LOS ANGELES, March 2, 2020 /PRNewswire/ -- Six leading researchers from different departments in the USC Marshall School of Business will present their latest research on a range of timely business topics at the school's annual research fair, entitled "Research Visions," on Friday, March 6. Summaries of the research papers are included below.
WHAT: USC Marshall "Research Visions" Research Fair
WHEN: Friday, March 6, presentations from Noon to 2:30 pm
WHERE: USC Hotel, Grand Ballroom; the event also will be available online via livestream video, beginning at 11:50 am here: https://livestream.com/accounts/11826764/events/9016801.
Videos of previous Research Fair events are also available online here: https://www.marshall.usc.edu/faculty-research/research-fair-videos
Tom Chang, Associate Professor of Finance and Business Economics, "Environment Economics":
- The physical environment has a much bigger impact on factors like productivity and cognitive function than most people realize. For example, air quality has been shown to affect everything from crime levels to test scores to how accurately MLB umpires call balls and strikes. The implications of this research for individual, firm, and economy wide productivity are considerable. In this talk, Professor Chang will summarize some of the key findings of this literature with a special focus on the environment in and around Hoffman Hall 504 – his campus office.
Anthony Dukes, Chair and Professor of Marketing, "When the Marketer Knows more about the Consumer than the Consumer":
- Can the marketer know more about the consumer than the consumer herself? And, if so, can this informational advantage be exploited? These questions are on many people's minds due to the rapid expansion of consumer data available to marketers. Professor Dukes' research explores the economic underpinnings of these situations to understand the importance of consumer suspicions in keeping companies "in check" so that they do not exploit their informational advantage.
Yingying Fan, Professor of Data Sciences and Operations, "Reproducible and Interpretable Inference with Big Data":
- Many contemporary large-scale applications with big data involve building complicated statistical models such as deep neural networks (DNNs). These models are usually difficult to interpret and produce results that are irreproducible. Selecting important features among a large number of available ones is a popularly used and effective tool for improving the interpretability and reproducibility of statistical inference results. To achieve feature selection in arbitrarily complicated high-dimensional models with controlled error rate, a new framework was recently proposed for model-X knockoffs in Candes, Fan, Janson and Lv (2018). In this talk, Professor Fan will give a high-level overview of the model-X knockoffs framework and discuss a new DNN architecture that we proposed for integrating the model-X knockoffs framework with deep learning models. The talk will close with an economic application.
Nathanael Fast, Associate Professor of Management and Organization, "The Psychology of Artificial Intelligence Adoption":
- The use of artificial intelligence to track employee behavior and direct human resource decisions is spreading rapidly, but little is known about how employees will respond to these new practices. Workers have historically been averse to tracking and monitoring, suggesting that they may reject AI-based technologies that perform these functions. On the other hand, people frequently complain about bias in HR decisions about hiring, pay, and promotions, suggesting that replacing biased human decision-making with AI may be a welcomed change. The present talk examines these two practices – behavior tracking and decision-making – separately, showing that a different psychology emerges for each. In particular, employees are more likely to welcome behavior tracking when there is no human involvement, because it eliminates concern about negative judgment. However, in direct contrast, employees view HR-algorithms as unfair, based on the perception that AI cannot consider context and engage in holistic reasoning. The talk will examine the implications of these two distinct tendencies and conclude with a discussion of how to effectively and fairly introduce AI-based practices into the workplace.
Christina Lubinski, Professor of Clinical Entrepreneurship, "Entrepreneurship Meets Nationalism: A Historical Exploration":
- Rising economic nationalism can create political and economic opportunities as well as threats for multinational firms. While most firms engage in "corporate diplomacy" to leverage political developments at home and abroad, few will openly admit to it, making the topic extremely difficult to study. Through a historical analysis of selected foreign multinationals in India from the 1880s to the 1970s, this talk conceptualizes "geopolitical jockeying", i.e. strategies to delegitimize rival multinationals based on politics, as well as "aspirational political practices," i.e. strategies that address the inherently future‐oriented aspects of nations. Based on a database collected from over a dozen different archives in India, Germany, US and UK, the discussion explores the set of political capabilities necessary to engage in these strategies and their impact. The talk uses history as a mirror for reflecting on the causes and consequences of economic nationalism for international strategy in our own time.
Richard Sloan, Accounting Circle Professor of Accounting, "Generally Unaccepted Accounting Principles":
- This talk summarizes an ongoing body of research illustrating how US Generally Accepted Accounting Principles (GAAP) are becoming irrelevant and obsolete. Firms are increasingly promulgating their own non-GAAP earnings numbers in order to overcome perceived shortcomings of GAAP. But the rise of non-GAAP earnings numbers is introducing new problems, particularly their susceptibility to manipulation by unscrupulous managers. This talk explains how we arrived at the current situation, provides examples and discusses possible solutions.
For additional information, or to speak with a USC Marshall researcher, please contact Matthew Simmons, USC Marshall Media Relations, at [email protected] or at 213-821-9868.
SOURCE USC Marshall School of Business