By Abby Weingarten
How is COVID-19 affecting human emotion? How are conversations about the pandemic unfolding on social media? New College alum Cory Costello wants to know.
“I’m looking at how people are regulating their emotions with respect to anxiety around COVID-19, and how people are talking about COVID-19 on Twitter,” Costello said. “The latter is with a specific eye towards how statements with generic ‘you’ (e.g., ‘if you get tested, you might not even get results for days’) might be functioning. The research is at a pretty early stage. We’re still in the process of collecting analyzing data.”
This is part of Costello’s ongoing research on the psychology of people’s emotions in online environments, as well as offline in daily life. In the spring, Costello earned a Ph.D. in psychology from the University of Oregon after defending his doctoral dissertation, Human and Computerized Personality Inferences from Digital Footprints on Twitter. He also holds a master’s degree in in psychology from Wake Forest University. But it was his undergraduate career at New College that first set the stage for his critical-thinking-based approach to his craft.
“I loved my time at New College, and it was where I gained my passion for research, science and especially statistics (thanks largely to Professor Duff Cooper),” said Costello, adding that his thesis was about the relation between anxious attachment, anxiety and motivation to drink alcohol, using an experimental procedure called priming. “It’s sort of interesting, looking back at how my experiences at New College affected the course of my career.”
As an academic, Costello found himself specifically interested in personality (who we are), interpersonal perception (what people think about us) and reputations (what people say and hear about us). He is fascinated by how these factors manifest in online environments, and the extent to which they are recoverable from digital footprints. One of the most impactful aspects of his time at New College was gaining this “perspective on science, research and learning,” he said.
“One example of this that always sticks out in my memory was from the first day of Professor Heidi Harley‘s ‘Cognitive Psychology’ seminar,” Costello recalled. “I remember her starting the class by saying something to the effect of, ‘In this course, we’re not going to focus on the answers. This is science, and the answers constantly change and evolve, and so there is little point to memorizing answers. Instead, we’re going to focus on the questions—how to ask the right questions in the right way—and the methods people employ to try to answer those questions.’ I think back on this advice a lot, and I think it had a profound impact on my graduate studies.”
Costello has spent most of his graduate career learning about methodology—how to measure psychological constructs with questionnaires, how to use downloadable data from the Internet (such as Tweets) to investigate psychological phenomena, how to design experiments with high realism that isolate mechanisms, and how to model data from various sources.
“It’s interesting because Professor Harley’s advice ended up being even more useful in this era of psychological science. That class took place in 2010, and the following year is often marked as the beginning of the ‘replication crisis’ in psychology, or the growing realization that standard methods in psychology were leading to unacceptable false positive rates,” Costello said. “This was a serious blow to the field and to a lot of people’s morale. But the perspective on science I gained at New College prepared me for this crisis. I just reminded myself of what Professor Harley said on the first day of ‘Cognitive Psychology,’ that the answers were always bound to change anyway, and now we (as a field) were collectively working on developing better means to answer the questions we care about.”
Costello continues to develop those means of answering questions. Kathleen Casto, Ph.D., an assistant professor of Psychology, Biopsychology and Neuroscience at New College, said Costello’s work is cutting-edge for multiple reasons.
“Not only does Cory implement complex data analysis techniques for modeling social perceptions and relations between people, but he also uses highly innovative methods such as using machine learning to extract digital footprints from online social networks,” Casto said. “Perhaps most importantly, Cory’s work is a prime example of the ways to improve the integrity and availability of science for all.”
Casto explained that Costello does this by openly sharing his data, as well as his data analysis software. He engages in a process called “pre-registration,” in which study hypotheses are publicly time-stamped and recorded in advance of collecting and analyzing data (this helps remove researcher bias and increase the replicability of study findings).
“Cory’s work is really a model to us all of how to do better science, and how to use creative implementations of data science to study human social behavior and mental health in a modern world,” Casto said.
New College laid the foundation for this work.
In Costello’s words: “I could go on and on about New College, how much I loved my time there, and how much it affected my life and career.”
Abby Weingarten is the editor/writer in the Office of Communications & Marketing.