This page lists the most recent projects I am currently working on.
Office work is an important part of the American workplace across sectors, involving tens of millions of workers. Further, doing office work is not just about the processing of paperwork and information; it also involves organizational knowledge, communication and relationship building, and other interpersonal skills. These twin issues of job reliance and job complexity raise concerns for the design of artificial intelligence-based office automation tools that use software agents, machine learning, and data analytics. Many such tools are designed without workers’ needs or organizational complexity in mind; mistakes here can lead to reduced quality of life and job loss for workers, and worse outcomes for firms and customers. This planning project will develop a multi-disciplinary research agenda toward the careful scientific study of both office work and the design of automation that affects it, with the goal of improving both office automation and training and outcomes for workers.
The project involves three main activities. The first is a comprehensive literature review using a combination of the critical review method, the systematized review framework, and a review of state-of-the-art office automation systems. The second involves qualitative analysis of a series of focus groups conducted with a broad sample of office workers about their daily work practices and challenges, their professional development and training, and their perceptions about the future of their job and of office automation. The third main activity is a three-day ideation workshop that will build on the findings from the first two activities. The research team will convene multidisciplinary academic and industry experts from human factors, computer science and artificial intelligence, economics, labor and human resource management, public policy, and equity and diversity to identify domains of office work and research themes that look most interesting, promising, and impactful for future research.
This project aims to serve the national interest by building institutional capacity for broadening participation in undergraduate engineering. Colleges of Engineering in predominantly white institutions continue to struggle to recruit, retain and successfully matriculate their non-white students. Using a systems approach, this project aims to identify catalysts that can unify programmatic efforts to produce greater institutional success in broadening participation. These catalysts can exist throughout an organization, and can take the form of people, knowledge, money, policies, built environments, and other resources. A coherent set of programmatic efforts can lead to greater transformative power and sustainability. The project will generate templates that other colleges of engineering in the United States can use to discover and understand their own capacities and to further their own recruitment, retention and graduation efforts for students from underrepresented groups. The ultimate aim of the project is to enable the success of a diverse pool of engineering graduates, and to strengthen the diversity of the engineering workforce.
The goals of the project are to generate systemwide capacity change catalysts in a College of Engineering at a predominantly white institution. The project conceptualizes the college as a complex sociotechnical organization with two subsystems at work. The social subsystem consists of people such as students, faculty, staff, and administrators. The technical subsystem consists of elements that can impact capacity building, including goals, policies, processes, programs, data, technology, and know-how. The project uses semi-structured interviews and ideation focus groups with students, faculty, and staff from admissions, advising, and career services as well as administrative leaders. Results will be used to model the sociotechnical system, to develop system capacity matrices and social focal role networks, and to generate capacity change catalysts. The project contributes to identifying and understanding, in-depth, the systemwide catalysts in predominantly white engineering colleges that will enable these colleges to improve their capacity to recruit, retain and successfully matriculate underrepresented students. The project also contributes a new sociotechnical systems lens and lays out a template for using the lens to model an educational unit. The NSF IUSE: EHR Program supports research and development projects to improve the effectiveness of STEM education for all students. Through the Institutional and Community Transformation track, the program supports efforts to transform and improve STEM education across institutions of higher education and disciplinary communities.
The goal of this RAPID project is to model decision-making among nurses when they triage patients during the COVID-19 pandemic. Triage nurses act as frontline gatekeepers and perform a difficult balancing act during a pandemic. They must not only ensure that patients who need immediate care get it in a timely manner but must also filter incoming patients to prevent infections and to reduce undue burden on hospital resources. The complexity and risk in their decisions are influenced by how the nurses themselves perceive the risk of a pandemic, and how they associate and project their risk perception with the information a patient provides. Conventional triage decision making criteria, protocols and processes based only on a linear, discrete, “single-symptom at a time” risk screening approach are woefully inadequate to tackle triage decisions in a pandemic of this scale and complexity. Nurses play a pivotal role in ensuring safe and timely patient care and in limiting the spread of COVID-19-like pandemics. The knowledge gained from this project about decision making and evidence-based practices during crises will benefit organizations worldwide, by proving data on factors that make decision-making. These data can drive training and guidelines development.
Triage nurses routinely make triage decisions about patients. But, during a pandemic, they make particularly complex and risky decisions. Triage decision making criteria and protocols must reflect a deep understanding of how nurses weigh patient symptoms, and match them to disease conditions, while also managing a multitude of complex, interrelated decision constraints, including their own risk perceptions, and limited, uncertain, confounding information, in the midst of a pandemic with major safety consequences. To identify the constraints nurses face when making triaging decisions, and to model the strategies they use when triaging, the project studies triage nurses from two large academic medical centers. The project retrospectively analyzes triage phone calls for patient risk screening, prospectively records screens nurses use as information sources, and interviews nurses about their constraints, strategies, risk perception and cognitive workload. The project maps a nurse’s patient-specific decision-making trajectories to reveal their constraints and how they managed them. Cumulatively, the data is expected to reveal generalizable strategies for triage decisions during catastrophic events.