I completed my PhD in Statistics with a specialization in Biostatistics at the University of Waterloo in 2010, and I joined the University of Saskatchewan that same year. My research lies broadly in Biostatistics, with particular focus on survival analysis, recurrent event modeling, joint modeling of longitudinal and time‑to‑event data, changepoint methods, and computational statistics. I develop advanced statistical methodology grounded in Bayesian inference and MCMC computation, motivated by applications in health, environmental, and interdisciplinary scientific research. For full details on my research contributions, supervision record, teaching, and service, please refer to my CV.

Selected Publications (Methodology)

• Rana M, Khan SA, and Pahwa P (2026). Covariate Measurement Error in Spatial Analysis of Count Data. Journal of Statistical Computation and Simulation, (accepted).
• Khan SA and Hossain S (2025). Segmented Bent-Cable Regression Model for Changepoint Data Analysis. Sankhya B, 87:  759-790.
• Forzley Q, Hossain S, and Khan SA (2023). Generalized Log-Logistic Proportional Hazard Model: A Non-Penalty Shrinkage Approach. Statistics, 57(6): 1511-1528.
• Khan SA and Basharat N (2021). Accelerated Failure Time Models for Recurrent Event Data Analysis and Joint Modeling. Computational Statistics, 37: 1569-1597.
• Hossain S and Khan SA (2020). Shrinkage Estimation of the Exponentiated Weibull Regression Model for Time-to-Event Data. Statistica Neerlandica, 74(4): 592-610.
• Khan SA (2018). Exponentiated Weibull Regression for Time-to-Event Data. Lifetime Data Analysis, 24(2): 328-354.

Selected Publications (Applied / Interdisciplinary)

• Krahn J, Hossain S, and Khan SA (2022). An Efficient Estimation Approach to Joint Modelling of Longitudinal and Survival Data. Journal of Applied Statistics, 50(15): 3031-3047.
• Khan SA and Kar SC (2018). Generalized Bent-Cable Methodology for Changepoint Data: A Bayesian Approach. Journal of Applied Statistics, 45(10): 1799-1812.
• Khan SA, Chiu GS, and Dubin JA (2013). Therapeutic Hypothermia: Quantification of the Transition of Core Body Temperature Using the Flexible Bent-Cable Model for Mixture Longitudinal Data. Australian and New Zealand Journal of Statistics, 55(4): 369-385.
• Chiu GS, Guttorp P, Westveld AH, Khan SA, and Liang J (2011). Latent Health Factor Index: A Statistical Modeling Approach for Ecological Health Assessment. Environmetrics, 22(3): 243-255.

Full publication list, software, talks, and supervision details are provided in the CV.

My teaching is guided by two central goals: to provide students with the knowledge they need for future study and professional success, and to motivate them by showing how statistics solves real‑world problems across many disciplines. I strive to make material clear and accessible, use application‑oriented examples, and cultivate an engaging classroom environment through discussion, curiosity‑driven questions, and open communication.

At the University of Saskatchewan, I have taught a wide range of undergraduate and graduate courses, including STAT 851 (Linear Models), STAT 850 (Mathematical Statistics and Inference), STAT 846 (Survival Analysis), STAT 845 (Statistical Methods for Research), STAT 834 (Advanced Experimental Design), STAT 812 (Computational Statistics), STAT 443 (Linear Statistical Models), STAT 348 (Sampling Techniques), STAT 345 (Design and Analysis of Experiments), STAT 246 (Introduction to Biostatistics), STAT 245 (Introduction to Statistical Methods), STAT 242 (Statistical Theory and Methodology), STAT 241 (Probability Theory), and STAT 103 (Elementary Probability).

My approach adapts naturally to different levels of instruction: undergraduate teaching emphasizes building intuition and motivation, while graduate teaching focuses on preparing students for scientific research and independent inquiry. Across all levels, my aim is to help students develop confidence, analytical skill, and an appreciation for statistics as a powerful interdisciplinary science.

I have extensive experience supervising students from both the Statistics and Biostatistics graduate programs, many of whom have successfully completed their degrees under my supervision or co‑supervision and moved on to strong academic, governmental, and industry careers.

I welcome applications from motivated MSc and PhD students broadly interested in Statistics/Biostatistics, including survival analysis, recurrent event modeling, joint modeling of longitudinal and time‑to‑event data, changepoint methods, and computational statistics. Projects typically involve Bayesian inference and MCMC and often draw on real datasets from health and environmental sciences.

If you are interested in pursuing graduate research with me, please email a brief statement of research interests, a CV, and unofficial transcripts. For general information about our graduate programs and the application process, please consult the graduate studies page of our department (for Statistics) and the Biostatistics program page (for Biostatistics).

Bayesian & MCMC methods Biostatistics Changepoint analysis Computational statistics Joint modeling Survival & event‑history analysis

My research lies broadly in Biostatistics, with a main focus on developing advanced statistical methodology for survival analysis, recurrent event modeling, and joint modeling of longitudinal and time‑to‑event data. I have advanced the use of accelerated failure time frameworks in recurrent event and joint modeling contexts, and work extensively with mixed‑effects models that link longitudinal and survival processes through shared random effects.

A major area of my work develops flexible parametric survival models, including generalizations of the exponentiated Weibull and log‑logistic families, enabling richer hazard structures and more accurate inference. I also investigate shrinkage estimation methods that improve estimation accuracy by reducing bias and enhancing efficiency, particularly within survival and joint modeling frameworks. My research is grounded in Bayesian inference and MCMC computation, which provide a principled foundation for estimation and uncertainty quantification in complex models.

I have developed highly flexible and interpretable changepoint models, including Bayesian bent‑cable formulations capable of representing non‑linear trends involving both gradual and abrupt transitions in time‑series data. These models offer clear regression‑coefficient interpretations and adapt well to diverse scientific applications.

In computational statistics, I am developing the R package JMR, designed to support the analysis of survival data — including standard survival models, recurrent event models, and joint models — along with residual diagnostics and dynamic prediction tools. Most components are implemented within a Bayesian MCMC framework, and the package is currently in progress, with a preliminary version available on GitHub.

I also maintain strong interests in applied and environmental statistics, including the analysis of time‑series data on air pollution and associated health outcomes, which originally motivated several strands of my methodological work.

Curriculum Vitae (PDF)