Almost all designs used in practice are constructed as if there is only one objective in the study. Rising experimental costs have led researchers to use efficient designs that can more adequately meet their practical goals and concerns. In particular, designs should be able to capture the multiple goals embedded in most studies, the possibly unequal importance of these goals and other constraints, such as model uncertainty, a restricted design space and potentially missing trials. We will construct multiple-objective optimal designs that incorporate these and other practical constraints. We also work with environmental health scientists and apply optimal design ideas to design a study for estimating the threshold of a joint compound in a teratology study using information in the dose response curves from the individual agents and the joint compound. Additionally, we develop cost-saving designs for Michaelks-Menten like mod sis used for studying pulmonary clearance of insoluble particles and find optimal designs for estimating biological responses defined by solutions of a differential equation. Innovative theory will be developed to construct these designs and understand their properties, including their sensitivities to model assumptions. ? ? To facilitate the use of optimal design ideas in practice, we create a user-friendly website with an array of algorithms for generating all types of designs for linear and nonlinear models, including designs to be developed in this research. The website will require input from the user before a tailored made design can be generated. Practitioners will be able to use the website to evaluate the efficiency of a design under a range of user-selected criteria and search for links to other websites where specialized design software is available. Such a website does not exist now. Our website will educate, inform and encourage practitioners to incorporate more optimal design ideas in their work, thereby saving substantial cost in their studies. ? ?

Agency
National Institute of Health (NIH)
Institute
National Institute of General Medical Sciences (NIGMS)
Type
Research Project (R01)
Project #
5R01GM072876-02
Application #
7113670
Study Section
Biostatistical Methods and Research Design Study Section (BMRD)
Program Officer
Okita, Richard T
Project Start
2005-09-01
Project End
2008-08-31
Budget Start
2006-09-01
Budget End
2007-08-31
Support Year
2
Fiscal Year
2006
Total Cost
$215,950
Indirect Cost
Name
University of California Los Angeles
Department
Biostatistics & Other Math Sci
Type
Schools of Public Health
DUNS #
092530369
City
Los Angeles
State
CA
Country
United States
Zip Code
90095
Wong, Weng Kee (2013) Web-based tools for finding optimal designs in biomedical studies. Comput Methods Programs Biomed 111:701-10
Dette, Holger; Pepelyshev, Andrey; Wong, Weng Kee (2012) Optimal designs for composed models in pharmacokinetic-pharmacodynamic experiments. J Pharmacokinet Pharmacodyn 39:295-311
Dette, Holger; Pepelyshev, Andrey; Wong, Weng Kee (2011) Optimal experimental design strategies for detecting hormesis. Risk Anal 31:1949-60
Holland-Letz, Tim; Dette, Holger; Pepelyshev, Andrey (2011) A geometric characterization of optimal designs for regression models with correlated observations. J R Stat Soc Series B Stat Methodol 73:239-252
López Fidalgo, J; Ortiz Rodríguez, I M; Wong, Weng Kee (2011) Design issues for population growth models. J Appl Stat 38:501-512
Wong, Weng Kee (2010) Discussion of Dette and Trampisch's paper ""A general approach to D-optimal designs for weighted univariate polynomial regression models"" J Korean Stat Soc 39:27-30
Bretz, Frank; Dette, Holger; Pinheiro, Jose C (2010) Practical considerations for optimal designs in clinical dose finding studies. Stat Med 29:731-42
Phoa, Frederick K H; Xu, Hongquan; Wong, Weng Kee (2009) The use of nonregular fractional factorial designs in combination toxicity studies. Food Chem Toxicol 47:2183-8
Crespi, Catherine M; Wong, Weng Kee; Mishra, Shiraz I (2009) Using second-order generalized estimating equations to model heterogeneous intraclass correlation in cluster-randomized trials. Stat Med 28:814-27
Chen, Ray-Bing; Wong, Weng Kee; Li, Kun-Yu (2008) Optimal minimax designs over a prespecified interval in a heteroscedastic polynomial model. Stat Probab Lett 78:1914-1921

Showing the most recent 10 out of 13 publications