Controlling the spotlight of attention: Visual span size and flexibility in schizophrenia
Highlights
► We measure visual span size among patients with schizophrenia. ► We measure how patients’ visual span size changes as a function of task demands. ► We use eye-tracking and the gaze-contingent window paradigm to measure visual span. ► Patients demonstrate smaller visual span sizes on some search conditions. ► Patients do not flexibly modulate their visual span as a function of task demands.
Introduction
The terms visual span, perceptual span, span of effective vision, and useful or functional field of view all refer to the area of the visual field from which information is extracted during each eye fixation (Rayner, 1998). A larger visual span size has been linked to more efficient performance on visual search tasks where one searches for a target among a number of non-target items (e.g., Bertera and Rayner, 2000, Rayner and Fisher, 1987, Reingold et al., 2001). In the presence of a wider visual span, more visual information is processed during each eye fixation and fewer fixations are required to complete the search.
Patients with schizophrenia (SCZ) are impaired on visual search tasks; that is, they are less efficient in detecting a target, especially when it is presented in a crowded display (e.g., Elahipanah et al., 2010, Fuller et al., 2006, Lieb et al., 1994, Mori et al., 1996). Patients are also known to have restricted attentional resources (e.g., Granholm, 1992, Granholm et al., 1997, Nuechterlein and Dawson, 1984); a deficit that impedes their performance on a wide variety of tasks requiring attention (e.g., Span of Apprehension Task; Continuous Performance Test). Given that research with healthy participants has shown that experimental manipulations that decrease the availability of attentional resources (e.g., divided attention conditions) cause a reduction in visual span size (e.g., Pomplun, Reingold, & Shen, 2001; Wood et al., 2006), we hypothesized that limited attentional resources among patients with SCZ serve to produce a smaller visual span resulting in impaired visual search performance. In a previous study (Elahipanah et al., 2010), it was found that patients’ visual search impairment was accentuated as a function of target eccentricity. Given that participants were instructed to fixate a central fixation cross at the onset of each search display, this finding is consistent with the hypothesis that patients with SCZ may have a smaller visual span and are, therefore, less likely to detect a peripheral target prior to executing the first saccade. However, in the aforementioned study, visual span size was inferred from the pattern of response times (RTs) and, consequently, it was not possible to rule out other explanations for patients’ disproportionate impairment in detecting more peripheral targets, such as different search strategies or eye movement patterns.
The current experiment aimed to investigate this hypothesis more directly by employing eye-tracking technology and a variant of the gaze-contingent moving window paradigm (McConkie & Rayner, 1975). In this paradigm, the identity of search items is obscured except those within a certain “window” that is continuously centered on the observer's current point of gaze, thereby restricting the visual information acquired during each eye fixation (see Fig. 1, Fig. 2). The observer's visual span size is then measured by varying the size of this window over successive trials according to an iterative algorithm to determine the smallest window size that does not significantly interfere with task performance (Pomplun et al., 2001, Reingold et al., 2001). In addition to comparing visual span size across patients and healthy participants, the current study also contrasted these groups in terms of their flexible modulation of visual span size as a function of task difficulty. Specifically, among healthy individuals, it has been shown that under more difficult visual search conditions, the size of the visual span decreases to accommodate the greater processing demands (Greene & Rayner, 2001Pomplun et al., 2001, Rayner and Fisher, 1987). This flexible modulation of visual span as a function of task demands suggests a normative online adjustment of the distribution of attention. In contrast, impaired modulation of visual span size is another possible deficit in the control of visual attention that might underlie the observed pattern of visual search deficits among patients with SCZ.
In the present study, we measured patients’ visual span size compared to healthy participants on two different visual search tasks and as a function of two different manipulations of task difficulty. As outlined below, both these manipulations have been shown to influence visual span size in healthy participants. We reasoned that the emergence of a consistent pattern of results across both tasks and conditions would more likely reflect a stable between-group difference rather than a finding that is restricted to a specific search condition. This issue is especially pertinent given prior demonstrations of the task-dependent nature of visual span.
The first search task in the present study (similarity task) employed two different levels of target–distractor similarity (T–D similarity). Decreasing the similarity between the target and distractors results in greater target discriminability, a more efficient visual search (Duncan & Humphreys, 1989), and larger visual spans (e.g., Rayner & Fisher, 1987) among healthy participants. The second task (familiarity task) contrasted search for an unfamiliar target among familiar distractors (U–F condition) with search for a familiar target among unfamiliar distractors (F–U condition). Among healthy participants, the former condition has been shown to produce a more efficient search than the latter condition; that is, a familiarity-based search asymmetry is observed (e.g., Frith, 1974, Richards and Reicher, 1978, Shen and Reingold, 2001). In addition, a U–F search is associated with larger visual spans than an F–U search (Greene & Rayner, 2001). It has been found that familiarity with distractors is the main determinant of search efficiency in the U–F search condition (Shen & Reingold, 2001). This familiarity effect is hypothesized to occur because unfamiliar distractors require extra processing, which results in a narrowing of the visual span so that fewer items are processed during each eye fixation (Greene & Rayner, 2001).
Thus, the main goal of the present study was to compare visual span size between patients and healthy participants on both the similarity and familiarity visual search tasks. In addition, the secondary goal of our study was to ascertain whether patients with SCZ can flexibly modulate their visual span across task conditions.
Section snippets
Participants
Thirty-two patients diagnosed with SCZ or Schizoaffective Disorder according to the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association [DSM-IV-TR], 2000) and 26 healthy control participants completed task 1; twenty-six patients and 24 healthy controls completed task 2.1 Patients were recruited
Similarity task
Both groups obtained above 98% accuracy across both conditions, suggesting minimal speed-accuracy trade-offs. After removal of trials with incorrect responses or outlier RT values, from the baseline (no-window) trials, median RT was obtained for each participant in each condition (low vs. high T–D similarity), and baseline search performance was evaluated with a 2(T–D similarity) × 2(group) mixed models ANOVA with T–D similarity as the within-subject factor and group as the between-subjects
Discussion
Patients in the current study demonstrated slower search across all search conditions. This finding is consistent with previous reports of impaired visual search performance among patients with SCZ (e.g., Elahipanah et al., 2010, Elahipanah et al., 2011, Fuller et al., 2006, Mori et al., 1996). The central hypothesis of the current study, however, was that patients with SCZ have a narrower visual span; that is, they process information from a smaller area of their visual field during each eye
Acknowledgements
Ava Elahipanah was supported by an NSERC scholarship during the course of the study. The authors are grateful to Jiye Shen for providing excellent technical and programming support. The authors would also like to thank Edward McAnanama for providing clinical support and assistance in recruiting and screening of participants.
References (49)
Diagnostic and statistical manual of mental disorders
(2000)- et al.
Eye movements and the span of the effective visual stimulus in visual search
Perception & Psychophysics
(2000) - et al.
Effects of similarity and history on neural mechanisms of visual selection
Nature Neuroscience
(1999) - et al.
Early-stage visual processing deficits in schizophrenia
Current Opinion in Psychiatry
(2005) - et al.
Dysfunction of early-stage visual processing in schizophrenia
The American Journal of Psychiatry
(2001) - et al.
Common neural circuitry supporting volitional saccades and its disruption in schizophrenia patients and relatives
Biological Psychiatry
(2008) - et al.
Attention in schizophrenia: Signal detection in the visual periphery
Journal of Nervous and Mental Disease
(1981) - et al.
Motion perception in schizophrenia
Archives of General Psychiatry
(1999) - et al.
Control of goal-directed and stimulus driven attention in the brain
Nature Reviews: Neuroscience
(2002) - et al.
Visual search and stimulus similarity
Psychological Review
(1989)