Photo of Demirtas, Hakan

Hakan Demirtas, PhD

Associate Professor

Epidemiology and Biostatistics


Building & Room:



1603 W. Taylor St.

Office Phone:

(312) 996-9841

CV Download:


Selected Publications

Demirtas, H. & Schafer, J. L. (2003). On the performance of random-coefficient
pattern-mixture models for non-ignorable drop-out. Statistics in Medicine, Volume 22,
Issue 16, 2553-2575.


Demirtas, H. (2005). Multiple imputation under Bayesianly smoothed pattern-mixture
models for non-ignorable drop-out. Statistics in Medicine, Volume 24, Issue 15, 2345-


Demirtas, H. & Hedeker, D. (2007). Gaussianization-based quasi-imputation and
expansion strategies for incomplete correlated binary responses. Statistics in Medicine,
Volume 26, Issue 4, 782-799.

Demirtas, H. & Hedeker, D. (2008). An imputation strategy for incomplete
longitudinal ordinal data. Statistics in Medicine, Volume 27, Issue 20, 4086-4093.

Demirtas, H. & Hedeker, D. (2011). A practical way for computing approximate
lower and upper correlation bounds. American Statistician, Volume 65, Issue 2, 104-109.

Demirtas, H., Hedeker, D. & Mermelstein, R. J. (2012). Simulation of massive public
health data by power polynomials. Statistics in Medicine, Volume 31, Issue 27, 3337-

Demirtas, H. (2016). A note on the relationship between the phi coefficient and the
tetrachoric correlation under nonnormal underlying distributions. American Statistician,
Volume 70, Issue 2, 143-148.

Demirtas, H., Ahmadian, R., Atis, S., Can, F. E. & Ercan, I. (2016). A nonnormal look
at polychoric correlations: Modeling the change in correlations before and after
discretization. Computational Statistics, Volume 31, Issue 4, 1385-1401.

Demirtas, H. & Vardar-Acar, C. (2017). Anatomy of correlational magnitude
transformations in latency and discretization contexts in Monte-Carlo studies (pp. 59-84).
In ICSA Book Series in Statistics, John Dean Chen and Ding-Geng (Din) Chen (Eds):
Monte-Carlo Simulation-Based Statistical Modeling. Singapore: Springer.

Demirtas, H. (2019). Inducing any feasible level of correlation to bivariate data with
any marginals. Forthcoming in American Statistician.
DOI: 10.1080/00031305.2017.1379438