Tackling Task Confounds to Understand Face Perception in Schizophrenia Spectrum Disorders

Abstract

People with schizophrenia spectrum disorders (SSD) experience difficulties interpreting information from faces. A longstanding debate questions whether these difficulties reflect: (1) a specific impairment in emotion face perception; (2) equal impairments across all aspects of face perception (e.g., identity and age perception, as well as emotion); or (3) broader cognitive-perceptual deficits that influence processes such as memory, attention and executive functioning. Face perception difficulties may contribute to the social functioning challenges that characterise SSD (Burns & Patrick, 2007), making it essential to understand which areas of face perception are impaired and to what degree. Despite extensive literature on face perception in SSD, methodological limitations have prevented the differentiation of task-specific from broader cognitive-perceptual impairments. This thesis outlines and implements the methodological requirements needed to reveal a clearer profile of SSD-related face perception abilities.

This thesis employed both meta-analytic and experimental approaches. First, Chapters 2 and 3 presented meta-analyses investigating whether SSD-related impairments are significantly larger for emotion compared to non-emotion face perception tasks. Sample and task characteristics were explored as moderators and potential sources of heterogeneity across the literature. Building on these studies, Chapter 4 developed novel experimental tasks matched on psychometric properties to measure face perception free from task-related confounds. Finally, Chapter 5 examined how SSD-related characteristics correlate with performance on these matched tasks in a general population sample, enabling investigation of face perception abilities while controlling for unwanted task-related artefacts.

The meta-analytic findings in Chapters 2 and 3 revealed significantly larger SSD-related impairments for emotion compared to non-emotion face perception. While this effect was robust across varying sample characteristics (i.e., symptom severity, intelligence matching), task characteristics influenced results. Using memory-dependent tasks inflated impairment size, while using highly easy tasks (that diverged from optimal, moderate difficulty) produced smaller impairments. When these confounds were controlled for by using matched experimental tasks in Chapter 5, SSD characteristics continued to predict poorer emotion perception performance relative to non-emotion (identity) face perception. However, this difference appeared to be driven by a relative strength in identity perception rather than a differential emotion impairment. These findings confirm that emotion face perception is a substantial area of difficulty in SSD. However, findings challenge assumptions about the mechanisms of its origins. Two key questions emerge from this research: (1) do emotion face perception impairments exist beyond general cognitive-perceptual difficulties; and (2) does this performance profile reflect a specific deficit or a relative strength? This thesis highlights challenges when measuring cognitive-perceptual abilities in SSD populations and introduces methods to overcome these challenges. Results highlight the need to minimise task-specific confounds and employ experimental paradigms that differentiate general or domain-specific impairments from areas of relative strength. Show more