superior selective attention; Joseph et al., 2009, O’Riordan, 2004) has recently been addressed by one particularly intriguing line of research that incorporates Lavie’s (1995) Load theory of attention and cognitive control (e.g., Remington et al., 2009, Remington, Swettenham, et al., 2012, Swettenham et al., 2014, Tillmann et al., 2015, Tillmann and Swettenham, 2017). Load theory suggests that task-irrelevant distractor stimuli will only be processed if there are enough cognitive resources left over after the primary target stimuli have been processed. In other words, the bigger the perceptual load of the main task, the less the ability to process additional stimuli. If the perceptual load is low, a ‘spill-over’ of attentional resources will occur and additional stimuli will be processed automatically. For example, perceptual load can be manipulated by altering the number of task-relevant stimuli in a display (e.g., number of items in a search task; Tillmann & Swettenham, 2017), or the perceptual processing requirement of a task (e.g., the subtlety of a line discrimination; Lavie, 2005, Tillmann et al., 2015).

Remington et al. (2009) hypothesized that individuals with ASD would have an enhanced visual perceptual load capacity as past research had shown that they performed better than control groups in visual search tasks when there were a large number of visual stimuli (e.g., O'Riordan, Plaisted, Driver, & Baron-Cohen, 2001). They explored perceptual load capacity in an ASD and a NT control group, by using a mix of both a visual search task (Treisman & Gelade, 1980) and a flanker task (Eriksen & Eriksen, 1974). Wherein both NT and ASD groups showed they were still processing distractor (flanker) stimuli at a low perceptual load search task (i.e., finding two target items amongst distractor stimuli) the NT group showed signs of distractor stimuli interference at a higher perceptual load (four target items), whereas the ASD group did not. At the highest perceptual load (i.e., six target items) both groups showed distractor stimuli interference. This suggests that there might be a higher perceptual load capacity in ASD individuals as it took six target items to exhaust their perceptual capacity, whereas NT individual’s perceptual capacity was exhausted after just four target items. These findings were further supported in a study of visual detection sensitivity (Remington, Campbell, et al., 2012, Remington, Swettenham, et al., 2012) and the detection of an unexpected task-irrelevant visual stimulus in an inattentional blindness task (Swettenham et al., 2014).

Such studies reviewed above, however, only tested the effects of perceptual load on selective attention in ASD in the visual domain. Therefore, it was not possible to generalise these findings across sensory modalities. Cross-modality research needs to be further explored as sensory input is not usually limited to one sense (e.g., vision; see Stein, 2012), and ASD individuals often present signs of overstimulation in all senses (Laurie, 2014). Harrison and Davies (2013) acknowledged the complexity of unpacking crossmodal attention research (see also, Klapatek et al., 2012, Liu et al., 2012, Peterson and Posner, 2012), and noted that previous research had shown that visual spatial attention can be modulated by emotional prosody cues but that it was unclear whether this “crossmodal modulation of visual attention is associated with the engagement or disengagement of attentional resources” (p. 247). The debate tends to center around the relative importance of bottom-up versus top-down processes (e.g., Mast and Frings, 2014, Mast et al., 2017) in determining attentional focus or attentional capture in crossmodal attention paradigms. For example, Mast et al. (2017) argued for the role of top-down ‘attentional control sets’ between vision audition but also acknowledged that “other studies have suggested that the detection of simultaneity across the senses and its influence on brain and cognitive processes is independent of the particular task, population or state of the individual (e.g., conscious awareness)” (p. 46). In other words, there is some support for bottom-up processes in selective attention, independent of task environment or top-down expectancies or attentional control sets.

Research has shown that NT individuals show an increased inattentional deafness of unexpected audio stimuli when the visual perceptual load is increased (Macdonald and Lavie, 2011, Molloy et al., 2015). Tillmann et al. (2015) examined the effects of perceptual load on crossmodal (i.e., visual and audio) selective attention in ASD. They explored inattentional deafness, and hypothesized that unlike NT individuals, ASD individuals would be able to detect the unexpected audio stimulus in higher visual load tasks as past research (e.g., Remington et al., 2009, Remington, Campbell, et al., 2012, Remington, Swettenham, et al., 2012, Swettenham et al., 2014) suggests that ASD individuals may have an enhanced perceptual capacity. In a similar procedure to Swettenham et al., 2014, Tillmann et al., 2015 manipulated the lengths of lines on a cross to create a low visual perceptual load condition (i.e., the lines were very clearly a different length and easy to discriminate), and a high visual perceptual load condition (i.e., the lines were much closer in length and more difficult to discriminate) (see Lavie, 2005). However, instead of using an unexpected visual stimulus (Swettenham et al., 2014), they used an unexpected auditory ‘beep’ presented simultaneously with the target cross. Tillmann et al. found that while both NT and ASD groups had similarly high detection rates in the low visual perceptual load condition, NT participants had significantly reduced detection rates of the unexpected audio stimulus in the high visual perceptual load condition. Importantly, detection rates remained high in the ASD group (supporting Macdonald and Lavie, 2011, Molloy et al., 2015). These findings further supported the proposition that there is an enhanced perceptual tcapacity in ASD (e.g., Remington et al., 2009, Remington, Campbell, et al., 2012, Remington, Swettenham, et al., 2012, Swettenham et al., 2014), and moreover, that this perceptual capacity is shared across sensory modalities.

As described in the DSM-5, two of the main impairments of ASD are diminished communication and social interaction abilities. As suggested by Tillmann et al. (2015) there is a clear need to explore whether stimuli that convey biological and socially relevant information (e.g., a person greeting another person, or a person’s face) would produce different results. For example, in the visual and spatial attention literature there is considerable debate concerning whether socially (or biologically) relevant cues “are special when it comes to triggering reflexive shifts in attention….[as] it is not clear whether all cues belong to the same category or whether social information has a distinct functional role in cueing attention” (Wilson, Soranzo, & Bertamini, 2017, p. 56; see also Birmingham and Kingstone, 2009, Langton et al., 2000, Ristic and Kingstone, 2012). There is some ambiguity in the literature in terms of operational definitions of what constitutes a socially relevant cue. For instance, Wilson et al. (2017) critiqued a visual perspective taking attentional interference study by Nielsen, Lance, Levy, and Holmes (2015) who employed a human avatar, an arrow, or a dual-colored block as distractor stimuli. Nielsen et al. argued that the magnitude of attentional interference effects caused by irrelevant distractor cues in cognitive tasks depends on the level or amount of social characteristics that the cue has or conveys. Furthermore, they explained their attentional interference or intrusion findings, according to different levels of attributed social-relevance to each of the three cues, designating the avatar as social, the arrow as semi-social, and the dual-colored block as non-social. Nielsen et al. suggested that the more social the distractor cue, the stronger observed intrusion effects will be in attention-based paradigms. However, it could reasonably be argued that the levels of attributed social ‘quality’ to the three cues was quite arbitrary.

Of interest to the present study, Remington, Campbell, and Swettenham (2012) employed socially relevant distractors (i.e., faces) in a visual search task. They found that NT individuals processed these distractors across all visual perceptual load levels, whereas ASD individuals only processed them at low loads. These findings are contradictory to past research (Remington et al., 2009) where non-socially meaningful visual stimuli (i.e., shapes) were used, suggesting that NT individuals give a special status to socially meaningful stimuli, and process them in an automatic fashion regardless of the perceptual load of a relevant task (see Lavie, Ro, & Russell, 2003). These findings fit in with Cherry’s (1953) ‘cocktail party phenomenon’, wherein people (NT) show a shift of attention when they hear socially meaningful auditory stimuli (e.g., their name) (cf. Wood & Cowan, 1995).

What remains to be examined is whether socially meaningful auditory stimuli would elicit greater detection rates in NT individuals across visual perceptual loads compared to ASD individuals. Exploring this area further will provide a clearer understanding of how selective attention functions in ASD, and shed some more light on how socially relevant stimuli are processed in ASD. This may have real world implications as it could help provide better understanding of how to support and care for ASD individuals. Furthermore, much of the research on attention in ASD has been conducted with child or adolescent populations (e.g., Tillmann & Swettenham, 2017). The present study employed an adult sample, and therefore, may help elucidate whether the findings with youthful populations to date also generalise across older age groups.

The present study used a line discrimination task to manipulate visual perceptual load (Lavie, 2005, Tillmann et al., 2015). Based on previous findings of enhanced perceptual capacity in ASD (Remington et al., 2009, Remington, Campbell, et al., 2012, Remington, Swettenham, et al., 2012, Swettenham et al., 2014), that attentional resources are allocated across sensory modalities in ASD (Tillmann et al., 2015), and that socially meaningful visual stimuli produce greater attentional demand in NT individuals (Remington, Campbell, et al., 2012, Remington, Swettenham, et al., 2012), the present study explored the rates of detection of unexpected Non-socially Meaningful (‘beep’), and Socially Meaningful (a voice saying ‘hi’) auditory stimuli across visual perceptual loads in NT and ASD groups. Based on the literature reviewed (e.g., Macdonald and Lavie, 2011, Remington, Campbell, et al., 2012, Remington, Swettenham, et al., 2012, Swettenham et al., 2014) we hypothesized that the ASD group would show an enhanced perceptual capacity that functions across sensory modalities, by having similarly high detection rates of an unexpected Non-socially Meaningful auditory stimulus across visual perceptual loads, whereas the NT group would show significantly reduced detection rates in the high visual perceptual load. Furthermore, we hypothesized that NT individuals would show similarly high detection rates of an unexpected Socially Meaningful auditory stimulus across both visual perceptual loads. As noted above, this research could help elucidate whether bottom-up or top-down processes have the greater influence on attentional capture (Mast et al., 2017). Moreover, this study could help explicate whether ASD persons have a genuinely enhanced perceptual capacity compared to NT individuals (Bayliss and Kritikos, 2011, Mayer, 2017). More specifically, if NT samples show significantly higher detection rates of an unexpected Socially Meaning auditory stimulus than to an unexpected Non-socially Meaningful auditory stimulus while engaged in a separate primary visual attention task (i.e., line discrimination), then it suggests, rather, that the central issue is likely one of disengagement of attention to irrelevant environmental stimuli (Harrison & Davies, 2013) in NT persons, and a failure to disengage in ASD samples.

Section snippets

Participants

One-hundred and twenty-one participants completed the present study. Of these, 59 were neurotypical (NT) adults (31 male, 28 female, M age = 32.46 yrs., SD = 3.68, age range: 18–65 yrs.), recruited from the University of Chichester’s participation pool for psychology students and advertisement via university email network, and snowball sampling. Sixty-two adults with a diagnosis of Autism Spectrum Disorder (ASD) took part (44 male, 18 female, M age = 26.15 yrs., SD = 1.98, age range:

Controls

The proportion of male participants in the NT group (53%) was different from the proportion of males in the ASD group (70%), but not significantly so, χ2(1) = 3.09, p = .079 (R Core Team, 2017). This is in line with the higher population prevalence of ASD among males than females. We assessed whether gender affected detection of the CAS and it did not (OR, 1.680; 95% CI, 0.74–3.988; p = .224). Gender was not included in further analyses.

As mentioned in the method, we included participants who

Discussion

In line with the first hypothesis, the results show that both Diagnostic group (NT vs. ASD) and Perceptual Load (Low vs. High) had a significant impact and interaction on detection rates of the Non-socially Meaningful critical auditory stimulus (CAS). Specifically, whereas both Diagnostic groups had similarly high detection rates in the Low Load Condition, the NT group showed significantly reduced detection rates in the High Load compared to the Low Load, and the ASD group showed no difference

Compliance with ethical standards

There was no funding provided for the present study.

Ethical approval: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed consent: Informed consent was obtained from all individual participants included in the study.

Conflict of interest

Ian Tyndall declares that he has no conflict of interest. Liam Ragless declares that he has no conflict of interest. Denis O’Hora declares that he has no conflict of interest.

References (60)

  • N. Adams et al.

    Inhibition in autism: Children with autism have difficulty inhibiting irrelevant distractors but not prepotent responses

    Journal of Autism and Developmental Disorders

    (2012)
  • American Psychiatric Association

    Diagnostic and statistical manual of mental disorders DSM-IV-TR

    (2000)
  • American Psychiatric Association

    Diagnostic and Statistical Manual of Mental Disorders

    (2013)
  • S. Baron-Cohen

    Autism and Asperger syndrome

    (2008)
  • S. Baron-Cohen et al.

    The autism-spectrum quotient (AQ): Evidence from Asperger syndrome/high functioning autism, males and females, scientists and mathematicians

    Journal of Autism and Developmental Disorders

    (2001)
  • A.P. Bayliss et al.

    Brief report: Perceptual load on the autism spectrum in typically developed individuals

    Journal of Autism and Developmental Disorders

    (2011)
  • E. Birmingham et al.

    Human social attention: A new look at past, present, and future investigations

    Annals of the New York Academy of Sciences

    (2009)
  • D. Broadbent

    Perception and communication

    (1958)
  • J. Burack

    Selective attention deficits in persons with autism: Preliminary evidence of an inefficient attentional lens

    Journal of Abnormal Psychology

    (1994)
  • E. Cherry

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

    The Journal of the Acoustical Society of America

    (1953)
  • S. Christ et al.

    Evidence for selective inhibitory impairment in individuals with autism spectrum disorder

    Neuropsychology

    (2011)
  • B. Eriksen et al.

    Effects of noise letters upon the identification of a target letter in a non-search task

    Perception & Psychophysics

    (1974)
  • D. Fein

    The neuropsychology of autism

    (2011)
  • U. Frith

    Autism: Explaining the Enigma

    (1989)
  • U. Frith

    Asperger and his syndrome

  • U. Frith et al.

    Autism and Asperger syndrome

    Focus on Autism and other Developmental Disabilities

    (1992)
  • F. Happé

    Studying weak central coherence at low levels: Children with autism do not succumb to visual illusions. A research note

    Journal of Child Psychology and Psychiatry

    (1996)
  • F. Happé et al.

    The weak coherence account: Detail-focused cognitive style in autism spectrum disorders

    Journal of Autism and Developmental Disorders

    (2006)
  • N.R. Harrison et al.

    Modulation of spatial attention to visual targets by emotional environmental sounds

    Psychology & Neuroscience

    (2013)
  • R. Joseph et al.

    Why is visual search superior in autism spectrum disorder?

    Developmental Science

    (2009)
  • View full text