Sergio Pozuelo-Campos

Un siglo separa la prueba de las tazas de té, nacida de un reto entre amigos en la estación experimental de Rothamsted, de los experimentos modernos. Lo que comenzó como una apuesta informal acabaría transformando la manera de hacer ciencia. El desafío, sin embargo, permanece intacto: distinguir los efectos reales de las simples fluctuaciones del azar.

Soft materials represent an interdisciplinary frontier in modern science, combining theoretical and experimental knowledge from diverse fields in both fundamental research and cutting‐edge technological applications. Hydrogels have garnered significant interest due to their unique properties. Nevertheless, practitioners often lack sufficient information on how to synthesize hydrogels, and the final composition is traditionally determined through trial and error. This study introduces a novel methodology to identify the optimal composition prior to experimentation. An ordinal regression model for mixture experiments to simulate the synthesis process, an adaptation of the Particle Swarm Optimization algorithm to tackle the complex optimization problem, and an open‐access interactive app to facilitate the calculations are proposed. This work represents a milestone in the hydrogel synthesis field: first, because of the savings in human, economic, and material resources it entails, the latter being consistent with the commitment to environmental sustainability; second, because it aligns with the principles of open science, providing an accessible and reproducible tool for the entire scientific community; and finally, because it seeks to overcome the barrier of expert knowledge to generalize the use of hydrogels. It represents the first step toward the development of an AI system specialized in the design of novel smart materials.

It is common in the optimal design literature to assume that the model has no random effects and it is known, apart from model parameters. Random effects models are widely applied across all disciplines, particularly in the life sciences and clinical studies. This paper assumes there are several plausible models and we want to find a design that optimally discriminates among models with random effects in them. A common design criterion for discriminating between models with fixed effects is the Kullback-Leibler (KL) divergence criterion. It is a maximin-type of criterion, not differentiable and is a 3-layer nested optimization problem over very distinct domains. Consequently, optimal designs for discriminating models are notoriously difficult to determine, not only analytically, but also computationally. This paper provides theoretical results for the KL-optimality criterion value for discriminating among random effects models and implements a nature-inspired metaheuristic algorithm to facilitate the search for an optimal discrimination design. Our methodology is quite general and applies to discriminating random effects models with multiple interacting experimental conditions which may be continuous or discrete. We provide two applications; the first finds a design that optimally discriminates among fractional polynomials with a single continuous variable, and the second identifies the best design to discriminate among several multi-factor random effects models.

The rapid spread of the COVID-19 worldwide led to the migration of the traditional education system based on the face-to-face classroom into an improvised online system, among many other preventive measures. Thus, all teaching methods had to be adapted to this new modality. This work is aimed at studying the viability of the online teaching of the subject of Applied Statistics in Health Sciences in higher education based on the teaching experience lived during COVID-19. In addition to this, possible technological difficulties and COVID-19-derived problems were investigated. A retrospective observational cross-sectional study was performed to analyze the students’ satisfaction according to the teaching methodologies in both face-to-face and online modalities. An exploratory and inferential analysis revealed that online teaching is feasible for the subject under study, although face-to-face learning still continues to significantly revert in favor of the quality of teaching. Therefore, further research is required to develop new online teaching methods given the feasibility of the proposal found in this research. Most of the students reported not having technological learning difficulties, whether related to their connectivity or technological resources, which did not have a significative impact on their teaching perception. Despite the psychological sequalae of COVID-19, this did not affect the students’ teaching satisfaction.

In optimal experimental design theory it is usually assumed that the response variable follows a normal distribution with constant variance. However, some works assume other probability distributions based on additional information or practitioner’s prior experience. The main goal of this paper is to study the effect, in terms of efficiency, when misspecification in the probability distribution of the response variable occurs. The elemental information matrix, which includes information on the probability distribution of the response variable, provides a generalized Fisher information matrix. This study is performed from a practical perspective, comparing a normal distribution with the Poisson or gamma distribution. First, analytical results are obtained, including results for the linear quadratic model, and these are applied to some real illustrative examples. The nonlinear 4-parameter Hill model is next considered to study the influence of misspecification in a dose-response model. This analysis shows the behavior of the efficiency of the designs obtained in the presence of misspecification, by assuming heteroscedastic normal distributions with respect to the D-optimal designs for the gamma, or Poisson, distribution, as the true one.