Optimizing Signal Detection: A Parametric Approach to Assessment and Training

Principal investigator: Spencer Lynn
Source: US Army Research Institute for the Behavioral and Social Sciences
Award: W911NF-16-1-0192
Dates: 5/16/16-5/15/17
Amount: $89,978

Can inaccuracies in a person’s subjective “cognitive model” of the operational environment be identified and corrected, to improve decision making? In our prior ARI-funded work we developed a signal detection theory (SDT) framework to define and manipulate environmental parameters in a social threat perception task and to measure individual differences predictive of threat detection abilities. Here, we propose to extend that work, developing means to quantitatively assess perceivers’ cognitive model of the environment, provide individually tailored training targeting a person’s environmental parameter “misestimate,” and describe neurophysiological (EEG) correlates of parameter estimation and training effectiveness. In a one-year project, 100 participants will complete a baseline social-threat perception test. Results will determine individual vulnerabilities to misestimating three environmental parameters known from SDT to control threat detection effectiveness. Subsequently, participants will receive a training protocol and a retest. We hypothesize that participants who receive training specific to their misestimated parameter will show greater improvement than participants who receive training on an accurately estimated parameter. We will assess how executive function and personality traits may modulate the efficacy of neurophysiological measures as putative markers of parameter estimation and training effectiveness. We will address High Priority Research Questions concerning: assessing learning processes and learner status to tailor training individually, linking constructs of adaptability to job performance, and determining neurophysiological individual differences related to core military skills. Therefore, if successful this project could transition to applications relevant to Army objectives in Learning in Formal & Informal Environments, Personnel Testing & Performance, and Psychophysiology of Individual Differences.

Optimizing Threat Detection Under Signal-Borne Risk

Principal investigator: Spencer Lynn
Source: US Army Research Institute for the Behavioral and Social Sciences
Contract: W5J9CQ-12-C-0028
Dates: 9/27/12-9/26/15
Amount: $434,499

Emotion perception research has revealed marked variability in people’s abilities to infer the emotional states of others. This variability is a function of (i) the uncertainty and risk in the environment inherent to perception (perceivers cannot be certain about what they are experiencing, and errors of perception may be costly) and (ii) factors internal to individual perceivers (physical and psychological states and traits). Using a novel utility-based signal detection framework, we will examine how individual differences in affective reactivity, executive function, and motivation contribute to this variability in perception and decision-making, under conditions of changing environmental uncertainty and risk.

The Utility of Threat Detection in Generalized Social Anxiety Disorder

Principal investigators (multi-PIs): Spencer Lynn, Naomi Simon
Source: National Institute of Mental Health
Award: R01 MH093394-01
Dates: 8/1/11-4/30/16
Amount: $1,954,208


Summary: During social interactions, we look into the face of another person and in the blink of an eye infer that person’s emotional state and their intentions. These perceptions inform decisions about what to do or say next. Generalized Social Anxiety Disorder (GSAD) is characterized by exaggerated concerns about negative evaluation and rejection in social situations. These symptoms have been quantified with signal detection theory (SDT). The application of SDT has led to novel approaches within anxiety research; a primary hypothesis, supported by several studies, has been that the “over-reactive” nature of the anxious state can be characterized as a bias to respond to or remember situations as more threatening than they in fact are. In spite of SDT’s power, its conventional use has been limited to simply quantifying differences in sensitivity, bias, and accuracy among perceivers. Left unanswered are questions of particular relevance to research and treatment: what causes the observed differences in bias and sensitivity? A critical barrier to answering this question is the current understanding of SDT in clinical research, which lacks a framework to predict or explain behavior, or in which to pose experimental questions about how mood and anxiety disorders influence the underlying mechanisms involved in threat perception. To bridge this barrier, we introduce a mathematical model of perceptual decision making that incorporates key insights from behavioral economics-utility and optimality- into a signal detection framework. Our primary objective is to use this novel framework to explain differences in threat perception among individuals with GSAD, anxious controls with generalized anxiety disorder (GAD), and non-psychiatrically-ill participants. Our secondary objective is to assess whether our framework could be used to improve interventions to reduce misperceptions of threat in GSAD. Our model is a unique conceptualization of perception (e.g., optimal detection, subjective miscalibration to underlying environmental parameters that influence overt behavior) that could eventually lead to improvements in cognitive-behavioral therapies by tailoring them to a patient’s individual perceptual decision-making impairment. To achieve our aims, we will recruit 100 individuals with GSAD and 100 individuals each from age- and gender-matched GAD and healthy populations. Participants will complete a suite of perceptual tasks to isolate which of several perceptual decision parameters cause misperceptions of social threat in GSAD. Successful characterization of GSAD along such lines will take the field in new directions by framing social threat perception as a decision made by attempting to optimize detection in the presence ambiguous sensory information and conflicting, risky consequences. As well, the novel theoretical developments represented by our model will broaden SDT’s usefulness deepening the insights it affords into the nature of cognitive processes.

Public Health Relevance: Generalized social anxiety disorder (GSAD) is characterized by frequent, debilitating misperceptions of threat and disapproval in non-threatening social circumstances. This research uses a novel theory and method for characterizing various pathways for disordered threat detection in GSAD. The findings will enable clinicians to build more effective behavioral and cognitive therapies by tailoring therapy to target an individual patient’s particular pathways to perceptual impairments.

Quantifying a Multiple Deficits Approach to Impaired Facial Affect Processing in Schizophrenia

Principal Investigator: Spencer Lynn
Source: Harvard Medical School, Department of Psychiatry
Award: Livingston Award
Dates: 7/1/2007-6/30/2008
Amount: $10,000

The objective of this study was to investigate the causes of emotional impairments and their specificity to schizophrenia by examining event related brain potentials elicited by an important social stimulus–emotional facial expressions–in the context of emotional and non-emotional impairments.

Neurophysiological Correlates of Facial Emotion Perception in Schizophrenia and Manic Psychosis

Role: Post-doctoral Fellow
Source: Clinical Research Training Program in Biological and Social/Developmental Psychiatry, Judge Baker Children’s Center, Department of Psychiatry, Harvard Medical School (T32 MH016259).
Dates: 7/1/2005-6/30/2007
Preceptor: Dean Salisbury, PhD, Cognitive Neuroscience Laboratory, McLean Hospital, Belmont, Massachusetts