Concurrent Generation of Binary, Ordinal, and Count Data with Specified Marginal and Associational Quantities in Pharmaceutical Sciences
Abstract
This manuscript is concerned with establishing a unified framework for concurrently generating data
sets that include three major kinds of variables (i.e., binary, ordinal, and count) when the marginal distributions and
a feasible association structure are specified for simulation purposes. The simulation paradigm has been commonly
utilized in pharmaceutical practice. A central aspect of every simulation study is the quantification of the model
components and parameters that jointly define a scientific process. When this quantification goes beyond the
deterministic tools, researchers often resort to random number generation (RNG) in finding simulation-based
solutions to address the stochastic nature of the problem. Although many RNG algorithms have appeared in the
literature, a major limitation is that most of them were not devised to simultaneously accommodate all variable types
mentioned above. Thus, these algorithms provide only an incomplete solution, as real data sets include variables of
different kinds. This work represents an important augmentation of the existing methods as it is a systematic attempt
and comprehensive investigation for mixed data generation. We provide an algorithm that is designed for generating
data of mixed marginals; illustrate its operational, logistical, and computational details; and present ideas on how it
can be extended to span more sophisticated distributional settings in terms of a broader range of marginal features
and associational quantities.
Keywords
Biserial correlation, phi coefficient, simulation, tetrachoric correlation, random number generation, mixed data
References
- [1] Demirtas H. A method for multivariate ordinal data generation given marginal distributions and correlations. Journal of Statistical Computation and Simulation, 2006; 76: 1017- 1025.
- [2] Demirtas H, Doganay B. Simultaneous generation of binary and normal data with specified marginal and association structures. Journal of Biopharmaceutical Statistics, 2012; 22: 223-236.
- [3] Demirtas H, Yavuz Y. Concurrent generation of ordinal and normal data. Journal of Biopharmaceutical Statistics, 2015; 25: 635-650.
- [4] Amatya A, Demirtas H. Simultaneous generation of multivariate mixed data with Poisson and normal marginals. Journal of Statistical Computation and Simulation, 2015a; 85: 3129-3139.
- [5] Emrich JL, Piedmonte MR. A method for generating high-dimensional multivariate binary variates. The American Statistician, 1991; 45: 302—304.
- [6] Demirtas H, Hedeker D. A practical way for computing approximate lower and upper correlation bounds. The American Statistician, 2011; 65: 104-109.
- [7] Demirtas H, Hedeker D. Computing the point-biserial correlation under any underlying continuous distribution. Communications in Statistics-Simulation and Computation, 2016; Anat. J. Pharm. Sci 2022; 45: 2744-2751.
- [8] Demirtas H, Ahmadian R, Atis S, Can FE, Ercan I. A nonnormal look at polychoric correlations: Modeling the change in correlations before and after discretization. Computational Statistics, 2016; 31: 1385-1401.
- [9] Ferrari PA, Barbiero A. Simulating ordinal data. Multivariate Behavioral Research, 2012; 47: 566-589.
- [10] Yahav I, Shmueli G. On generating multivariate Poisson data in management science applications. Applied Stochastic Models in Business and Industry, 2012; 28: 91—102.