51ÉçÇøºÚÁÏ

Collaborators: Angelica Lim (51ÉçÇøºÚÁÏComputing Science), Bonnie Lau (University of Washington), Max Donelan (51ÉçÇøºÚÁÏBiomedical Physiology & Kinesiology), and Yue Wang (51ÉçÇøºÚÁÏLinguistics). 

Many people interact with machines through vocal interfaces like smartphones or smart speakers. These systems increasingly mimic human conversational patterns through a process called entrainment, which can improve user experience. This research explores how infants respond to artificial voices, focusing on phonetic convergence—a form of speech alignment seen in adults when they interact with vocal interfaces.

The study has three goals:

1. Develop a method using facial electromyography (EMG) to measure infants’ sympathetic speech behaviors; 
2. Test whether machine-like voices hinder or delay phonetic convergence; and 
3. Examine if voices resembling caregivers enhance or accelerate this behavior. 

Findings will offer new tools for studying early language development and inform the design of infant-directed technologies. As digital media use among infants grows, this research will encourage the development of more effective and socially responsive vocal interfaces for developmental populations.

This work aims to investigate whether revealing the intricate and connected nature of shared complex systems – defined here as systems that are more complex than the sum of their parts, such as the human brain, a living cell, or power grid – can evoke feelings of awe to promote prosociality. This research will provide novel insight into human prosociality, a topic at the core of human morality. In addition, this research will deepen our understanding of awe in psychology by examining whether some experiences – shared or common to many people– are more effective at promoting awe than solitary experiences of awe. Ultimately, this work will provide insight into new sources of awe and everyday prosociality, helping people, charities and policy makers understand daily positive experiences as well as rare occasions that inspire people to engage in prosocial action.

Teachers play a vital role in the education system, but there is considerable debate about how to enhance teaching quality. One natural approach grounded in economic principles is pay-for-performance — offering financial rewards to teachers based on the improvement in their students' test scores. Many school districts have implemented performance incentives, but evaluations of such programs have yielded mixed conclusions — some programs help, but others do not. One possible reason for the conflicting results is that only certain teachers are motivated by extra pay.

This project explores the possibility that only some teachers respond to monetary incentives by investigating both the introduction and termination of a pay-for-performance program called Mission Possible that ran in North Carolina from 2006 to 2017. If teachers who choose to work in schools with incentive programs are exactly those teachers who respond to monetary incentives, then we expect there to be a dramatic asymmetry between the program’s introduction and termination. Specifically, we would expect little effect when the program is introduced as the majority of teachers are not responsive to incentives. Over time, however, incentive-responsive teachers sort into schools offering the incentives, causing a large drop in test scores upon the program's termination. This research will thus uncover the impact of pay-for-performance on student test scores and shed light on whether improvements are driven by existing teachers responding to incentives or by the gradual movement of incentive-responsive teachers into participating schools.

Professor Richard Frank’s project aims to explore the present challenge of deepfake technology. This research will investigate how effectively people can identify celebrity deepfakes—highly realistic, AI-generated videos that depict public figures saying or doing things they never actually did.

Supported by a talented research team including PhD student Emmet Robins, MA student Maneh Rostomyan, and computer science undergrads Jyotir Mayor and Ritika Goyal, the project seeks to understand who is good at identifying deepfakes, who is not, and why. Additionally, they will explore how confident individuals feel about their judgments and whether short training programs or financial incentives improve detection skills. 

As part of the project, the team plans to create a public-facing website where people can test their skills, explore examples of deepfakes, and learn how to spot red flags in deepfake scams.

Decarbonizing Canada’s economy to limit the impacts of climate change represents an enormous societal challenge. Understanding how the complex mix of climate policies employed impacts the cost of living for Canadian households represents a pressing policy issue for ensuring a sustainable and equitable energy transition.

Using spatial micro-simulation techniques, I aim to evaluate the distributional impacts of alternative decarbonization approaches across income levels, sectors, and regions in Canada. This approach involves using household consumption data from the Survey of Household Spending, assessing the emissions intensity of consumed goods and services across the country, estimating policy-induced price changes under alternative policy combinations, and simulating demand changes in response to these prices.

This research will generate evidence to help inform climate policy design and implementation at both the federal and provincial levels in pursuit of equitable, sustainable, and cost-effective decarbonization pathways and address pressing concerns about inequality and energy affordability in Canada.