Cory Costello is a PhD candidate in psychology at the University of Oregon. He is a member of the Personality and Social Dynamics Lab (PI: Sanjay Srivastava). His research interests include Personality (who we are), Interpersonal Perception (what people think about us), Reputations (what people say and hear about us), how these are manifest in online environments, and the extent to which they’re recoverable from digital footprints. He combines survey, experimental, and data scientific methods in pursuit of these interests. He is an open science and R enthusiast. His methodological interests include social network analysis, predictive modelling / machine learning, text analysis, structural equation modelling, hierarchical linear modelling.
PhD in Personality/Social Psychology, 2020 (expected)
University of Oregon
MA in Psychology, 2014
Wake Forest University
BA in Psychology, 2012
New College of Florida (The State's Honors College)
A Naturalistic, laboratory paradigm for studying the spread of reputational information.
Work examining longitudinal personality change and stability in the life and time project
The past decade has seen rapid growth in research linking stable psychological characteristics (i.e., traits) to digital records of online behavior in Online Social Networks (OSNs) like Facebook and Twitter, which has implications for basic and applied behavioral sciences. Findings indicate that a broad range of psychological characteristics can be predicted from various behavioral residue online, including language used in posts on Facebook (Park et al., 2015) and Twitter (Reece et al., 2017), and which pages a person ‘likes’ on Facebook (e.g., Kosinski, Stillwell, & Graepel, 2013). The present proposal seeks to examine the extent to which the accounts a user follows on Twitter – their Twitter friends – can predict individual differences in self-reported anxiety, depression, post-traumatic stress, and anger. Studying Twitter friends offers distinct theoretical and practical advantages for researchers, including the potential for less overt impression management and better capturing passive users. By incorporating best practices in open science and machine learning, we aim to provide unbiased estimates of predictive accuracy for predicting Mental Health from Twitter friends. Our findings will have implications for theories linking psychological traits to behavior online, applications seeking to infer psychological characteristics from records of online behavior, and for informing discussions of how such applications could affect users’ privacy.
Reputations are critical in human social life: they allow people to share and act on information about one another, even when they have never met. Reputations can be conceptualized as information about a target person that is stored in networks of perceivers and transmitted through either direct interaction or hearsay. We present a novel paradigm that integrates the network approach with concepts and methods from interpersonal perception research. We apply that paradigm to study the relative valence, consensus, accuracy, and consequences of personality trait information in hearsay-based reputations. In two preregistered studies (N = 260 and 369), we created naturalistic micronetworks in the lab in which participants interacted and got to know one another, then later described each other to naïve third parties. Across studies, hearsay-based reputations were less positive than direct reputations and self-perceptions. Hearsay reputations demonstrated appreciable consensus (agreement) with direct reputations but only modest accuracy, suggesting that they provide a robust but substantially inaccurate signal. Neither hearsay consensus nor accuracy were moderated by individual differences in extraversion, empathy, or perspective-taking. In Study 2 we experimentally manipulated participants’ impression-formation goals, but this had no detectable impact on hearsay consensus or accuracy. Hearsay reputations were consequential, affecting the extent to which perceivers thought targets would be good leaders or friends. These results provide initial insights into reputation networks and suggest several important future directions for the network approach to reputations. We also present open materials and data analysis software for others to extend the reputations-as-networks approach.
In a large, nationally-representitive sample, we find mean-level change toward more inclusive value priorities and high rank-order stability of values over time.