About

John Lee is an Associate Dean for Faculty and Academic Affairs at the College of Education at NC State University. His role involves overseeing faculty and academic programs within the college. The page does not provide specific details about his research focus, background, or key contributions.

Research topics

• Computer Science
• Sociology
• Psychology
• Engineering
• Mathematics education
• Pedagogy
• Engineering ethics
• Knowledge management
• Management science
• Process management

Selected publications

• Predicting Postoperative Complications After Glossectomy And Tongue Reconstruction With Machine Learning

  Zenodo (CERN European Organization for Nuclear Research) · 2026-05-08

  articleOpen access1st authorCorresponding

  PURPOSE: Glossectomy, which encompasses procedures of partial to total tongue resection, is a critical intervention for advanced oral cavity and oropharyngeal cancers. Despite its importance, there is currently limited understanding regarding complication rates and risk predictors following glossectomy and tongue reconstruction. Leveraging machine learning, we developed robust models to predict post-operative complications for glossectomy and tongue reconstruction, aiming to characterize patient risk profiles to guide future peri-operative planning. METHODS: We developed 5 machine learning models with the following algorithms: stacked ensemble (Stack), neural network (NN), light global boosting machine (LightGBM), support vector classifiers (SVC), and tuned logistic regression (LR), to predict the following four primary post-operative complications: mortality, bleeding-related, aspiration-related (pneumonia, reintubation, ventilator dependance), and surgical (dehiscence and infection) complications. To train our models, we analyzed a cohort of 7,485 tongue cancer patients receiving glossectomy with or without immediate reconstruction from the 2008-2020 American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database. Model performance was comprehensively evaluated using AUROC, accuracy, precision, recall, and F1-score, with confusion matrices providing insight into classification performance analysis. Calibration was assessed with the integrated calibration index and Brier scores. SHAP analysis was performed to identify the most influential features driving model prediction. RESULTS: In our cohort of 7,485 patients, the mean age was 62.0 ± 13.3 years with a mean BMI of 28.0 ± 8.2. Elective procedures accounted for 92.9% of our cohort and 74.1% of patients underwent a partial glossectomy. For aspiration-related complications, LR achieved the highest test AUROC (0.782, 95% CI 0.730-0.828), while LightGBM demonstrated a slighter worse AUROC (0.780, 95% CI 0.730-0.830) but superior recall and precision (0.751 and 0.550 vs 0.681 and 0.544). Mortality prediction was more challenging, with LR demonstrating the highest AUROC (0.723 95% CI 0.531-0.893), recall (0.687), and precision (0.505). LightGBM heavily underperformed in predicting mortality, with a testing AUROC of 0.155. For bleeding-related complications, LR once again attained the highest AUROC (0.905, 95% CI 0.871-0.929). However, LightGBM (AUROC 0.896, 95% CI 0.863-0.922), had the highest F1-score, accuracy, and precision compared to other models. For surgical complications, all models showed similar performance (test AUROC range 0.713-0.772), with LightGBM and Stack emerging as the models with the best sensitivity and specificity. Age, BMI, ASA class, reconstruction type, glossectomy type, concurrent procedures, and tumor location were determined as significant predictors of postoperative complications according to SHAP analysis. CONCLUSIONS: Machine learning models, particularly tuned logistic regression and LightGBM, offer robust performance for predicting post-operative complications following glossectomy and reconstruction. While LR consistently demonstrated the highest discrimination, LightGBM often provided superior sensitivity and precision for certain outcomes. Though external validation is warranted, greater adoption of artificial intelligence in predicting complications, as demonstrated with the development of these models, may aid in risk stratification and peri-operative planning to improve future glossectomy and reconstructive outcomes. *Source: https://ps-rc.org/meeting/Program/2026/CS73.cgi*

  PublisherDOI

• 778 CRISPR screens reveal innate immune checkpoints regulating NK cell sensitivity across solid tumors

  Regular and Young Investigator Award Abstracts · 2025-11-01

  articleOpen access
  PublisherOA PDFDOI

• Ongoing Formative Evaluation and Quality Improvement in an Interprofessional Family Support Program

  OTJR Occupational Therapy Journal of Research · 2025-12-19

  article

  The Raising Families Project delivered three, 12-week cohorts of interprofessional services alongside graduate student training, where weekly evaluation surveys provided formative data. This article examines quality improvement via ongoing formative program evaluation. The purpose of the current study was to gather and analyze Raising Families Project participant feedback relating to program delivery logistics, to improve ongoing and future program delivery and quality. A total of 347 weekly evaluation surveys were collected from caregivers ( n = 37), students ( n = 35), and clinicians ( n = 7). Descriptive statistics and reflexive thematic analysis were utilized to analyze survey data. Five themes were developed related to quality improvement, namely immediate vs. sustained adjustments, logistical enhancements, challenging flexibility, collaborative benefits/varied meaning, revealing complexity of participants’ preferences, alongside iterative improvements resulting from feedback. Findings suggest the need to include formative evaluation, to embed the caregiver/family in interprofessional approaches, and to consider individuals’ needs in practice.

  PublisherDOI

• Empowering Families Through Interprofessional Collaboration: A Program Evaluation of the Raising Families Project

  American Journal of Occupational Therapy · 2025-07-30

  article

  Abstract Date Presented 04/03/2025 An interdisciplinary faculty team piloted Raising Families, a postpandemic, sustainable caregiver-and-young-child model of family support. This study explored perceptions of caregivers, students, and clinicians who collaborated in program delivery. Primary Author and Speaker: Zahava Friedman Contributing Authors: Miara Joy Mandap, Brianna Somma, Sabrina Kenny, Jessica Latawiec, John Lee, Kate Nealon, Keri Giordano, Kelly Sullivan-Jones

  PublisherDOI

• BRIDGING SPECIALTIES: TIMELY DIAGNOSIS FOR INTERSTITIAL LUNG DISEASE IN THE STATE OF WEST VIRGINIA: A QUALITY IMPROVEMENT PROJECT

  CHEST Journal · 2024-09-18

  articleOpen access1st authorCorresponding
  PublisherOA PDFDOI

• Digital Citizenship, Values and Cultural Dynamism

  International Journal on Research in STEM Education · 2024-05-05 · 1 citations

  articleOpen access

  This paper examines the global shift towards digital citizenship triggered by COVID-19 and its role in mediating cultural tensions in a rapidly digitizing world. Utilizing mixed methods, the study draws from two projects: the first assesses the engagement of 315 Australian adolescents with values in their science education, and the second investigates digital citizenship practices among 303 university faculty members in Saudi Arabia. The findings highlight significant sociocultural differences in digital engagement and underscore the varying impacts of digital globalization across different educational and national contexts. The paper argues for a proactive educational strategy that encourages critical engagement with digital tools to navigate and reconcile these cultural dynamics effectively. By exploring the interactions between digital technology providers, users, and regulatory bodies, the study provides insights into the complexities of digital responsibility and the potential of education to foster a balanced digital citizenship. This approach suggests moving beyond mere technological integration to embrace a pedagogy that is responsive to the ethical challenges posed by global digital interactions.

  PublisherOA PDFDOI

• Using Inquiry to Teach About Religion in Middle and High School Classrooms

  2023-09-20

  book-chapter
  PublisherDOI

• Introduction

  2022-07-25

  book-chapterSenior author
  PublisherDOI

• Inquiry-Based Practice in Social Studies Education

  2022 · 18 citations

  Senior authorCorresponding
  • Sociology
  • Sociology
  • Pedagogy

  Now in its second edition, Inquiry-Based Practice in Social Studies Education: Understanding the Inquiry Design Model presents a conceptual base for shaping the classroom experience through inquiry-based teaching and learning. Using their Inquiry Design Model (IDM), the authors present a field-tested approach for ambitious social studies teaching. They do so by providing a detailed account of inquiry’s scholarly roots, as well as the rationale for viewing questions, tasks, and sources as inquiry’s foundational elements. Based on work done with classroom teachers, university faculty, and state education department personnel, this book encourages readers to transform classrooms into places where inquiry thrives as everyday practice. The second edition includes a new chapter highlighting three ways that the blueprint acts as an assessment and curriculum system, and includes updated and enhanced references throughout the book. Both pre-service and in-service teachers are sure to learn strategies for developing the reinforcing elements of IDM, from planning inquiries to communicating conclusions and taking informed action. The updated curricular and pedagogical examples included make this practical book essential reading for researchers, students of pre-service and in-service methods courses, and professional development programs.

  PublisherDOI

• Assessment and Inquiry

  2022-07-25

  book-chapterSenior author
  PublisherDOI

Frequent coauthors

• T. Brent Gunnoe

  University of Virginia

  56 shared

• Thomas R. Cundari

  University of North Texas

  55 shared

• Jeffrey L. Petersen

  West Virginia University

  51 shared

• Kathy Swan

  University of Kentucky

  32 shared

• Karl A. Pittard

  Norfolk State University

  25 shared

• David Hicks

  Aalborg University

  23 shared

• Zhuofeng Ke

  18 shared

• Paul D. Boyle

  16 shared

Labs

• John Lee LabPI