Professor Jong Sung Kim received his Ph.D. in Statistics through the Department of Statistics and Actuarial Science, University of Iowa in 1999. He joined the faculty at PSU that same year. Professor Kim has made many contributions to the department including being instrumental in changing the department's name from Mathematical Science to Mathematics and Statistics, and participating in launching a new Ph.D. program in Mathematical Science, which includes Statistics, and a new MS program in Statistics.
As a teacher, Professor Kim is dedicated to the success of his students, as well as to the maintenance and improvement of the statistics program. He teaches courses from the students' point of view and is well known for his patience, encouragement and thoroughness. Professor Kim is tireless in his attention to student success. He assigns projects to all of his graduate courses, which is unusual in that it requires a tremendous amount of additional work for an instructor. Professor Kim, however, believes his effort results in continued success in maintaining and improving the current statistics program. In fact, a number of course projects have a strong potential for Master's theses or projects. In both his undergraduate and graduate courses he thoroughly explains concepts and examples, as evidenced by numerous students' evaluations. The key to this success is that he teaches his students the difference between random and fixed quantities as he believes that is the most fundamental concept for statistics. Students who fail to see the difference often fail to perform well in statistics programs. Lastly, all of his graduate students have secured jobs before graduation due to his careful training and presentation of every possible opportunity.
Dr. Kim has made important contributions in the areas of semiparametric inference and survival analysis in statistics, which has earned him national and international recognition. One significant contribution in those areas is his work on "Maximum likelihood estimation for the proportional hazards model with partly interval-censored data" (Journal of the Royal Statistical Society, B, 2003). This work has many important applications in biostatistics (such as in the analysis of survival data in cancer clinical trials) in which the observation times to an event of interest are often subject to right censoring, left censoring, or interval censoring. Rigorous theoretical investigation of this problem has not been done in the literature, although the proportional hazards model for the right-censored data has been studied extensively. A major difficulty is that Cox's partial likelihood approach for right-censored data, which does not involve the unknown baseline hazard, cannot be easily applied in the setting of arbitrarily censored data. Dr. Kim considered the maximum likelihood estimation approach. He proved the asymptotic distributional results of the maximum likelihood estimators of the regression parameter and the cumulative baseline hazard function. Theoretically, this problem is in the scope of semiparametric inference, because it involves both a finite-dimensional parameter (the regression coefficient) and an infinite-dimensional parameter (the baseline hazard). This is the first paper that provides theoretical justification for likelihood-based inference in this high-dimensional semiparametric model. Dr. Kim also generalized the missing information principle for finite-dimensional model to the semiparametric Cox model in estimating the standard errors of the maximum likelihood estimators. Dr. Kim extended this line of work to handle the truncated data in "Efficient estimation for the proportional hazards model with left-truncated and 'case 1' interval-censored data" (Statistica Sinica, 2003).
Other important publications include: a) "Introduction to SAS and Selected Textbook Examples by SAS Code" for Survival Analysis Using S: Analysis of Time-to-Event Data, with Mara Tableman, Chapman & Hall/CRC (2004); b) "Survival Analysis Using S: Analysis of Time to Event Data", Texts in Statistical Science, with Mara Tableman, Boca Raton: Chapman & Hall/CRC (2004); c) Contributed a chapter solution to "Solution Manual for Survival Analysis Using S: Analysis of Time to Event Data," with Mara Tableman, prepared by Peter Sparks, a former graduate student in Master of Science in Statistics program, (Fall 2003).
Dr. Kim has secured several research grants: a) Co-investigator: NIH (National Institute of Health) grants "Trichloroethylene Exposure Assessment," $159,640 for 2003 – 2005. (PI: Jan Semenza); b) Co-investigator (statistical consultant): National Institute of Justice (NIJ) grant "Trajectories of Violent Offending and Risk Status across Adolescence and Early Adulthood" with James Nash, PI, $35,000.00 for 2004 – 2005; c) Statistical consultant for Longitudinal Study of Adult Literacy (LASL) at PSU, PI: Steve Reder; d) PI for Faculty Enhancement grant "Censored Q-Q Plot: A Tool for Checking Population Heteroscedasticity," $4,700 for 2005 – 2006; e) Professional Travel Grant for 2005 WNAR-International Biometric Society/Institute of Mathematical Statistics Meeting, $850 for 2005.
Current work includes longitudinal data modeling and analysis for identifying distinct trajectories of aggressive behavior and risk over time for adolescents and young adults, identifying factors that have significant effects on adult literacy, and investigating whether countries can affect the timing of International Monetary Fund (IMF) program participation through their macroeconomic policies during the inter-program years (years without any IMF programs); development of asymptotic theory and methods required for analysis for a number of interval-censored medical and reliability data; creating a new statistical module for such data with his Ph.D. program student; and creating easy-to-see graphics-oriented software for right-censored data.