Cousins: Charles Darwin, Sir Francis Galton and the birth of eugenics

Sir Francis Galton, scientist, African Explorer and statistician, was a key figure in statistical history. He was the man who devised the statistical concepts of regression and correlation. He was also Charles Darwin's cousin. And, inspired by his reading of Darwin, he was the founder of eugenics: the science of improving the human race through selective breeding. Nicholas Gillham tells of a darker side to statistics and heredity.     

Grey seals red in tooth and claw: how Darwin helps model their population

Grey seals in Britain are a conservation success, brought back from the brink of extinction by protection in the early 20th century. The public love them. Fishermen hate them. The resurgence of the grey guzzlers threatens their livelihood, they say. Len Thomas shows how Charles Darwin's ideas are used to determine seal numbers and population dynamics. Surprisingly, Darwin's theories have also influenced the statistical tools that are used.     

A genealogy of Florence Nightingale, Charles Darwin, Francis Galton and Francis Ysidro Edgeworth with special reference to their Italian connections and an annexe on Beatrice Webb and Charles Booth

In this article we show that several leading natural scientists, statisticians and social scientists born between 1730 and 1930 are closely related by marriage, thereby forming what Annan (Studies in social history: a tribute to C. M. Trevelyan, Longmans, Green, London, pp 241–287, 1955) has named an Intellectual Aristocracy. We also establish that the first three individuals mentioned in our title had family connections with Italy.     

Basic aspects of evolution strategies

Evolution strategies (ESs) are a special class of probabilistic, direct, global optimization methods. They are similar to genetic algorithms but work in continuous spaces and have the additional capability of self-adapting their major strategy parameters. This paper presents the most important features of ESs, namely their self-adaptation, as well as their robustness and potential for parallelization which they share with other evolutionary algorithms.Besides the early (1 + 1)-ES and its underlying theoretical results, the modern ( + )-ES and (, )-ES are presented with special emphasis on the self-adaptation of strategy parameters, a mechanism which enables the algorithm to evolve not only the object variables but also the characteristics of the probability distributions of normally distributed mutations. The self-adaptation property of the algorithm is also illustrated by an experimental example.The robustness of ESs is demonstrated for noisy fitness evaluations and by its application to discrete optimization problems, namely the travelling salesman problem (TSP).Finally, the paper concludes by summarizing existing work and general possibilities regarding the parallelization of evolution strategies and evolutionary algorithms in general.     

Professional development of statisticians in the pharmaceutical sector: evolution over the past decade and into the future

The adoption of The International Conference on Harmonization Tripartite Guideline: Statistical Principles for Clinical Trials (ICH-E9) has provided a foundation for the application of statistical principles in clinical research and raised awareness of the value of a statistical contribution to the wider pharmaceutical R&D process. In addition, over the past decade globalization of the pharmaceutical R&D process and the measures taken to address reduced productivity and spiralling costs have impacted on the roles and career opportunities for statisticians working in the pharmaceutical sector. This has enhanced the need for continuing professional development to equip statisticians with the skills to fully contribute to creating innovative solutions. In the future, key areas of focus are the establishment of professional standards for statistical work and increasing the collaboration between statisticians working in industry, regulatory agencies and academia. In addition, the diversity of roles and potential career paths for statisticians embarking on a career in the pharmaceutical sector emphasizes the importance of mentoring and coaching. For the more experienced statisticians, there are unprecedented opportunities to lead and innovate.     

Modelling the evolution of distributions: an application to Major League baseball

Summary.  We develop Bayesian techniques for modelling the evolution of entire distributions over time and apply them to the distribution of team performance in Major League baseball for the period 1901–2000. Such models offer insight into many key issues (e.g. competitive balance) in a way that regression-based models cannot. The models involve discretizing the distribution and then modelling the evolution of the bins over time through transition probability matrices. We allow for these matrices to vary over time and across teams. We find that, with one exception, the transition probability matrices (and, hence, competitive balance) have been remarkably constant across time and over teams. The one exception is the Yankees, who have outperformed all other teams.     

Modelling the evolution of a random set utilizing covariate information with applications to nowcasting

ABSTRACT

We propose models that allow us to capture the evolution of objects over time and more importantly, we provide forecasts to describe an object at future unobserved states utilizing information from the current state along with covariate information. We view objects as random sets and proceed to model them in a hierarchical Bayesian framework and estimate the model parameters using a Markov chain Monte Carlo scheme. We illustrate the methodology with an application to nowcasting of severe weather precipitation fields as obtained from weather radar images, where the severe storm cells are treated as random sets and the wind velocity is used to inform the distributions of the model parameters.     

The evolution of life expectancy in the capital city of Warsaw in 1931-1980

The author presents estimates of life expectancy in Warsaw, Poland, by age and sex for the period 1931-1980. The estimates involve modifications of previously calculated life tables for 1931 to ensure greater compatibility with more recent estimates.     

Functional approaches for predicting land use with the temporal evolution of coarse resolution remote sensing data

The sensor SPOT 4/Végétation gives every day satellite images of Europe with medium spatial resolution, each pixel corresponding to an area of 1 r km 2 1 r km. Such data are useful to characterize the development of the vegetation at a large scale. The pixels, named "mixed' pixels, aggregate information of different crops and thus different themes of interest (wheat, corn, forest, …). We aim at estimating the land use when observing the temporal evolution of reflectances of mixed pixels. The statistical problem is to predict proportions with longitudinal covariates. We compared two functional approaches. The first relies on varying-time regression models and the second is an extension of the multilogit model for functional data. The comparison is achieved on a small area on which the land use is known. Satellite data were collected between March and August 1998. The functional multilogit model gives better predictions and the use of composite vegetation index is more efficient.     

Geographic distance,the measurement of socioeconomic development,and the extent of migration

A study of migration in the region of Wojewodztwo Wloclawskie, Poland, is presented for the period 1975-1979 using a gravitational model. The model includes migration variables and variables measuring the level of socioeconomic development.     

Gosset,Fisher, and the t Distribution

Bias has different sources. Measurement errors create “bad” data and biased estimates. But selection biases occur even with “good” data and can be both subtle and large in magnitude.

Selection biases are not easily detected by internal examination of the data. Detection is more likely by comparison with external data sources.     

Randomization,balance, and the validity and efficiency of design-adaptive allocation methods

Few topics have stirred as much discussion and controversy as randomization. A reading of the literature suggests that clinical trialists generally feel randomization is necessary for valid inference, while biostatisticians using model-based inference often appear to prefer nearly optimal designs irrespective of any induced randomness. Dissection of the methods of treatment assignment shows that there are five basic approaches; pure randomizers, true randomizers, quasi-randomizers, permutation testers, and conventional modelers. Four of these have coherent design and analysis strategies, even though they are not mutually consistent, but the fifth and most prevalent approach (quasi-randomization) has little to recommend it. Design-adaptive allocation is defined, it is shown to provide valid inference, and a simulation indicates its efficiency advantage. In small studies, or large studies with many important prognostic covariates or analytic subgroups, design-adaptive allocation is an extremely attractive method of treatment assignment.     

A Note on the Variances of the Tests of Kendall,Jonckheere, and Terpstra

Recently, some papers have been appearing that have noticed some mistakes in the calculations in the statistical software in simple and classical tests (Wilcoxon–Mann-Whitney and Log-Rank). This article examines the variance formulas for the Kendall test, the Jonckheere test, and the Terpstra test. We show in this article that these variances (in spite of what is said in the literature and in the statistical software), in the presence of ties and r ≥ 3 groups, are only equal approximately and are exactly equal for two groups and ties as we prove in the Appendix.     

The demographic,social, and occupational distribution of the population

Some results from Poland's 1985 micro-census of population are presented. They concern population changes by age, sex, marital status, educational status, occupation, and income. The data are presented separately for rural and urban areas.     

Return rates in studies of life history evolution: Are biases large?

Studies of life history evolution in passerine birds often depend on examination of annual survival probability of adult birds. Most studies rely on return rates (proportion of marked individuals released in one year that are recaptured in the next year) to estimate annual survival probability. Yet, return rate includes both the probability of survival and the probability of recapturing or resighting the bird in the next time interval. We use numerical estimation to illustrate the increasing bias in return rate as an estimator of annual survival probability as recapture/resighting probability decreases. Recapture/resighting probability is normally assumed to be high and relatively invariant for recapture/resighting studies of color-banded territorial birds. We tested this assumption through examination of 11 color-banding studies of passerines. These studies showed that recapture/resighting probabilities vary strongly and cannot be generalized as high. In short, return rates generally are poor estimators of annual survival probabilities and use of return rates may strongly bias relationships explored in comparative studies or bias results of experiments to test survival costs of reproduction. Recapture/resighting probabilities should be estimated in all studies that attempt to estimate annual survival probabilities.     

Self-avoiding walks,the language of science,and Fibonacci numbers

The self-avoiding walk, restricted to a strip, is considered in the context of linguistic combinatorics.