Selection bias | EPFL Graph Search

Selection bias is the bias introduced by the selection of individuals, groups, or data for analysis in such a way that proper randomization is not achieved, thereby failing to ensure that the sample obtained is representative of the population intended to be analyzed. It is sometimes referred to as the selection effect. The phrase "selection bias" most often refers to the distortion of a statistical analysis, resulting from the method of collecting samples. If the selection bias is not taken into account, then some conclusions of the study may be false. Types Sampling bias Sampling bias is systematic error due to a non-random sample of a population, causing some members of the population to be less likely to be included than others, resulting in a biased sample, defined as a statistical sample of a population (or non-human factors) in which all participants are not equally balanced or objectively represented. It is mostly classified as a subtype of selection bias,

Empirical Evidence on the Effectiveness of Shareholder Democracy

Maxime Thibault B Couvert

This thesis is composed of two essays that study the effectiveness of shareholder democracy. In the first essay, I analyze the firm value implications of shareholder-initiated proposals. I show that managerial resistance precludes half of shareholder-initiated proposals from reaching the ballot stage. I construct a novel dataset of excluded and withdrawn proposals from the Securities and Exchange Commission's responses to managers' exclusion requests. An examination of announcement returns to the exclusion and withdrawal decisions reveals that non-voted proposals have a value-destroying nature. Special interest investors pursuing self-serving agendas and retail investors advocating for one-size-fits-all reforms explain the harmful character of non-voted proposals. I correct for the selection bias induced by managerial resistance and show that focusing only on voted proposals overstates the shareholder proposals-driven value creation. In the second essay, I study the impact of mutuals funds' environmental, social, and governance (ESG) preferences on portfolio firms. I construct measures of mutual funds' ESG preferences from their proxy voting guidelines in which mutual funds announce how they generally vote on the different ballot items at the shareholder meetings of their portfolio firms. I manually collect 17,000 of these voting policies for a sample of 29 of the largest U.S. mutual fund families over 2006-2018. I find that voting policies are a major predictor of funds' voting behavior. Exploiting staggered changes in funds' voting policies as an instrument, I show that investee companies adopt their mutual fund shareholders' preferred governance provisions. This adoption is the result of mutual fund shareholders' active voting. Announced voting policies also stimulate strategic proposal submissions by non-mutual fund shareholders.

EPFL2021

Practical issues with modeling extreme Brazilian rainfall

Paulo Vitor Da Costa Pereira, Anthony Christopher Davison, Isolde Santos Previdelli

Accurately quantifying extreme rainfall is important for the design of hydraulic structures, for flood mapping and zoning and for disaster management. In order to produce maps of estimates of 25-year rainfall return levels in Brazil, we selected 893 shorter and 104 longer rainfall time series from the Agencia Nacional de Aguas (ANA), and applied the framework of extreme value theory. Care was needed to reduce the impact of poor data. Estimates of the shape parameter of the extreme-value model fitted to rainfall data are typically biased, so we discuss an empirical correction that takes into account not only the sample-size bias, but also a so-called penultimate approximation that we use to inform a Bayesian spatial latent variable model for the annual rainfall maxima. This model accounts for subtle patterns of spatial variation in the data and provides plausible return level estimates.

BRAZILIAN STATISTICAL ASSOCIATION2021

Linking micro and macroevolution in the presence of migration

Sophie Myriam Hautphenne, Laurent Lehmann, Nicolas Salamin

Understanding macroevolutionary patterns is central to evolutionary biology. This involves the process of divergence within a species, which starts at the microevolutionary level, for instance, when two sub populations evolve towards different phenotypic optima. The speed at which these optima are reached is controlled by the degree of stabilising selection, which pushes the mean trait towards different optima in the different subpopulations, and ongoing migration that pulls the mean phenotype away from that optimum. Traditionally, macro phenotypic evolution is modelled by directional selection processes, but these models usually ignore the role of migration within species. Here, our goal is to reconcile the processes of micro and macroevolution by modelling migration as part of the speciation process. More precisely, we introduce an Ornstein-Uhlenbeck (OU) model where migration happens between two subpopulations within a branch of a phylogeny and this migration decreases over time as it happens during speciation. We then use this model to study the evolution of trait means along a phylogeny, as well as the way phenotypic disparity between species changes with successive epochs. We show that ignoring the effect of migration in sampled time-series data biases significantly the estimation of the selective forces acting upon it. We also show that migration decreases the expected phenotypic disparity between species and we analyse the effect of migration in the particular case of niche filling. We further introduce a method to jointly estimate selection and migration from time-series data. Our model extends traditional quantitative genetics results of selection and migration from a microevolutionary time frame to multiple speciation events at a macroevolutionary scale. Our results further support that not accounting for gene flow has important consequences in inferences at both the micro and macroevolutionary scale. (C) 2019 The Authors. Published by Elsevier Ltd.

ACADEMIC PRESS LTD- ELSEVIER SCIENCE LTD2020

Linking micro and macroevolution in the presence of migration

Sophie Myriam Hautphenne, Laurent Lehmann, Nicolas Salamin

ACADEMIC PRESS LTD- ELSEVIER SCIENCE LTD2020