Article
Automation & Control Systems
Mario Becerra, Peter Goos
Summary: Discrete choice experiments are commonly used to quantify consumer preferences, but designing choice experiments involving mixtures of ingredients has been largely neglected. The I-optimality criterion is more suitable for precise predictions in experiments with mixtures, as it focuses on estimated statistical models. In this research, Bayesian I-optimal designs are shown to outperform their Bayesian D-optimal counterparts in terms of predicted utility variance.
CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS
(2021)
Article
Food Science & Technology
Mario Becerra, Peter Goos
Summary: This article demonstrates how to quantify and model consumer preferences for food products that are viewed as processed mixtures, by considering the impact of both mixture ingredient proportions and process variable settings. The combination of mixture-process variable experiments and discrete choice experiments is used to achieve this. D- and I-optimal designs for choice experiments involving mixtures and process variables are also generated and compared using examples.
FOOD QUALITY AND PREFERENCE
(2023)
Article
Management
Martin Szydlowski
Summary: The study explores the optimal disclosure and financing policies for an entrepreneur funding a project with uncertain cash flows. It suggests that truthfully revealing project cash flows above a certain threshold is optimal in the Bayesian persuasion framework. The entrepreneur's financing choice is influenced by a trade-off between persuading investors and relinquishing cash flow rights. If adverse selection occurs post-disclosure, a pooling equilibrium may be the unique outcome.
MANAGEMENT SCIENCE
(2021)
Article
Statistics & Probability
Tamar Haizler, David M. Steinberg
Summary: This study uses Bayesian probit regression to model factor effects and constructs sequential designs that embed attractive features of estimation and exploitation in online experiments.
Article
Computer Science, Interdisciplinary Applications
Yuna Zhao, Dennis K. J. Lin, Min-Qian Liu
Summary: The study focused on efficient fractional OofA designs in the field of order-of-addition designs, introducing a systematic construction method for a class of OofA orthogonal arrays (OofA-OAs) and demonstrating their superiority over other fractional OofA designs for the predominant pair-wise ordering (PWO) model, as well as their balance property. Additional research was conducted on the capacity of OofA-OAs for estimating different models.
COMPUTATIONAL STATISTICS & DATA ANALYSIS
(2022)
Article
Mathematics, Applied
Duc-Lam Duong, Tapio Helin, Jose Rodrigo Rojo-Garcia
Summary: We study the stability properties of the expected utility function in Bayesian optimal experimental design. We provide a framework for this problem in a non-parametric setting and prove the convergence rate of the expected utility with respect to a likelihood perturbation. The assumptions set out for the general case are satisfied in the specific case of non-linear Bayesian inverse problems with Gaussian likelihood, and the stability of the expected utility with respect to perturbations to the observation map is regained. Numerical simulations are used to demonstrate the theoretical convergence rates in three different examples.
Article
Economics
Anthony J. Hatswell
Summary: This article presents a method for incorporating weights into the synthesis of health-state utility values (HSUVs), allowing more relevant studies to have greater influence. Using four case studies, the results showed that the single preferred value (SPV) approach had differences compared to meta-analysis.
Article
Statistics & Probability
Jonathan Stallrich, Katherine Allen-Moyer, Bradley Jones
Summary: Traditionally, optimal screening designs for arbitrary run sizes are generated using the D-criterion with fixed factor settings at +/-1, but this article identifies cases where D-optimal designs have undesirable estimation variance properties. It argues that A-optimal designs generally minimize variances closer to their minimum value. New insights about the criteria are gained through studying their coordinate-exchange formulas. The study confirms the existence of D-optimal designs with only +/-1 settings for main effect and interaction models in both blocked and unblocked experiments. Arbitrary manipulation of a coordinate between [-1,1] can result in infinitely many D-optimal designs with different variance properties.
Article
Computer Science, Interdisciplinary Applications
Saubhagya S. Rathore, Grace E. Schwartz, Scott C. Brooks, Scott L. Painter
Summary: In this study, a Bayesian joint-fitting scheme is proposed to calibrate the entire biogeochemical model at once by simultaneously fitting all available datasets using the MCMC method. The joint fitting of datasets allows for complete uncertainty propagation and parameter estimates informed by all available data.
ENVIRONMENTAL MODELLING & SOFTWARE
(2022)
Article
Mathematics
Dongying Wang, Sumin Wang
Summary: For order-of-addition experiments, the response is affected by the addition order of the experimental materials. A predictive model and optimal designs are of interest for optimizing the response. The paper introduces D-, A-, and M.S.-optimal fractional PWO designs and proposes an efficient algorithm for generating these designs. Numerical simulation results show that the generated designs have high efficiency compared to the full PWO design, with only a small fraction of runs.
Article
Multidisciplinary Sciences
R. Alshenawy, Hanan Haj Ahmad, Ali Al-Alwan
Summary: This paper discusses two prediction methods for predicting the non-observed units under progressive Type-II censored samples, and provides inference on the unknown parameters of the Marshall-Olkin model. Through simulation studies and evaluation on real data examples, the best prediction method is found.
Article
Mathematics
Belmiro P. M. Duarte
Summary: We propose mixed-integer semidefinite programming formulations for finding exact optimal designs for linear models and locally optimal designs for nonlinear models. The strategy involves generating candidate treatments, formulating the optimal design problem as a mixed-integer semidefinite program, and solving it using appropriate solvers. We also use semidefinite programming-based formulations to find equivalent approximate optimal designs for comparison.
Article
Mathematical & Computational Biology
Cyrus Mehta, Apurva Bhingare, Lingyun Liu, Pralay Senchaudhuri
Summary: This paper develops optimal decision rules for sample size re-estimation in two-stage adaptive group sequential clinical trials. The initial sample size specification of such trials is usually adequate to detect a realistic treatment effect with good power but insufficient to detect the smallest clinically meaningful treatment effect. It is difficult for sponsors to make an up-front commitment to power a study adequately to detect the smallest effect. If the interim data enter a promising zone, it is easier to justify increasing the sample size.
STATISTICS IN MEDICINE
(2022)
Article
Oncology
Emily C. Zabor, Alexander M. Kaizer, Nathan A. Pennell, Brian P. Hobbs
Summary: This study proposes three randomized designs for early phase biomarker-guided oncology clinical trials, using optimal efficiency predictive probability method to monitor multiple biomarker subpopulations for futility. The simulation study demonstrates that potentially smaller phase II trials can be designed efficiently using randomization and futility stopping to obtain more information before phase III studies.
FRONTIERS IN ONCOLOGY
(2022)
Article
Statistics & Probability
Mingyao Ai, Zhiqiang Ye, Jun Yu
Summary: In this study, we investigate the locally D-optimal designs for models with general link functions under the partial proportional odds assumption. We derive the necessary and sufficient conditions for the positive definiteness of the Fisher information matrix and propose an efficient algorithm to search for optimal designs that can handle both discrete and continuous design fields. Numerical examples demonstrate the advantages of the proposed designs over existing ones.
Article
Economics
Jeroen Luyten, Sandy Tubeuf, Roselinde Kessels
Summary: Despite direct inquiries showing that respondents prioritize essential workers, elderly individuals, and those with pre-existing conditions, observations of competitive choices among individuals for vaccine allocation revealed different preferences, with a lower priority for elderly individuals and respondents divided into two clusters.
Article
Public, Environmental & Occupational Health
Niek Mouter, Annamarie de Ruijter, G. Ardine de Wit, Mattijs S. Lambooij, Maarten van Wijhe, Job van Exel, Roselinde Kessels
Summary: The willingness to take a COVID-19 vaccine is high among adults in the Netherlands, but a considerable proportion prefers to delay their decision to vaccinate until experiences of others are known. Age, education level, and vaccine attributes influence the willingness to vaccinate.
SOCIAL SCIENCE & MEDICINE
(2022)
Article
Engineering, Chemical
Ines Nulens, Rasheda Peters, Rhea Verbeke, Douglas M. Davenport, Cedric Van Goethem, Bart De Ketelaere, Peter Goos, Kumar Varoon Agrawal, Ivo F. J. Vankelecom
Summary: This study evaluates the influence of interfacial polymerization on the morphology and performance of polyamide thin film composite (TFC) membranes, considering varying monomer concentrations and organic phases with different physico-chemical characteristics. Two new descriptors, 'MPD supply' and 'TMC supply', are introduced to describe the synthesis-structure-performance relationship of polyamide TFC membranes. The study finds that well-performing membranes can be prepared until one of the monomers is added in excess, which depends on the interplay of MPD and TMC supply. The insights provided in this study can contribute to reducing the carbon footprint of reverse osmosis membrane synthesis and operation.
JOURNAL OF MEMBRANE SCIENCE
(2023)
Article
Engineering, Industrial
Mohammed Saif Ismail Hameed, Jose Nunez Ares, Peter Goos
Summary: Experimental data are often highly structured and allow for tailored analysis methods, such as orthogonal minimally aliased response surface (OMARS) designs. In this work, we improve upon existing methods and generalize the analysis framework for the entire family of OMARS designs. Through extensive simulations, we demonstrate the effectiveness of our customized method in identifying active effects.
JOURNAL OF QUALITY TECHNOLOGY
(2023)
Article
Health Care Sciences & Services
Jeroen Luyten, Roselinde Kessels
Summary: This study replicated a DCE experiment to investigate the impact of the COVID-19 pandemic on preferences for vaccine priority setting. The results showed that during the pandemic, people attached less importance to the number of vaccine-preventable deaths compared to before, while the importance of the vaccine's contribution to disease eradication and certainty about vaccine effectiveness increased. This suggests that the pandemic had some influence on preferences, although the impact was relatively small.
MEDICAL DECISION MAKING
(2023)
Article
Computer Science, Theory & Methods
Alan R. Vazquez, Weng Kee Wong, Peter Goos
Summary: This article presents the application of two-level screening designs in the manufacturing industry and proposes methods for constructing two-level Q(B)-optimal designs, including a mixed-integer programming algorithm and a heuristic algorithm. Numerical experiments show that these methods are effective in finding both small and large designs.
STATISTICS AND COMPUTING
(2023)
Article
Engineering, Industrial
Jose Nunez Ares, Peter Goos
Summary: The family of orthogonal minimally aliased response surface (OMARS) designs includes traditional response surface designs and definitive screening designs. The key features of OMARS designs are orthogonality for main effects and no aliasing with two-factor interaction effects or quadratic effects. This article presents a method to block OMARS designs in order to estimate main effects independently and confound interaction and quadratic effects as little as possible. The new blocking method for OMARS designs offers flexibility in choosing the number of runs, blocks, and block sizes, often outperforming existing arrangements in literature and software.
JOURNAL OF QUALITY TECHNOLOGY
(2023)
Article
Statistics & Probability
Alexandre Bohyn, Eric D. Schoen, Peter Goos
Summary: Designs for screening experiments often have factors with two levels only. By adding four-level factors, we can include multi-level categorical factors or quantitative factors with quadratic or third-order effects. We used three methods to generate a catalogue of designs with both two-level and four-level factors. Comparing the efficiencies of the methods, we demonstrated the usefulness of the catalogue by revisiting the motivating examples.
JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES C-APPLIED STATISTICS
(2023)
Article
Statistics & Probability
Arno Strouwen, Bart M. M. Nicolai, Peter Goos
Summary: Current experimental design techniques for dynamical systems often only consider measurement noise and not process noise. The Fisher information matrix is used to quantify the information content of experimental designs, but it depends on unknown model parameters. This paper combines Bayesian experimental design and adaptive experimental design methods to develop a robust methodology that averages the Fisher information matrix over a prior distribution and updates information as measurements are gathered.
STATISTICAL PAPERS
(2023)
Article
Economics
Alvaro A. Gutierrez-Vargas, Michel Meulders, Martina Vandebroek
Summary: This article explores the use of a model-based recursive partitioning algorithm to model preference heterogeneity. The algorithm constructs a decision tree based on statistical tests of preference parameter stability and uses Mixed Logit (MIXL) model with alternative-specific attributes at the tree's leaves. Simulation results demonstrate the algorithm's ability to recover diverse tree-like data generating processes with structural breaks in taste parameters. Furthermore, the algorithm outperforms MIXL in fitting stated choice data for environmental impact preferences and offers information for variable selection in Latent Class (LC) models.
JOURNAL OF CHOICE MODELLING
(2023)
Article
Education, Scientific Disciplines
Byran J. Smucker, Nathaniel T. Stevens, Jacqueline Asscher, Peter Goos
Summary: The design and analysis of experiments (DOE) is a crucial part of statistics education, especially with the increasing complexity of modern production processes and large-scale online experiments. This article provides an extensive review of DOE pedagogy and presents five perspectives on the subject, including one derived from a survey of DOE instructors. It offers insights into current teaching approaches and potential future developments of DOE instruction.
JOURNAL OF STATISTICS AND DATA SCIENCE EDUCATION
(2023)
Article
Food Science & Technology
Eline Van Wayenbergh, Niels A. Langenaeken, Nore Struyf, Peter Goos, Imogen Foubert, Christophe M. Courtin
Summary: Food fortification is an effective strategy against vitamin A deficiency, and cereal bran can be used as a natural stabilising agent. The stabilising effect of wheat bran on vitamin A is stronger for samples with high antioxidant capacity, high bound lipid content, and low lipase activity. Wheat bran antioxidants and bound lipids protect vitamin A from degradation during storage, while endogenous lipase activity counteracts the stabilising effect.
FOOD RESEARCH INTERNATIONAL
(2023)
Article
Hospitality, Leisure, Sport & Tourism
Mario Becerra, Matteo Balliauw, Peter Goos, Bruno De Borger, Benjamin Huyghe, Thomas Truyts
Summary: Analyzing the determinants of fans' willingness-to-pay for tickets is important for football clubs and federations to optimize their ticket offering and increase revenue.
INTERNATIONAL JOURNAL OF SPORTS MARKETING & SPONSORSHIP
(2023)
Article
Economics
Samson Yaekob Assele, Michel Meulders, Martina Vandebroek
Summary: This paper proposes new approaches for determining the sample size in discrete choice experiment (DCE) studies. The new rule of thumb and regression-based method improve the estimation of sample size compared to existing approaches, especially for large settings.
JOURNAL OF CHOICE MODELLING
(2023)
Article
Economics
Thanh Xuan Hua, Roselinde Kessels, Guido Erreygers
Summary: Receiving remittances has a positive impact on household savings and expenditure patterns in Vietnam. It increases the amount and rate of saving, and leads to more spending on health, assets, and house repairs.