Date of Award


Document Type


Degree Name


Organizational Unit

College of Arts Humanities and Social Sciences, Psychology

First Advisor

Kateri McRae, Ph.D.

Second Advisor

Timothy Sweeny

Third Advisor

Peter Sokol-Hessner


Binocular rivalry, Emotion, Facial expressions, Perception


A central debate in defining emotional space is whether emotions are organized categorically (e.g., fear, happy, disgust) or continuously (i.e., along the independent dimensions of valence and arousal). Emotional facial expressions are one tool often leveraged in trying to define emotional space. Faces are rich sources of social and emotional information. Faces, like emotions, can be organized in either categorical (e.g., happy, sad) or continuous (e.g., open-closed) ways. Therefore, understanding the relatedness of emotional facial expressions to each other may shed light on the underlying structure of emotions. Binocular rivalry (BR) is a tool which can be leveraged to measure the relatedness of two percepts. When each eye is presented with a different image, the visual system is forced to resolve the images into a coherent percept by either selecting one percept to dominate or blending the two images. BR was employed across three experiments with emotional expressions (happy, fear, disgust, sadness, and neutral) to quantify similarities and differences in how the visual system responds to emotional faces. In Study 1, emotional faces dominated over neutral faces. In Study 2, emotion-emotion conflict was explored, and results suggest a positivity bias in emotion perception, as happy faces dominated over all negative faces. In addition, fear dominated over disgust and sad faces. In Study 3, the role of top-down, directed attention on perception was tested by asking participants to direct their attention to the presence or absence of positivity or negativity. Results suggest that the positivity bias observed in Study 2 is enhanced by directed attention towards positivity. Overall, these studies demonstrate that emotion expression information is processed preferentially compared to neutral expressions, that emotion-emotion conflict can be characterized by both positivity and fear biases, and that top-down attention can modulate these biases. Results from these studies were not consistent with any continuous models that were tested. Therefore, results can be interpreted as supporting a categorical emotion model in which happy and fear are prioritized compared to other emotions.

Publication Statement

Copyright is held by the author. User is responsible for all copyright compliance.

Rights Holder

Daniel Stephen Lumian


Received from ProQuest

File Format




File Size

87 p.