On Presenting Results of Statistical Analyses in Scientific Journals

In many biological and other scientific journals, a reader's understanding of a paper to the argument of which statistical methods and analyses are important is often impeded by confusions of terminology and ambiguities of symbols. This is not solely because statistics is a difficult subject for biologists! If an editor were to formulate and make known a code of statistical symbols, abbreviations, and technical terms that in his journal will be regarded as part of the normal language of science, an author could use these without need for explanation each time. Every author would remain free to depart from the code, provided that he defined clearly his own usages. Such a policy, supported by the journal's referees, would do much to remove the frequent necessity for a reader to guess an author's meaning. Similarly considerations apply to the use of statistical software packages, where there is an evident need for an author to declare what software (if any) he has used, in much the same way as, by established custom, he will carefully specify his experimental materials and methods where these in any respect differ from the obvious. The present paper is written to stimulate constructive debate, and in no way to dogmatize on the merits or faults of particular statistical methods. Its underlying spirit is that the author of a scientific communication has a duty to describe the making of his observations, the conduct of his computations, and the performance of his computations with a clarity that would permit their repetition by another scientist who has access to the appropriate facilities and resources.     

Which statistics should tropical biologists learn?

Tropical biologists study the richest and most endangered biodiversity in the planet, and in these times of climate change and mega-extinctions, the need for efficient, good quality research is more pressing than in the past. However, the statistical component in research published by tropical authors sometimes suffers from poor quality in data collection; mediocre or bad experimental design and a rigid and outdated view of data analysis. To suggest improvements in their statistical education, we listed all the statistical tests and other quantitative analyses used in two leading tropical journals, the Revista de Biología Tropical and Biotropica, during a year. The 12 most frequent tests in the articles were: Analysis of Variance (ANOVA), Chi-Square Test, Student's T Test, Linear Regression, Pearson's Correlation Coefficient, Mann-Whitney U Test, Kruskal-Wallis Test, Shannon's Diversity Index, Tukey's Test, Cluster Analysis, Spearman's Rank Correlation Test and Principal Component Analysis. We conclude that statistical education for tropical biologists must abandon the old syllabus based on the mathematical side of statistics and concentrate on the correct selection of these and other procedures and tests, on their biological interpretation and on the use of reliable and friendly freeware. We think that their time will be better spent understanding and protecting tropical ecosystems than trying to learn the mathematical foundations of statistics: in most cases, a well designed one-semester course should be enough for their basic requirements.     

Regarding the confusion between the population concept and Mayr's "population thinking"

Ernst Mayr said that one of Darwin's greatest contributions was to show scholars the way to population thinking, and to help them discard a mindset of typological thinking. Population thinking rejects a focus on a central representative type, and emphasizes the variation among individuals. However, Mayr's choice of terms has led to confusion, particularly among biologists who study natural populations. Both population thinking and the concept of a biological population were inspired by Darwin, and from Darwin the chain for both concepts runs through Francis Galton who introduced the statistical usage of "population" that appears in Mayr's population thinking. It was Galton's "population" that was modified by geneticists and biometricians in the early 20th century to refer to an interbreeding and evolving community of organisms. Under this meaning, a population is a biological entity and so paradoxically population thinking, which emphasizes variation at the expense of dwelling on entities, is usually not about populations. Mayr did not address the potential for misunderstanding but for him the important part of the population concept was that the organisms within a population were variable, and so he probably thought there should not be confusion between population thinking and the concept of a population.     

Inflated impact factors? The true impact of evolutionary papers in non-evolutionary journals

Amongst the numerous problems associated with the use of impact factors as a measure of quality are the systematic differences in impact factors that exist among scientific fields. While in theory this can be circumvented by limiting comparisons to journals within the same field, for a diverse and multidisciplinary field like evolutionary biology, in which the majority of papers are published in journals that publish both evolutionary and non-evolutionary papers, this is impossible. However, a journal's overall impact factor may well be a poor predictor for the impact of its evolutionary papers. The extremely high impact factors of some multidisciplinary journals, for example, are by many believed to be driven mostly by publications from other fields. Despite plenty of speculation, however, we know as yet very little about the true impact of evolutionary papers in journals not specifically classified as evolutionary. Here I present, for a wide range of journals, an analysis of the number of evolutionary papers they publish and their average impact. I show that there are large differences in impact among evolutionary and non-evolutionary papers within journals; while the impact of evolutionary papers published in multidisciplinary journals is substantially overestimated by their overall impact factor, the impact of evolutionary papers in many of the more specialized, non-evolutionary journals is significantly underestimated. This suggests that, for evolutionary biologists, publishing in high-impact multidisciplinary journals should not receive as much weight as it does now, while evolutionary papers in more narrowly defined journals are currently undervalued. Importantly, however, their ranking remains largely unaffected. While journal impact factors may thus indeed provide a meaningful qualitative measure of impact, a fair quantitative comparison requires a more sophisticated journal classification system, together with multiple field-specific impact statistics per journal.     

Integrating animal temperament within ecology and evolution

Temperament describes the idea that individual behavioural differences are repeatable over time and across situations. This common phenomenon covers numerous traits, such as aggressiveness, avoidance of novelty, willingness to take risks, exploration, and sociality. The study of temperament is central to animal psychology, behavioural genetics, pharmacology, and animal husbandry, but relatively few studies have examined the ecology and evolution of temperament traits. This situation is surprising, given that temperament is likely to exert an important influence on many aspects of animal ecology and evolution, and that individual variation in temperament appears to be pervasive amongst animal species. Possible explanations for this neglect of temperament include a perceived irrelevance, an insufficient understanding of the link between temperament traits and fitness, and a lack of coherence in terminology with similar traits often given different names, or different traits given the same name. We propose that temperament can and should be studied within an evolutionary ecology framework and provide a terminology that could be used as a working tool for ecological studies of temperament. Our terminology includes five major temperament trait categories: shyness-boldness, exploration-avoidance, activity, sociability and aggressiveness. This terminology does not make inferences regarding underlying dispositions or psychological processes, which may have restrained ecologists and evolutionary biologists from working on these traits. We present extensive literature reviews that demonstrate that temperament traits are heritable, and linked to fitness and to several other traits of importance to ecology and evolution. Furthermore, we describe ecologically relevant measurement methods and point to several ecological and evolutionary topics that would benefit from considering temperament, such as phenotypic plasticity, conservation biology, population sampling, and invasion biology.     

Reducing ambiguity in describing plant–insect interactions: "preference", "acceptability" and "electivity"

Ecologists and evolutionary biologists have a common interest in plant–insect interactions. Ecologists develop terminology describing patterns of association between plants and insects, while evolutionary biologists use the same words to denote potentially heritable traits of individuals. Use of the same terms to describe both traits of the interaction and traits of the organisms hinders communication. An example is "preference", often used by ecologists to denote properties of the plant–insect interaction and by behavioural or evolutionary biologists to denote insect traits. The existing term "electivity" could be incorporated into the lexicon of plant–insect interactions to supplant the ecological use of "preference". The term "preference" would then denote a behavioural trait of the insect. The mirror-image trait of the plant would be "acceptability". This could be a step towards a common terminology that would be usable by both ecologists and evolutionists.     

Representing Statistical Distributions for Uncertain Parameters in LCA. Relationships between mathematical forms,their representation in EcoSpold,and their representation in CMLCA (7 pp)

Introduction Statistical information for LCA is increasingly becoming available in databases. At the same time, processing of statistical information is increasingly becoming easier by software for LCA. A practical problem is that there is no unique unambiguous representation for statistical distributions.- Representations. This paper discusses the most frequently encountered statistical distributions, their representation in mathematical statistics, EcoSpold and CMLCA, and the relationships between these representations.- The distributions. Four statistical distributions are discussed: uniform, triangular, normal and lognormal.- Software and examples. An easy to use software tool is available for supporting the conversion steps. Its use is illustrated with a simple example.Discussion This paper shows which ambiguities exist for specifying statistical distributions, and which complications can arise when uncertainty information is transferred from a database to an LCA program. This calls for a more extensive standardization of the vocabulary and symbols to express such information. We invite suppliers of software and databases to provide their parameter representations in a clear and unambiguous way and hope that a future revision of the ISO/TS 14048 document will standardize representation and terminology for statistical information.     

Biometry and the IBS—Strength through Diversity

Summary .  The International Biometric Society (IBS) brings together members from a diversity of cultural backgrounds, organized into geographically based Regions and National Groups, and covering a diverse range of interests, in terms of both methodological topics and application areas. We briefly reflect on how the historical development of our science, society, and international conferences reflects this diversity, with a focus on the history of the British and Irish Region of the IBS. Then, by considering the cultural/geographical diversity of the society, and the scientific diversity of the society and biometricians, we identify both some strengths of the society (diverse topics for meetings arranged across the world, application of biometrical methods to diverse application areas, management of the society by members from a diversity of backgrounds) and also some current challenges (electronic delivery of journals and other information, the diversity of application areas addressed by members of the society, improving links with the scientific societies of those who motivate our research). Finally, we illustrate the diversity of scientific problems that each of us face in our roles as biometricians.     

Development of a software tool for in silico simulation of Escherichia coli using a visual programming environment

This study describes the development of a software tool, EcoSim, to assist users in implementing quantitative in silico simulation easily. It consists of four parts: extracellular environment and constraints setting mode, table for optimal metabolic flux distribution and chart for changes of substrate concentration, dynamic flux distribution viewer and dynamic hierarchical regulatory network viewer. Representation of a hierarchical regulatory network was constructed with defined modeling symbols and weight in the central Escherichia coli metabolism. All programming procedures for EcoSim were accomplished in a visual programming environment (LabVIEW). To illustrate quantitative in silico simulation with EcoSim, this program was performed on E. coli using glucose and acetate as carbon sources. The simulation results were in agreement with the experimental data obtained from the literature. EcoSim can be used to assist biologists and engineers in predicting and interpreting dynamic behaviors of E. coli under a variety of environmental conditions.     

Effect size, confidence interval and statistical significance: a practical guide for biologists

Null hypothesis significance testing (NHST) is the dominant statistical approach in biology, although it has many, frequently unappreciated, problems. Most importantly, NHST does not provide us with two crucial pieces of information: (1) the magnitude of an effect of interest, and (2) the precision of the estimate of the magnitude of that effect. All biologists should be ultimately interested in biological importance, which may be assessed using the magnitude of an effect, but not its statistical significance. Therefore, we advocate presentation of measures of the magnitude of effects (i.e. effect size statistics) and their confidence intervals (CIs) in all biological journals. Combined use of an effect size and its CIs enables one to assess the relationships within data more effectively than the use of p values, regardless of statistical significance. In addition, routine presentation of effect sizes will encourage researchers to view their results in the context of previous research and facilitate the incorporation of results into future meta-analysis, which has been increasingly used as the standard method of quantitative review in biology. In this article, we extensively discuss two dimensionless (and thus standardised) classes of effect size statistics: d statistics (standardised mean difference) and r statistics (correlation coefficient), because these can be calculated from almost all study designs and also because their calculations are essential for meta-analysis. However, our focus on these standardised effect size statistics does not mean unstandardised effect size statistics (e.g. mean difference and regression coefficient) are less important. We provide potential solutions for four main technical problems researchers may encounter when calculating effect size and CIs: (1) when covariates exist, (2) when bias in estimating effect size is possible, (3) when data have non-normal error structure and/or variances, and (4) when data are non-independent. Although interpretations of effect sizes are often difficult, we provide some pointers to help researchers. This paper serves both as a beginner's instruction manual and a stimulus for changing statistical practice for the better in the biological sciences.     

What is the subject of science "bioinformatics"?

The paper is concerned with some problems of terminology, in particular the term "bioinformatics". In the last few years, the term "bioinformatics" has been intensively used among molecular biologists to indicate a subject that is only a constituent of genomics and is considered to involve a computer-assisted analysis of all data on nucleotide sequences of DNA. However, a wide circle of scientists, including biologists, physicists, mathematicians, and specialists in the field of cybernetics, informatics, and other disciplines have accepted and accept, as a rule, the "bioinformatics" as a synonym of science cybernetics and as a successor of this science. In this case, the subject of science "bioinformatics" should embrace not only genomics but practically all sections of the biological science. It should involve a study of information processes (storage, transfer, and processing of information, etc.) participating in the regulation and control at all levels of living systems, from macromolecules to the brain of higher animals and human.     

Problems of terminology in the teaching of plant water relations

The considerable variation and confusion in terminology used in the teaching of plant water relations is discussed. The concept of water potential is described briefly and compared with the older concepts of diffusion pressure deficit and suction pressure. The advantages of using water potential terminology are considered. Reference is made to some well-known student texts to illustrate the conflicting meanings of osmotic terms and symbols, and attention is drawn to the problems which arise when authors mix old and new terms. It is recommended that the new terminology should be incorporated into school examination syllabuses and textbooks as soon as possible.     

Sampling strategies for distances between DNA sequences

An international effort is now underway to obtain the DNA sequence for the entire human genome (Watson and Jordan, 1989, Genomics 5, 654-656; Barnhart, 1989, Genomics 5, 657-660). This Human Genome Initiative will generate sequence data from several species other than humans, and will result in several copies per species of at least some regions of the genome. Although the project has generated much interest, it is but one aspect of the widespread effort to generate DNA sequence data. Published sequences are collected in common databases, and release 63 of GenBank in March 1990 contained 40,127,752 bases from 33,337 reported sequences (News from GenBank 3; Mountain View, California: Intelligenetics, Inc., 1990). Large though this database is, it is only about 1% of the number of bases in the human genome. Interpretations of data of such magnitude are going to require the collaborative efforts of biometricians and molecular biologists, and an aim of this paper is to show that there is also a role for readers of this journal in the design of surveys of DNA sequences. Discussion here will center on the use of sequence data in evolutionary studies, where some region of DNA is sequenced in several different species. The object is to infer the evolutionary history of that particular region, or of the species themselves. Statistical issues in the very important studies on sequences to locate and characterize regions responsible for human diseases will not be addressed here. We will discuss appropriate ways of measuring distances between DNA sequences and of predicting the sampling properties of the distances. There are procedures for inferring evolutionary histories for a set of elements that depend on a matrix of distances between each pair of elements, and the precision of resulting trees must be influenced by the precision of the distances. We will show that account needs to be taken of two sampling processes--the sampling of sequences by the investigator ("statistical sampling"), and the sampling of genetic material involved in the formation of offspring from a parental population ("genetic sampling").     

The state of the art

Graphical representation of molecular and cellular features is a key form of communication in structural biology in which abstract symbols of an economical and visually appropriate kind, pseudo-color coding, and dynamic animations all play their parts. Accordingly, it should not be surprising that many structural biologists--like traditional biologists before them--are talented artists who also express themselves on "non-scientific" topics. This article illustrates the approaches of and pictures by several practicing scientist-artists-mainly, in this sampling, electron microscopists.     

The philosophy of modelling or does the philosophy of biology have any use?

Biologists in search of answers to real-world issues such as the ecological consequences of global warming, the design of species'' conservation plans, understanding landscape dynamics and understanding gene expression make decisions constantly that are based on a ‘philosophical’ stance as to how to create and test explanations of an observed phenomenon. For better or for worse, some kind of philosophy is an integral part of the doing of biology. Given this, it is more important than ever to undertake a practical assessment of what philosophy does mean and should mean to biologists. Here, I address three questions: should biologists pay any attention to ‘philosophy’; should biologists pay any attention to ‘philosophy of biology’; and should biologists pay any attention to the philosophy of biology literature on modelling? I describe why the last question is easily answered affirmatively, with the proviso that the practical benefits to be gained by biologists from this literature will be directly proportional to the extent to which biologists understand ‘philosophy’ to be a part of biology, not apart from biology.     

The ``Evolutionary Synthesis' of George Udny Yule

This article discusses the work ofGeorge Udny Yule in relation to theevolutionary synthesis and thebiometric-Mendelian debate. It has generallybeen claimed that (i.) in 1902, Yule put forththe first account showing that the competingbiometric and Mendelian programs could besynthesized. Furthermore, (ii.) the scientificfigures who should have been most interested inthis thesis (the biometricians W. F. RaphaelWeldon and Karl Pearson, and the MendelianWilliam Bateson) were too blinded by personalanimosity towards each other to appreciateYule's proposal. This essay provides adetailed account of (i.), maintaining thatYule's 1902 proposal is better understood as areduction, not a synthesis of the two programs.The results of this analysis are then used toevaluate (ii.), where I will instead argue thatBateson and the biometricians had good reasonsto avoid endorsing Yule's account.